There is some evidence that Apples approach doesn't work all that well either. Rumors of marginal Mini sales continue to surface and iMac sales seem to be the result of people not having a choice.
Yeah... that's the only reason I've considered an imac at any point. All in ones have a ton of issues. They really aren't built for heavy heavy use to the degree of a tower with superior airflow. The display quality is just ok. I'm guessing if I had to, I could get used to it but it's sort of a higher end consumer grade display. It's not really meant for those concerned with color and detail. Regarding the mac pro people who buy them often use them for a very long time. It's not so easy to do this when the hardware sits so long. The logic board is the same one used in 2009. The processor in the baseline model is a slight bump from that because intel discontinued the other and replaced the w3520 with the w3530. If I bought one of these today, I would be concerned about how long it will retain current support. Apple has been really inconsistent in this regard, so I don't think anything past the next OS is a real guarantee.
Quote:
Originally Posted by wizard69
Think about it a bit, most of Apples laptops are a better deal than the Mini.
In my limited testing many of the laptops offer better performance too. The mini is still invested into laptop parts so the cost to performance ratio is poor relative to desktop equivalents. The mini server brings you up to $1000 starting without keyboard or mouse (and back to integrated graphics). At that point the 21.5" imac at $1200 would be a better purchase overall. You lose the virtual core function of the i7 design but you're moving up to desktop processors and away from integrated graphics that share system memory.
Quote:
Originally Posted by wizard69
And no external storage boxes, especially over Thunderbolt, are not a solution.
You know I don't disagree with you here, but I'll end up with external storage regardless of how many bays they put in it. I simply said before that for my purposes if it at least had multiple thunderbolt ports of a later generation/higher bandwidth, I'd probably grudgingly go that route rather than continue to wait and see.
Yeah... that's the only reason I've considered an imac at any point. All in ones have a ton of issues. They really aren't built for heavy heavy use to the degree of a tower with superior airflow. The display quality is just ok. I'm guessing if I had to, I could get used to it but it's sort of a higher end consumer grade display. It's not really meant for those concerned with color and detail. Regarding the mac pro people who buy them often use them for a very long time. It's not so easy to do this when the hardware sits so long. The logic board is the same one used in 2009. The processor in the baseline model is a slight bump from that because intel discontinued the other and replaced the w3520 with the w3530. If I bought one of these today, I would be concerned about how long it will retain current support. Apple has been really inconsistent in this regard, so I don't think anything past the next OS is a real guarantee.
Sadly I agree. For my needs an iMac wouldn't be to bad if one could easily deal with the storage limitations. If you go the Thunderbolt route though you end up pouring out a pretty good chunk of change for that solution. If I had a choice I'd go with a PCI Express card supporting 10G Ethernet or Fiber Channel, but of course Apple does not have an economical solution for that.
Quote:
In my limited testing many of the laptops offer better performance too. The mini is still invested into laptop parts so the cost to performance ratio is poor relative to desktop equivalents. The mini server brings you up to $1000 starting without keyboard or mouse (and back to integrated graphics). At that point the 21.5" imac at $1200 would be a better purchase overall. You lose the virtual core function of the i7 design but you're moving up to desktop processors and away from integrated graphics that share system memory.
The problem with the Mini is that it is purposely limited. Even with Laptop chips they could go for more performance, instead they cheap out and provide midrange laptop performance. The cynic in me says they do that to entice people to buy laptops.
I'm not knocking the Mini as much as saying it is less than optimal for somebodies primary computer.
Quote:
You know I don't disagree with you here, but I'll end up with external storage regardless of how many bays they put in it. I simply said before that for my purposes if it at least had multiple thunderbolt ports of a later generation/higher bandwidth, I'd probably grudgingly go that route rather than continue to wait and see.
Well some do have that problem. Fortunately I do not, at least not at the moment need that much external storage. These day though adding 4TB of storage to a desktop should be easy.
Sadly I agree. For my needs an iMac wouldn't be to bad if one could easily deal with the storage limitations. If you go the Thunderbolt route though you end up pouring out a pretty good chunk of change for that solution. If I had a choice I'd go with a PCI Express card supporting 10G Ethernet or Fiber Channel, but of course Apple does not have an economical solution for that.
Is 10G or fiber channel really that cost effective in any form? Feel free to correct me here but I've never personally seen cheap cards for those standards. I imagine by next year we'll start to see more cost effective thunderbolt solutions and possibly the ability to retrofit backplanes to existing esata solutions on the dumb box end. Intel was very very very slow in releasing an SDK or any other materials to aid development of thunderbolt accessories, and PCs aren't really seeing those ports until next year. Give it a little time and it may become very usable. It's just that I'd like to see a minimum of two ports on a given machine. Daisy chaining is a really bad option for high bandwidth items like drive enclosures populated by SSDs and displays because either could consume a very large portion of its bandwidth in either direction. Displayport 1.2 actually supports 17.28 Gb/s of video bandwidth with thunderbolt supporting 10 in each direction. My point being that we really need more than a single port on any given machine if they are to be used for both external displays and storage.
Quote:
Originally Posted by wizard69
The problem with the Mini is that it is purposely limited. Even with Laptop chips they could go for more performance, instead they cheap out and provide midrange laptop performance. The cynic in me says they do that to entice people to buy laptops.
I'm not knocking the Mini as much as saying it is less than optimal for somebodies primary computer.
I agree with you there, and they do really go pretty tight on it especially with the lack of a keyboard and mouse. Should you desire the Apple versions, that's around $100 extra for the two items assuming you don't go wireless. I sometimes feel like they reverse engineer these things to hit a certain price point while retaining margins.
Quote:
Originally Posted by wizard69
Well some do have that problem. Fortunately I do not, at least not at the moment need that much external storage. These day though adding 4TB of storage to a desktop should be easy.
It is. 4TB is nothing though I've had single jobs generate 20GB worth of data. It also requires one or more backups.
Is 10G or fiber channel really that cost effective in any form? Feel free to correct me here but I've never personally seen cheap cards for those standards.
Cheap - no but they are well accepted standards. That means a storage server using the standards would be viable for some time. Interestingly I just saw a bit on a new PCI Express 3.0 card that implements several of these protocols in one card.
Quote:
I imagine by next year we'll start to see more cost effective thunderbolt solutions and possibly the ability to retrofit backplanes to existing esata solutions on the dumb box end. Intel was very very very slow in releasing an SDK or any other materials to aid development of thunderbolt accessories, and PCs aren't really seeing those ports until next year.
Slow! Turtles move faster than Intel here.
By the way unless Intel releases the standard publicly I not to sure how well it will go over in the long run. The using community has been burnt to many times in the past by proprietary standards, so unless Intel and Apple open this up I can't see mass acceptance.
Quote:
Give it a little time and it may become very usable. It's just that I'd like to see a minimum of two ports on a given machine. Daisy chaining is a really bad option for high bandwidth items like drive enclosures populated by SSDs and displays because either could consume a very large portion of its bandwidth in either direction. Displayport 1.2 actually supports 17.28 Gb/s of video bandwidth with thunderbolt supporting 10 in each direction. My point being that we really need more than a single port on any given machine if they are to be used for both external displays and storage.
Very true about port congestion! People still don't grasp the concept that this is a serial link. One good point though is that we have lines going in each direction. In the end though without additional ports I can see many a professional user getting a little perturbed with port.
Quote:
I agree with you there, and they do really go pretty tight on it especially with the lack of a keyboard and mouse. Should you desire the Apple versions, that's around $100 extra for the two items assuming you don't go wireless. I sometimes feel like they reverse engineer these things to hit a certain price point while retaining margins.
I don't think it is price point as much as it is resisting the temptation to exceed the performance levels of their laptops at anyone time. It is very much a conscious effort on their part. Let's face it, the Mini is a desktop machine there is no battery nor cooling limitation to speak of. Further the unit including the discreet GPU is just pathetic considering the extra cost and marginal GPU implementation.
Quote:
It is. 4TB is nothing though I've had single jobs generate 20GB worth of data. It also requires one or more backups.
Yeah the more that I think about it the more that I would agree. After all today that is just four low profile laptop drives. For me though it would be nice to have that in a base machine. This brings up another Mini issue, what is with the puny disk offerings? I can understand the entry level models but what about the options, it isn't like 1TB disks are unheard of in California or even 750GB models.
Oh about backups those should very much be on an external system. Internally though I want room for all my data and code. That isn't asking a lot and frankly easy these days for me to cover with laptop drives. That is if I had an XMac with a few drive bays.
Cheap - no but they are well accepted standards. That means a storage server using the standards would be viable for some time. Interestingly I just saw a bit on a new PCI Express 3.0 card that implements several of these protocols in one card.
Slow! Turtles move faster than Intel here.
By the way unless Intel releases the standard publicly I not to sure how well it will go over in the long run. The using community has been burnt to many times in the past by proprietary standards, so unless Intel and Apple open this up I can't see mass acceptance.
I thought they were going to do so at some point. That could have been a misinterpretation or wishful thinking on my part. Anyway yeah they're way too slow about it. They need people to feel like they're really behind it and provide a good environment for third party development. Just placing it on chipsets doesn't replace the need for a quality SDK.
Quote:
Originally Posted by wizard69
Very true about port congestion! People still don't grasp the concept that this is a serial link. One good point though is that we have lines going in each direction. In the end though without additional ports I can see many a professional user getting a little perturbed with port.
Yep....and they were very misleading about this. Apple and intel have pressed the idea that you can just hook up whatever you want to via a single thunderbolt port, and this really isn't the case.
Quote:
Originally Posted by wizard69
I don't think it is price point as much as it is resisting the temptation to exceed the performance levels of their laptops at anyone time. It is very much a conscious effort on their part. Let's face it, the Mini is a desktop machine there is no battery nor cooling limitation to speak of. Further the unit including the discreet GPU is just pathetic considering the extra cost and marginal GPU implementation.
You might be right. I thought the minis got quite hot though? Perhaps this has changed more than I realized. I just don't consider the quad/mini server option a truly viable alternative to other things in the surrounding price points as anything other than a light duty server. Quad core machines have basically become the norm at this point so the dual core version isn't a true consideration for me.
Quote:
Originally Posted by wizard69
Yeah the more that I think about it the more that I would agree. After all today that is just four low profile laptop drives. For me though it would be nice to have that in a base machine. This brings up another Mini issue, what is with the puny disk offerings? I can understand the entry level models but what about the options, it isn't like 1TB disks are unheard of in California or even 750GB models.
Oh about backups those should very much be on an external system. Internally though I want room for all my data and code. That isn't asking a lot and frankly easy these days for me to cover with laptop drives. That is if I had an XMac with a few drive bays.
If SSDs continue to drop in price per GB, I could see a compact machine with 2-4 2.5" bays to cut heat in a somewhat confined enclosure. They've offered this option on the mac pro, but it is still an extremely expensive route. It may take a few years but we'll see a further drop in the popularity of 2.5" HDDs. I agree regarding backups. I also believe it's important to maintain actual offline backups of critical data. eSATA type enclosures aren't that bad a solution. You just have to be very careful what you buy. You want one with powerful cooling, which often means having to blow out dust every few months. If you're on a budget you want to completely avoid raid solutions. Cheap raids suck, and many people don't understand how they work at all. I've spent a fair amount of time explaining to others why they have to back up their raid solutions and that parity striping will not save them in the event of data corruption, controller failure, or a power surge of any kind (including the power supply on the unit failing).
I thought they were going to do so at some point. That could have been a misinterpretation or wishful thinking on my part. Anyway yeah they're way too slow about it. They need people to feel like they're really behind it and provide a good environment for third party development. Just placing it on chipsets doesn't replace the need for a quality SDK.
They need to do something here to make TB a public standard. Right now it is too obscure to be considered a major player. As far as Intel goes you almost get the impression that the release was too early.
The problem with widespread implementation has already caused the USB people to consider developing an alternative. To this end I don't see TB being acceptable to the industry as a whole until you can get independent chipsets to implement the interface. More importantly the port needs to be implemented on micro controllers if reasonable prices are to be had. Further embedded people won't be willing to buy Intel hardware.
Quote:
Yep....and they were very misleading about this. Apple and intel have pressed the idea that you can just hook up whatever you want to via a single thunderbolt port, and this really isn't the case.
Sadly they will learn. However many users will do fine with the TB port as it's performance will depend upon the work load.
Quote:
You might be right. I thought the minis got quite hot though? Perhaps this has changed more than I realized.
I don't have a new one to test, however the last case redesign was a major one. It appears that one goal was much better cooling. That being said if cooling was an issue I don't think we would be seeing racks of these used as servers.
Now a server might not be as heavily loaded as you might in your usage. Given the right work load you can heat up just about any computer.
Quote:
I just don't consider the quad/mini server option a truly viable alternative to other things in the surrounding price points as anything other than a light duty server. Quad core machines have basically become the norm at this point so the dual core version isn't a true consideration for me.
I can understand the disappointment about the dual cores but that just goes to support my point, Apple castrated the machines to make their laptops look good. It is positively storage that they put a quad core in one model and then called it a server.
Quote:
If SSDs continue to drop in price per GB, I could see a compact machine with 2-4 2.5" bays to cut heat in a somewhat confined enclosure. They've offered this option on the mac pro, but it is still an extremely expensive route.
Apple needs to make use of one of the emerging PCI - Express standards for solid state storage cards. The goal should be to use commodity parts that leverage the fast interface of PCI-E. The use of cards should lead to smaller hardware and easier cooling.
Quote:
It may take a few years but we'll see a further drop in the popularity of 2.5" HDDs. I agree regarding backups. I also believe it's important to maintain actual offline backups of critical data. eSATA type enclosures aren't that bad a solution. You just have to be very careful what you buy. You want one with powerful cooling, which often means having to blow out dust every few months.
A good and cheap solution. If nothing else data should be backed up here.
Quote:
If you're on a budget you want to completely avoid raid solutions. Cheap raids suck, and many people don't understand how they work at all.
In the past I implemented RAIDs on some of my Linux machines. Software based Linux RAIDs can be a good low cost option but you are right you need to understand them. In my case implementing them was easy. Luck was on my side as I never had to rebuild a volume. In any event I didn't do the RAIDs back then for data security.
For a long while there I outgrew my hardware before failure.
Quote:
I've spent a fair amount of time explaining to others why they have to back up their raid solutions and that parity striping will not save them in the event of data corruption, controller failure, or a power surge of any kind (including the power supply on the unit failing).
Yep even data center RAIDs with all of their fancy protections go down. Sometimes for a very long time. Even a minor failure takes forever to rebuild.
Even backups of RAIDs can get corrupted though so back ups of critical data really requires multiple approaches that don't depend upon other systems. Sometimes a CD, disk or tape in a safe deposit box makes a lot of sense. However even that is a pain with a reasonably large RAID.
Back ups are very important but I'm not sure if people get it. In some of the forums I visit new college students will ask about what computer they should buy. For the most part that is easy, buy a Mac unless your program has specific requirements. I always stress the need to consider a backup program so that they don't suffer due to some sort of failure. What interests me is how many listen to the advice.
They need to do something here to make TB a public standard. Right now it is too obscure to be considered a major player. As far as Intel goes you almost get the impression that the release was too early.
The problem with widespread implementation has already caused the USB people to consider developing an alternative. To this end I don't see TB being acceptable to the industry as a whole until you can get independent chipsets to implement the interface. More importantly the port needs to be implemented on micro controllers if reasonable prices are to be had. Further embedded people won't be willing to buy Intel hardware.
If they did in fact make it easier/cheaper to implement, it could work out well. It remains significantly cheaper than some of the other high data bandwidth solutions. It's just that right now I think a single port which is also tied to display output is an issue. Display resolution will continue to increase significantly. Medical grade panels have achieved some pretty insane numbers in the past few years. Laptop resolution has shot way up. Right now the desktop panels are a bit stuck, but this will change. Look at how many people like the crisp look of the retina display and the number of magazines migrating to digital content delivery.
Quote:
Originally Posted by wizard69
I can understand the disappointment about the dual cores but that just goes to support my point, Apple castrated the machines to make their laptops look good. It is positively storage that they put a quad core in one model and then called it a server.
They put a quad core processor and weaker graphics in it which was also the only i7 variant in the line. It makes sense as a low power server really. The i7 variants have hyper threading enabled which would help such a machine process more simultaneous requests. If they built out/up a bit from the mini design, they could have a decent desktop machine. Marvin referenced something similar but I didn't like the example he picked (I felt the example of a machine that uses a very low wattage power source might not work out well if too many things are plugged into the machine).
Quote:
Originally Posted by wizard69
Apple needs to make use of one of the emerging PCI - Express standards for solid state storage cards. The goal should be to use commodity parts that leverage the fast interface of PCI-E. The use of cards should lead to smaller hardware and easier cooling.
That's a great idea. PCI-E is still a solid interface. I don't think everyone realizes that it doesn't have to be implemented entirely at the back of the computer.
Quote:
Originally Posted by wizard69
Back ups are very important but I'm not sure if people get it. In some of the forums I visit new college students will ask about what computer they should buy. For the most part that is easy, buy a Mac unless your program has specific requirements. I always stress the need to consider a backup program so that they don't suffer due to some sort of failure. What interests me is how many listen to the advice.
I agree with you here. Some things require me to boot into Windows and emulation doesn't deliver the required performance. When in Windows I tend to stay off the internet, and turn off anything that annoys me. This includes a bunch of cursor animations (if you use a graphics tablet the ripple effect every time you click on something is just way too annoying if you work as fast as I do) and Aero. Run like this it runs fine.
I don't feel like quoting your lengthy posts and go over every single point, but to recap:
You say Apple is missing out on a huge opportunity by not offering the xMac.
Fact: all computer sales are down and declining. Only Apple's sales grow. You seem to think that their sales grow in spite of their "incomplete" or restricted line-up. I reckon it is more logical to assume that their sales improve precisely because they don't offer the same muck that everybody else has been churning out forever.
And PC's aren't an investment anymore, not for the middle-class anyway. They have become a commodity. People have them repaired about as frequent as they have their shoes resoled.
If you want to know what is going to happen to the desktop, you only need to look at how 20-year olds use them. If it is not for gaming or some other specific activity, they don't have one and they don't want one. They have a laptop or netbook but they spent most of their time on their smartphones.
Fact: all computer sales are down and declining. Only Apple's sales grow. You seem to think that their sales grow in spite of their "incomplete" or restricted line-up. I reckon it is more logical to assume that their sales improve precisely because they don't offer the same muck that everybody else has been churning out forever.
The desktop form factor is still better for a lot of professionals. The guys that own machines like a mac pro, also own macbook pros and typically iphones. Having a solid option for users that need power and bandwidth does benefit Apple. Right now they haven't been paying much attention to the mac pro. I'd like to see them build a followup that they can be enthusiastic about today, because there will still be a market for it. They need something for those that work with multiple 1900x1200 professional grade displays (displays that cost more than an imac) and a multitude of hard drives. Apple has been all about ecosystems and this would fit in as an integral part if they just decide how to proceed. Otherwise it will get to the point where they try to kill off the mac pro and end up with another pr fiasco. Remember how they brought back firewire on one of their laptops some time ago and the recent FCP X thing? Having stuff like this helps those who need it maintain an all Apple setup.
Sadly I've been having to use bootcamp lately. I customized the living hell out of windows so that it wouldn't irritate me, and I still hate it.
Quote:
Originally Posted by zeph
If you want to know what is going to happen to the desktop, you only need to look at how 20-year olds use them. If it is not for gaming or some other specific activity, they don't have one and they don't want one. They have a laptop or netbook but they spent most of their time on their smartphones.
You're thinking too old. You should look at how teenagers use their phones and computers. Apple has always been very good at being a cool brand given the early associations with its popularity amongst graphic designers, photographers, and other creative professions, speaking of which many of those guys still own mac pros just because it allows for a stable computer with a large amount of ram. If they moved away from this design and built upward from the mac mini instead, I would totally buy it. If we could get one that used desktop processors because the high end laptop variant would be too expensive, at an average gpu, better ram capacity, and 2-3 second generation TB ports, we might have an extremely popular machine for all of these guys that want xmacs. You could keep it small enough to hold its appeal to the mini crowd but give it enough power to act like a true desktop. I'm not talking only about cpu, but about the machine as a whole.
Sorry if you felt left out, man. I use a MacPro but only for audio. I am amazed at how the iPhone has usurped so many of what was previously desktop duties.
For work nothing but a MacPro will do (or a similarly spec'd Windows workstation) and that's only because there is no xMac. If there was one, I would be dead in the middle of its target market. A 2600-i7 with a few hdd bays and pci slots would be sufficient for me, but alas.
Yet I understand Apple's strategy and I think they have proven to be right about this. Too bad for me, now I have to pay $1000 more for a MacPro or go with Windows. But for most people, an iMac suffices, even as the hub of their digital life. Outside of work, my iMac does all I need it to do. Shucks, a MacMini would probably suffice (albeit with an external superdrive).
I think I'm right about the kids, though. I see it all around me, especially girls. They use a netbook/laptop and a smartphone. And they don't like clunky boxes. For them the desktop is utterly irrelevant. There will always be a market for workstations/desktops but I do believe SJ is right in that the vast majority no longer needs them.
Now, in the absence of the xMac, I say that dropping the price of an entry-level MacPro to $1999 would constitute an acceptable compromise.
Good discussion, but to return to the main question, are any of you willing to venture some advice?
I'm currently running a Mac Pro 1,1 (2006) that is rapidly nearing the end of life, and I will need to replace it withing the next few months.
If I had to buy one today, my choice would be the 6-core Westmere with the Radeon 5870- I work in photography and graphic design, so more cores would be unused.
To be frank, after looking at Bare Feats, the high-end iMac appears to rival the Pro in performance, and I'd be tempted, but I'm not going to abandon my investment in displays.
So the question I have is: assuming that a new Mac Pro will be introduced within the next six months or so, how much of an improvement would it be over the current models?
These rumors about the next chips from Intel, integrated graphics, and so on are interesting, but how much of a difference will they mean for someone like me? In other words, if we're talking about cutting times in half, I can definitely wait- but if it's going to be a marginal improvement, I wonder if it might be better to get a 6-core and upgrade the graphics card in a couple of years?
I think I'm right about the kids, though. I see it all around me, especially girls. They use a netbook/laptop and a smartphone. And they don't like clunky boxes. For them the desktop is utterly irrelevant. There will always be a market for workstations/desktops but I do believe SJ is right in that the vast majority no longer needs them.
Now, in the absence of the xMac, I say that dropping the price of an entry-level MacPro to $1999 would constitute an acceptable compromise.
I think you're right there too. The mac pro used to start at $1999 with mid level hardware choices rather than the cheapest in the xeon line. I've gone over that before though. If they did something cool with it, sales potential still exists, just not in the volume of the macbook air and iphones. I almost think a lot of it has to do with brand image, like they don't want the brand to feel too corporate/commercial (note "feel"). They want it to maintain a coolness factor, and really no other brand has ever pulled that off with computers. If I had a nice, reasonably cost effective solution for an up to date workstation that runs OSX (and not a hackintosh) I would take it. Consider that mac pros are used in video capture and still photography quite a lot for their ability to shuffle through and process data. They're even taken on location. If Apple made them a bit lighter without killing the ability for easy expansion entirely, they might see a nice boost in sales from the professional market. As I've said it could even be accomplished by building upward from the mini rather than reverse engineering the mac pro.
Quote:
Originally Posted by Joe Blue
These rumors about the next chips from Intel, integrated graphics, and so on are interesting, but how much of a difference will they mean for someone like me? In other words, if we're talking about cutting times in half, I can definitely wait- but if it's going to be a marginal improvement, I wonder if it might be better to get a 6-core and upgrade the graphics card in a couple of years?
While a lot of things are becoming more gpu dependent, much of what I am guessing you rely on is still very cpu dependent (photoshop, illustrator, maybe InDesign and Capture One, Acrobat, etc), and no I don't think the 6 core is a good investment at this point. If it keeps going the way it's going, I may just buy the next imac and plug my eizo cg243w (my point was I'm invested into expensive displays too) into one of the thunderbolt ports. There's nothing stopping you from using a professional grade display there. It should support at least one extra just fine. Regarding graphics, there's no telling what the backward compatibility on future cards from Apple will be. The gpu in the imac is just ok. It's kind of a high end mobile card. The 5870 which is still a $200 upgrade came out on the PC side just over 2 years ago. IT is no longer new and shiny. Really you wouldn't be taking a big loss on graphics going with the imac, and a couple years from now it would be a PCI-E 3.0 card that may or may not work with that 6 core machine.
Something to note here.... the 6 core is a $1200 upgrade to go from a $300 processor to a $600 processor. Really I'd suggest waiting for the next major bump and jumping on whatever gets the best boost at that time. Really i'm trying to tell you I think that build is a waste of $4000.
Now if you post either in the thread or via pm what kind of problems your current computer is having, I will see if I can offer any help. There are ways to diagnose spinning wheel issues and stuff like that. Even kernel panics can sometimes be something as minor as a stick of bad ram (and replacing one in that model was cheap last time I looked).
While a lot of things are becoming more gpu dependent, much of what I am guessing you rely on is still very cpu dependent (photoshop, illustrator, maybe InDesign and Capture One, Acrobat, etc), and no I don't think the 6 core is a good investment at this point.
You are correct- it's all photography and print design stuff- no longer considered 'high-end', but still requiring a bit more than consumer-level machines offer. I agree with your evaluation of the 6-core, but remember, I'm running dual 2.66 Ghz Xeons: while the 6-core seems awfully steep to me, it is also a very big upgrade. I'm just trying to figure out whether I should hang on for a few more months for something even better.
Quote:
Originally Posted by hmm
If it keeps going the way it's going, I may just buy the next imac and plug my eizo cg243w (my point was I'm invested into expensive displays too) into one of the thunderbolt ports. There's nothing stopping you from using a professional grade display there. It should support at least one extra just fine.
Yes, the iMac seems to be the more prudent choice, but I get ill when I consider that either my Eizo or my NEC would have to gather dust- having the option to install an extra card might end up being a deal-breaker.
Quote:
Originally Posted by hmm
Now if you post either in the thread or via pm what kind of problems your current computer is having, I will see if I can offer any help. There are ways to diagnose spinning wheel issues and stuff like that. Even kernel panics can sometimes be something as minor as a stick of bad ram (and replacing one in that model was cheap last time I looked).
Thanks for the offer, but nothing that serious is happening. It's just getting old, and things are wearing out, so I'm trying to figure out which would be better: buying replacement parts (which are becoming scarce), getting a new machine, or hanging on until whatever's next comes out.
You are correct- it's all photography and print design stuff- no longer considered 'high-end', but still requiring a bit more than consumer-level machines offer. I agree with your evaluation of the 6-core, but remember, I'm running dual 2.66 Ghz Xeons: while the 6-core seems awfully steep to me, it is also a very big upgrade. I'm just trying to figure out whether I should hang on for a few more months for something even better.
Yes, the iMac seems to be the more prudent choice, but I get ill when I consider that either my Eizo or my NEC would have to gather dust- having the option to install an extra card might end up being a deal-breaker.
Thanks for the offer, but nothing that serious is happening. It's just getting old, and things are wearing out, so I'm trying to figure out which would be better: buying replacement parts (which are becoming scarce), getting a new machine, or hanging on until whatever's next comes out.
Ahh what NEC are you using? The older ones shifted shifted pretty fast and their calibration software had so many bugs. That and colorcomp was mostly useless compared to Eizo's method of uniformity correction. That's why I switched. I'm not sure what your needs are like but I only end up needing a display of that quality as my primary display so having an imac screen as the secondary wouldn't bother me as much.
Really the mac pro has the same chipset it received in 2009 and it's hardly economical for a machine that will look extremely dated the second an updated model is available. You've held onto the same one for 5 years, so it tells me that you don't change computers very often. It makes it hard for me to suggest spending $4000+ before tax on a configured machine that already looks dated in many aspects (graphics card, logic board, no usb3, no thunderbolt, processor released over a year and a half ago) While I don't feel the mac pros take priority for Apple, there are new processors appropriate to such a machine being debuted in November. It doesn't mean we'll see anything right away, but an updated baseline model may exceed the performance of that 6 core machine at a lower price point. The machine is plenty fast. I just don't think it's a great investment right now if this is going to be the machine you use for the next 5 years or so.
I'm not sure what is wearing out specifically. If it's an issue of a dated hard drive, keep in mind that you could opt to update such a part and pass it on to the new machine later (the only exception being a 3.5" drive in an imac, I almost forgot Apple switched to a proprietary one there). Apple's upgrade prices on such items are often not very competitive.
Last thing just wondering what do you use for storage currently? Internal? eSATA? SAS?
As such don't risk your business on anything in this forum.
Quote:
Originally Posted by Joe Blue
Hi guys,
Good discussion, but to return to the main question, are any of you willing to venture some advice?
I'm currently running a Mac Pro 1,1 (2006) that is rapidly nearing the end of life, and I will need to replace it withing the next few months.
Try to hold out for the next round of bumps or new machines. It is best to buy Apple hardware relatively close to the release date.
Quote:
If I had to buy one today, my choice would be the 6-core Westmere with the Radeon 5870- I work in photography and graphic design, so more cores would be unused.
I realize every bodies work habits are different but don't assume that more cores won't be useful in the future. At this point I consider four cores to be the minimal for just about anybody, professional users should consider more depending upon their needs.
Quote:
To be frank, after looking at Bare Feats, the high-end iMac appears to rival the Pro in performance, and I'd be tempted, but I'm not going to abandon my investment in displays.
You don't automatically loose that investment. You will be taking a bit of a step backward in GPU performance potential which "could" be a problem.
Quote:
So the question I have is: assuming that a new Mac Pro will be introduced within the next six months or so, how much of an improvement would it be over the current models?
That is a really good question and is pretty much what this discussion is about. Well in a sense. The real killer here is the length of time Intel has taken to deliver Xeons based upon Sandy Bridge. Given that they do a proper job the performance increase could be significant. Why? Well if you remove the GPU you then have a different thermal configuration, this could allow Intel to bump base frequency significantly over the iMac. Supposedly suitable chips have been announced, (I'm not following Intel lately) in any event something to consider.
Quote:
These rumors about the next chips from Intel, integrated graphics, and so on are interesting, but how much of a difference will they mean for someone like me? In other words, if we're talking about cutting times in half, I can definitely wait- but if it's going to be a marginal improvement, I wonder if it might be better to get a 6-core and upgrade the graphics card in a couple of years?
Well that is the question. I suspect a significant improvement over the current machine. Not that I've been wrong here before.
Ahh what NEC are you using? The older ones shifted shifted pretty fast and their calibration software had so many bugs. That and colorcomp was mostly useless compared to Eizo's method of uniformity correction. That's why I switched. I'm not sure what your needs are like but I only end up needing a display of that quality as my primary display so having an imac screen as the secondary wouldn't bother me as much.
I've been using the NEC PA30 as my layout display (or, I did until my Radeon XT1900 died) and the Eizo (same as yours) for detail work. Neither have given me a spot of trouble, I'd swear by both of them. Unfortunately, he Eizo has to make do with the old GForce 7300 that came standard on that model. As you can imagine, the gpu would be the first priority regardless of what I bought.
I've heard that the 5770 can be used in this machine with some workaround- and it's only $250 or so- that might be another option to tide me over.
Quote:
Originally Posted by hmm
While I don't feel the mac pros take priority for Apple, there are new processors appropriate to such a machine being debuted in November. It doesn't mean we'll see anything right away, but an updated baseline model may exceed the performance of that 6 core machine at a lower price point.
Yes, this is what I'm trying to determine- I'd be disappointed if they simply added more cores, since that seemed to be the direction for a while, and wouldn't help me much at all. I haven't heard much about the new processors, other than they're due in November, and really I just want to know what sort of boost they will give over the current model. If it's substantial, and will show up in the next three or four months, I can make do until then; if it's not, I'll need to try something else (like the iMac), and budget for something else in a year or two.
Quote:
Originally Posted by hmm
The machine is plenty fast. I just don't think it's a great investment right now if this is going to be the machine you use for the next 5 years or so.
That's what's holding me back- besides the possibilities of new processors, I'd like to see how other technologies like Thunderbolt pan out.
Quote:
Originally Posted by hmm
Last thing just wondering what do you use for storage currently? Internal? eSATA? SAS?
I've recently replaced most of my HDs with an eye towards a new machine. Internally, I have a 250 Gb WD for my system, and 1Tb raid-0 stripe (two 500 GB Hitachis) for data (all SATA). Those are backed up by external firewire drives. Whenever I do get a new machine, I plan on swapping the system to an SSD.
I realize every bodies work habits are different but don't assume that more cores won't be useful in the future. At this point I consider four cores to be the minimal for just about anybody, professional users should consider more depending upon their needs.
My personal benchmark is Photoshop, which according to what I've read, actually performs better on 6-8 cores than on 12. Apparently it is Adobe's problem to fix- and I haven't heard anything to indicate that it will be happening. So, if there's a big difference in price, I'll settle for the midrange.
Quote:
Originally Posted by wizard69
The real killer here is the length of time Intel has taken to deliver Xeons based upon Sandy Bridge. Given that they do a proper job the performance increase could be significant.? Supposedly suitable chips have been announced, (I'm not following Intel lately) in any event something to consider.
Exactly: when is the big question, immediately followed by how suitable? Personally, I want to believe the talk that it will be sooner rather than later, but know better than to put any stock in that.
It means the base model has 4 physical cores and 8 virtual. You're looking at a significantly higher clock speed and a newer architecture, so it should be a nice boost to performance.
Quote:
Originally Posted by Joe Blue
I've been using the NEC PA30 as my layout display (or, I did until my Radeon XT1900 died) and the Eizo (same as yours) for detail work. Neither have given me a spot of trouble, I'd swear by both of them. Unfortunately, he Eizo has to make do with the old GForce 7300 that came standard on that model. As you can imagine, the gpu would be the first priority regardless of what I bought.
That sucks that your card died . Yeah you know I've heard that NEC improved considerably with the PA series. I might give them another try one day, but I like my Eizo. My older NEC is set up next to it for non critical tasks and as a secondary display. I don't map the tablet to it though. Freehand drawing requires something close to a 1:1 mapping ratio and I do a lot of detail work.
Quote:
Originally Posted by Joe Blue
I've heard that the 5770 can be used in this machine with some workaround- and it's only $250 or so- that might be another option to tide me over.
That's not a terrible option. It's still a bit expensive for an older computer, but it's not terrible.
Quote:
Originally Posted by Joe Blue
Yes, this is what I'm trying to determine- I'd be disappointed if they simply added more cores, since that seemed to be the direction for a while, and wouldn't help me much at all. I haven't heard much about the new processors, other than they're due in November, and really I just want to know what sort of boost they will give over the current model. If it's substantial, and will show up in the next three or four months, I can make do until then; if it's not, I'll need to try something else (like the iMac), and budget for something else in a year or two.
That's what's holding me back- besides the possibilities of new processors, I'd like to see how other technologies like Thunderbolt pan out.
How thunderbolt will be implemented here is undetermined. It's possible they won't all be routed through integrated graphics. This seems like more of a workaround for mobile devices with constrained space, and Intel has announced plans for thunderbolt on the PC side as well and they have confirmed that the way we see it on the current Apple machines isn't the only possible style of implementation. Thunderbolt is known to support PCIe protocols but I don't know the exact details of it will be included, only that they've stated it will be included in chipsets coming in 2012. Apple may wait for this and usb3 to be available at the chipset level before releasing an updated mac pro.
Regarding processors they do seem to have retained a quad core configuration. I'd have to find the SKU's but you must remember the current 6 core machine uses an older architecture. It's still on Westmere (die shrink of Nehalem) rather than the Sandy Bridge generation. Overall there are a lot of newer IO standards right on the horizon so it is somewhat of an awkward time to buy. PCIe 3.0 is not yet implemented, and we're waiting on thunderbolt (on the mac pros) and usb3. The potential for a huge step up from the current lineup is there. It's just a matter of what Apple does in the end.
Quote:
Originally Posted by Joe Blue
I've recently replaced most of my HDs with an eye towards a new machine. Internally, I have a 250 Gb WD for my system, and 1Tb raid-0 stripe (two 500 GB Hitachis) for data (all SATA). Those are backed up by external firewire drives. Whenever I do get a new machine, I plan on swapping the system to an SSD.
If hard drive speed and SSDs are a concern of yours, the mac pro does not seem to have 6Gb/s SATA implemented yet. This feeds into what I was saying before. An updated machine "could" make the older ones feel very dated very fast just because of where we're at in terms of newer IO standards. USB2 has been the norm for quite a few years. SATA2 has lasted for a few years. PCIe 2.0 has been in mac pros since 2008? There are a lot of things they could implement, and this machine is still largely comprised of a 2009 layout internally. That's why I'm saying it's kind of a bad time to buy if you're on a lengthy repurchasing cycle.
If hard drive speed and SSDs are a concern of yours, the mac pro does not seem to have 6Gb/s SATA implemented yet. This feeds into what I was saying before. An updated machine "could" make the older ones feel very dated very fast just because of where we're at in terms of newer IO standards. USB2 has been the norm for quite a few years. SATA2 has lasted for a few years. PCIe 2.0 has been in mac pros since 2008? There are a lot of things they could implement, and this machine is still largely comprised of a 2009 layout internally. That's why I'm saying it's kind of a bad time to buy if you're on a lengthy repurchasing cycle.
I was waiting for the new MacPro but it didn't come and I had to pull the trigger. That's why I got a pre-owned '09 quad. Basically the same and with the firmware tweak you can drop in a hex and 1333 memory. Saved myself $1000 that way. I actually consider it to be my own personal xMac!
I thought about getting a new one but I felt like I was gonna pay $1000 more for a 140 Mhz cpu speedbump, a nicer gpu (that I don't need) and 360GB of extra hdd space. Combined with the prospect of a replacement coming this year, it just felt like a raw deal.
Now I'm happy with my '09 and will hold out for IvyBridge. If Apple hasn't EOL'd the MacPro by then.
It means the base model has 4 physical cores and 8 virtual. You're looking at a significantly higher clock speed and a newer architecture, so it should be a nice boost to performance.
Sorry, these numbers are for which chips- the current Mac Pros, or the upcoming release?
Quote:
Originally Posted by hmm
That sucks that your card died .
To be honest, I wasn't surprised. In my experience, the Radeon XT1900 was a dog from from the start- it ran hot and sounded like a hair dryer, and while nominally an upgrade (the only one available at the time), it wasn't much of an improvement at all. Frankly, it's put me off ATI- if they're the only option available, I'll take it, but I'd rather try something else.
Quote:
Originally Posted by hmm
I don't map the tablet to it though. Freehand drawing requires something close to a 1:1 mapping ratio and I do a lot of detail work.
I miss Freehand! What tablet do you use? I don't do much vector work, but I think I might get a tablet to take advantage of the gesture support- anything to take the strain from the wrist.
Quote:
Originally Posted by hmm
PCIe 3.0 is not yet implemented, and we're waiting on thunderbolt (on the mac pros and usb3. The potential for a huge step up from the current lineup is there. It's just a matter of what Apple does in the end.
Yes, this is what I'm curious about- while I'm all set regarding storage for the next few years (I hedged my bets and made sure that all my external drives are both Firewire and USB 3), I wonder which of these will take off, and which will slowly fade away... Faster storage would be nice- it's not bad now, but 6Gb/s SATA would considerably speed up Photoshop.
Comments
There is some evidence that Apples approach doesn't work all that well either. Rumors of marginal Mini sales continue to surface and iMac sales seem to be the result of people not having a choice.
Yeah... that's the only reason I've considered an imac at any point. All in ones have a ton of issues. They really aren't built for heavy heavy use to the degree of a tower with superior airflow. The display quality is just ok. I'm guessing if I had to, I could get used to it but it's sort of a higher end consumer grade display. It's not really meant for those concerned with color and detail. Regarding the mac pro people who buy them often use them for a very long time. It's not so easy to do this when the hardware sits so long. The logic board is the same one used in 2009. The processor in the baseline model is a slight bump from that because intel discontinued the other and replaced the w3520 with the w3530. If I bought one of these today, I would be concerned about how long it will retain current support. Apple has been really inconsistent in this regard, so I don't think anything past the next OS is a real guarantee.
Think about it a bit, most of Apples laptops are a better deal than the Mini.
In my limited testing many of the laptops offer better performance too. The mini is still invested into laptop parts so the cost to performance ratio is poor relative to desktop equivalents. The mini server brings you up to $1000 starting without keyboard or mouse (and back to integrated graphics). At that point the 21.5" imac at $1200 would be a better purchase overall. You lose the virtual core function of the i7 design but you're moving up to desktop processors and away from integrated graphics that share system memory.
And no external storage boxes, especially over Thunderbolt, are not a solution.
You know I don't disagree with you here, but I'll end up with external storage regardless of how many bays they put in it. I simply said before that for my purposes if it at least had multiple thunderbolt ports of a later generation/higher bandwidth, I'd probably grudgingly go that route rather than continue to wait and see.
Yeah... that's the only reason I've considered an imac at any point. All in ones have a ton of issues. They really aren't built for heavy heavy use to the degree of a tower with superior airflow. The display quality is just ok. I'm guessing if I had to, I could get used to it but it's sort of a higher end consumer grade display. It's not really meant for those concerned with color and detail. Regarding the mac pro people who buy them often use them for a very long time. It's not so easy to do this when the hardware sits so long. The logic board is the same one used in 2009. The processor in the baseline model is a slight bump from that because intel discontinued the other and replaced the w3520 with the w3530. If I bought one of these today, I would be concerned about how long it will retain current support. Apple has been really inconsistent in this regard, so I don't think anything past the next OS is a real guarantee.
Sadly I agree. For my needs an iMac wouldn't be to bad if one could easily deal with the storage limitations. If you go the Thunderbolt route though you end up pouring out a pretty good chunk of change for that solution. If I had a choice I'd go with a PCI Express card supporting 10G Ethernet or Fiber Channel, but of course Apple does not have an economical solution for that.
In my limited testing many of the laptops offer better performance too. The mini is still invested into laptop parts so the cost to performance ratio is poor relative to desktop equivalents. The mini server brings you up to $1000 starting without keyboard or mouse (and back to integrated graphics). At that point the 21.5" imac at $1200 would be a better purchase overall. You lose the virtual core function of the i7 design but you're moving up to desktop processors and away from integrated graphics that share system memory.
The problem with the Mini is that it is purposely limited. Even with Laptop chips they could go for more performance, instead they cheap out and provide midrange laptop performance. The cynic in me says they do that to entice people to buy laptops.
I'm not knocking the Mini as much as saying it is less than optimal for somebodies primary computer.
You know I don't disagree with you here, but I'll end up with external storage regardless of how many bays they put in it. I simply said before that for my purposes if it at least had multiple thunderbolt ports of a later generation/higher bandwidth, I'd probably grudgingly go that route rather than continue to wait and see.
Well some do have that problem. Fortunately I do not, at least not at the moment need that much external storage. These day though adding 4TB of storage to a desktop should be easy.
Sadly I agree. For my needs an iMac wouldn't be to bad if one could easily deal with the storage limitations. If you go the Thunderbolt route though you end up pouring out a pretty good chunk of change for that solution. If I had a choice I'd go with a PCI Express card supporting 10G Ethernet or Fiber Channel, but of course Apple does not have an economical solution for that.
Is 10G or fiber channel really that cost effective in any form? Feel free to correct me here but I've never personally seen cheap cards for those standards. I imagine by next year we'll start to see more cost effective thunderbolt solutions and possibly the ability to retrofit backplanes to existing esata solutions on the dumb box end. Intel was very very very slow in releasing an SDK or any other materials to aid development of thunderbolt accessories, and PCs aren't really seeing those ports until next year. Give it a little time and it may become very usable. It's just that I'd like to see a minimum of two ports on a given machine. Daisy chaining is a really bad option for high bandwidth items like drive enclosures populated by SSDs and displays because either could consume a very large portion of its bandwidth in either direction. Displayport 1.2 actually supports 17.28 Gb/s of video bandwidth with thunderbolt supporting 10 in each direction. My point being that we really need more than a single port on any given machine if they are to be used for both external displays and storage.
The problem with the Mini is that it is purposely limited. Even with Laptop chips they could go for more performance, instead they cheap out and provide midrange laptop performance. The cynic in me says they do that to entice people to buy laptops.
I'm not knocking the Mini as much as saying it is less than optimal for somebodies primary computer.
I agree with you there, and they do really go pretty tight on it especially with the lack of a keyboard and mouse. Should you desire the Apple versions, that's around $100 extra for the two items assuming you don't go wireless. I sometimes feel like they reverse engineer these things to hit a certain price point while retaining margins.
Well some do have that problem. Fortunately I do not, at least not at the moment need that much external storage. These day though adding 4TB of storage to a desktop should be easy.
It is. 4TB is nothing though
Is 10G or fiber channel really that cost effective in any form? Feel free to correct me here but I've never personally seen cheap cards for those standards.
Cheap - no but they are well accepted standards. That means a storage server using the standards would be viable for some time. Interestingly I just saw a bit on a new PCI Express 3.0 card that implements several of these protocols in one card.
I imagine by next year we'll start to see more cost effective thunderbolt solutions and possibly the ability to retrofit backplanes to existing esata solutions on the dumb box end. Intel was very very very slow in releasing an SDK or any other materials to aid development of thunderbolt accessories, and PCs aren't really seeing those ports until next year.
Slow! Turtles move faster than Intel here.
By the way unless Intel releases the standard publicly I not to sure how well it will go over in the long run. The using community has been burnt to many times in the past by proprietary standards, so unless Intel and Apple open this up I can't see mass acceptance.
Give it a little time and it may become very usable. It's just that I'd like to see a minimum of two ports on a given machine. Daisy chaining is a really bad option for high bandwidth items like drive enclosures populated by SSDs and displays because either could consume a very large portion of its bandwidth in either direction. Displayport 1.2 actually supports 17.28 Gb/s of video bandwidth with thunderbolt supporting 10 in each direction. My point being that we really need more than a single port on any given machine if they are to be used for both external displays and storage.
Very true about port congestion! People still don't grasp the concept that this is a serial link. One good point though is that we have lines going in each direction. In the end though without additional ports I can see many a professional user getting a little perturbed with port.
I agree with you there, and they do really go pretty tight on it especially with the lack of a keyboard and mouse. Should you desire the Apple versions, that's around $100 extra for the two items assuming you don't go wireless. I sometimes feel like they reverse engineer these things to hit a certain price point while retaining margins.
I don't think it is price point as much as it is resisting the temptation to exceed the performance levels of their laptops at anyone time. It is very much a conscious effort on their part. Let's face it, the Mini is a desktop machine there is no battery nor cooling limitation to speak of. Further the unit including the discreet GPU is just pathetic considering the extra cost and marginal GPU implementation.
It is. 4TB is nothing though
Yeah the more that I think about it the more that I would agree. After all today that is just four low profile laptop drives. For me though it would be nice to have that in a base machine. This brings up another Mini issue, what is with the puny disk offerings? I can understand the entry level models but what about the options, it isn't like 1TB disks are unheard of in California or even 750GB models.
Oh about backups those should very much be on an external system. Internally though I want room for all my data and code. That isn't asking a lot and frankly easy these days for me to cover with laptop drives. That is if I had an XMac with a few drive bays.
Cheap - no but they are well accepted standards. That means a storage server using the standards would be viable for some time. Interestingly I just saw a bit on a new PCI Express 3.0 card that implements several of these protocols in one card.
Slow! Turtles move faster than Intel here.
By the way unless Intel releases the standard publicly I not to sure how well it will go over in the long run. The using community has been burnt to many times in the past by proprietary standards, so unless Intel and Apple open this up I can't see mass acceptance.
I thought they were going to do so at some point. That could have been a misinterpretation or wishful thinking on my part. Anyway yeah they're way too slow about it. They need people to feel like they're really behind it and provide a good environment for third party development. Just placing it on chipsets doesn't replace the need for a quality SDK.
Very true about port congestion! People still don't grasp the concept that this is a serial link. One good point though is that we have lines going in each direction. In the end though without additional ports I can see many a professional user getting a little perturbed with port.
Yep....and they were very misleading about this. Apple and intel have pressed the idea that you can just hook up whatever you want to via a single thunderbolt port, and this really isn't the case.
I don't think it is price point as much as it is resisting the temptation to exceed the performance levels of their laptops at anyone time. It is very much a conscious effort on their part. Let's face it, the Mini is a desktop machine there is no battery nor cooling limitation to speak of. Further the unit including the discreet GPU is just pathetic considering the extra cost and marginal GPU implementation.
You might be right. I thought the minis got quite hot though? Perhaps this has changed more than I realized. I just don't consider the quad/mini server option a truly viable alternative to other things in the surrounding price points as anything other than a light duty server. Quad core machines have basically become the norm at this point so the dual core version isn't a true consideration for me.
Yeah the more that I think about it the more that I would agree. After all today that is just four low profile laptop drives. For me though it would be nice to have that in a base machine. This brings up another Mini issue, what is with the puny disk offerings? I can understand the entry level models but what about the options, it isn't like 1TB disks are unheard of in California or even 750GB models.
Oh about backups those should very much be on an external system. Internally though I want room for all my data and code. That isn't asking a lot and frankly easy these days for me to cover with laptop drives. That is if I had an XMac with a few drive bays.
If SSDs continue to drop in price per GB, I could see a compact machine with 2-4 2.5" bays to cut heat in a somewhat confined enclosure. They've offered this option on the mac pro, but it is still an extremely expensive route. It may take a few years but we'll see a further drop in the popularity of 2.5" HDDs. I agree regarding backups. I also believe it's important to maintain actual offline backups of critical data. eSATA type enclosures aren't that bad a solution. You just have to be very careful what you buy. You want one with powerful cooling, which often means having to blow out dust every few months. If you're on a budget you want to completely avoid raid solutions. Cheap raids suck, and many people don't understand how they work at all. I've spent a fair amount of time explaining to others why they have to back up their raid solutions and that parity striping will not save them in the event of data corruption, controller failure, or a power surge of any kind (including the power supply on the unit failing).
I thought they were going to do so at some point. That could have been a misinterpretation or wishful thinking on my part. Anyway yeah they're way too slow about it. They need people to feel like they're really behind it and provide a good environment for third party development. Just placing it on chipsets doesn't replace the need for a quality SDK.
They need to do something here to make TB a public standard. Right now it is too obscure to be considered a major player. As far as Intel goes you almost get the impression that the release was too early.
The problem with widespread implementation has already caused the USB people to consider developing an alternative. To this end I don't see TB being acceptable to the industry as a whole until you can get independent chipsets to implement the interface. More importantly the port needs to be implemented on micro controllers if reasonable prices are to be had. Further embedded people won't be willing to buy Intel hardware.
Yep....and they were very misleading about this. Apple and intel have pressed the idea that you can just hook up whatever you want to via a single thunderbolt port, and this really isn't the case.
Sadly they will learn. However many users will do fine with the TB port as it's performance will depend upon the work load.
You might be right. I thought the minis got quite hot though? Perhaps this has changed more than I realized.
I don't have a new one to test, however the last case redesign was a major one. It appears that one goal was much better cooling. That being said if cooling was an issue I don't think we would be seeing racks of these used as servers.
Now a server might not be as heavily loaded as you might in your usage. Given the right work load you can heat up just about any computer.
I just don't consider the quad/mini server option a truly viable alternative to other things in the surrounding price points as anything other than a light duty server. Quad core machines have basically become the norm at this point so the dual core version isn't a true consideration for me.
I can understand the disappointment about the dual cores but that just goes to support my point, Apple castrated the machines to make their laptops look good. It is positively storage that they put a quad core in one model and then called it a server.
If SSDs continue to drop in price per GB, I could see a compact machine with 2-4 2.5" bays to cut heat in a somewhat confined enclosure. They've offered this option on the mac pro, but it is still an extremely expensive route.
Apple needs to make use of one of the emerging PCI - Express standards for solid state storage cards. The goal should be to use commodity parts that leverage the fast interface of PCI-E. The use of cards should lead to smaller hardware and easier cooling.
It may take a few years but we'll see a further drop in the popularity of 2.5" HDDs. I agree regarding backups. I also believe it's important to maintain actual offline backups of critical data. eSATA type enclosures aren't that bad a solution. You just have to be very careful what you buy. You want one with powerful cooling, which often means having to blow out dust every few months.
A good and cheap solution. If nothing else data should be backed up here.
If you're on a budget you want to completely avoid raid solutions. Cheap raids suck, and many people don't understand how they work at all.
In the past I implemented RAIDs on some of my Linux machines. Software based Linux RAIDs can be a good low cost option but you are right you need to understand them. In my case implementing them was easy. Luck was on my side as I never had to rebuild a volume. In any event I didn't do the RAIDs back then for data security.
For a long while there I outgrew my hardware before failure.
I've spent a fair amount of time explaining to others why they have to back up their raid solutions and that parity striping will not save them in the event of data corruption, controller failure, or a power surge of any kind (including the power supply on the unit failing).
Yep even data center RAIDs with all of their fancy protections go down. Sometimes for a very long time. Even a minor failure takes forever to rebuild.
Even backups of RAIDs can get corrupted though so back ups of critical data really requires multiple approaches that don't depend upon other systems. Sometimes a CD, disk or tape in a safe deposit box makes a lot of sense. However even that is a pain with a reasonably large RAID.
Back ups are very important but I'm not sure if people get it. In some of the forums I visit new college students will ask about what computer they should buy. For the most part that is easy, buy a Mac unless your program has specific requirements. I always stress the need to consider a backup program so that they don't suffer due to some sort of failure. What interests me is how many listen to the advice.
They need to do something here to make TB a public standard. Right now it is too obscure to be considered a major player. As far as Intel goes you almost get the impression that the release was too early.
The problem with widespread implementation has already caused the USB people to consider developing an alternative. To this end I don't see TB being acceptable to the industry as a whole until you can get independent chipsets to implement the interface. More importantly the port needs to be implemented on micro controllers if reasonable prices are to be had. Further embedded people won't be willing to buy Intel hardware.
If they did in fact make it easier/cheaper to implement, it could work out well. It remains significantly cheaper than some of the other high data bandwidth solutions. It's just that right now I think a single port which is also tied to display output is an issue. Display resolution will continue to increase significantly. Medical grade panels have achieved some pretty insane numbers in the past few years. Laptop resolution has shot way up. Right now the desktop panels are a bit stuck, but this will change. Look at how many people like the crisp look of the retina display and the number of magazines migrating to digital content delivery.
I can understand the disappointment about the dual cores but that just goes to support my point, Apple castrated the machines to make their laptops look good. It is positively storage that they put a quad core in one model and then called it a server.
They put a quad core processor and weaker graphics in it which was also the only i7 variant in the line. It makes sense as a low power server really. The i7 variants have hyper threading enabled which would help such a machine process more simultaneous requests. If they built out/up a bit from the mini design, they could have a decent desktop machine. Marvin referenced something similar but I didn't like the example he picked (I felt the example of a machine that uses a very low wattage power source might not work out well if too many things are plugged into the machine).
Apple needs to make use of one of the emerging PCI - Express standards for solid state storage cards. The goal should be to use commodity parts that leverage the fast interface of PCI-E. The use of cards should lead to smaller hardware and easier cooling.
That's a great idea. PCI-E is still a solid interface. I don't think everyone realizes that it doesn't have to be implemented entirely at the back of the computer.
Back ups are very important but I'm not sure if people get it. In some of the forums I visit new college students will ask about what computer they should buy. For the most part that is easy, buy a Mac unless your program has specific requirements. I always stress the need to consider a backup program so that they don't suffer due to some sort of failure. What interests me is how many listen to the advice.
I agree with you here. Some things require me to boot into Windows and emulation doesn't deliver the required performance. When in Windows I tend to stay off the internet, and turn off anything that annoys me. This includes a bunch of cursor animations (if you use a graphics tablet the ripple effect every time you click on something is just way too annoying if you work as fast as I do) and Aero. Run like this it runs fine.
I don't feel like quoting your lengthy posts and go over every single point, but to recap:
You say Apple is missing out on a huge opportunity by not offering the xMac.
Fact: all computer sales are down and declining. Only Apple's sales grow. You seem to think that their sales grow in spite of their "incomplete" or restricted line-up. I reckon it is more logical to assume that their sales improve precisely because they don't offer the same muck that everybody else has been churning out forever.
And PC's aren't an investment anymore, not for the middle-class anyway. They have become a commodity. People have them repaired about as frequent as they have their shoes resoled.
If you want to know what is going to happen to the desktop, you only need to look at how 20-year olds use them. If it is not for gaming or some other specific activity, they don't have one and they don't want one. They have a laptop or netbook but they spent most of their time on their smartphones.
@Wizard69:
Bleh no love
Fact: all computer sales are down and declining. Only Apple's sales grow. You seem to think that their sales grow in spite of their "incomplete" or restricted line-up. I reckon it is more logical to assume that their sales improve precisely because they don't offer the same muck that everybody else has been churning out forever.
The desktop form factor is still better for a lot of professionals. The guys that own machines like a mac pro, also own macbook pros and typically iphones. Having a solid option for users that need power and bandwidth does benefit Apple. Right now they haven't been paying much attention to the mac pro. I'd like to see them build a followup that they can be enthusiastic about today, because there will still be a market for it. They need something for those that work with multiple 1900x1200 professional grade displays (displays that cost more than an imac) and a multitude of hard drives. Apple has been all about ecosystems and this would fit in as an integral part if they just decide how to proceed. Otherwise it will get to the point where they try to kill off the mac pro and end up with another pr fiasco. Remember how they brought back firewire on one of their laptops some time ago and the recent FCP X thing? Having stuff like this helps those who need it maintain an all Apple setup.
Sadly I've been having to use bootcamp lately. I customized the living hell out of windows so that it wouldn't irritate me, and I still hate it.
If you want to know what is going to happen to the desktop, you only need to look at how 20-year olds use them. If it is not for gaming or some other specific activity, they don't have one and they don't want one. They have a laptop or netbook but they spent most of their time on their smartphones.
You're thinking too old. You should look at how teenagers use their phones and computers. Apple has always been very good at being a cool brand given the early associations with its popularity amongst graphic designers, photographers, and other creative professions, speaking of which many of those guys still own mac pros just because it allows for a stable computer with a large amount of ram. If they moved away from this design and built upward from the mac mini instead, I would totally buy it. If we could get one that used desktop processors because the high end laptop variant would be too expensive, at an average gpu, better ram capacity, and 2-3 second generation TB ports, we might have an extremely popular machine for all of these guys that want xmacs. You could keep it small enough to hold its appeal to the mini crowd but give it enough power to act like a true desktop. I'm not talking only about cpu, but about the machine as a whole.
Bleh no love
Sorry if you felt left out, man. I use a MacPro but only for audio. I am amazed at how the iPhone has usurped so many of what was previously desktop duties.
For work nothing but a MacPro will do (or a similarly spec'd Windows workstation) and that's only because there is no xMac. If there was one, I would be dead in the middle of its target market. A 2600-i7 with a few hdd bays and pci slots would be sufficient for me, but alas.
Yet I understand Apple's strategy and I think they have proven to be right about this. Too bad for me, now I have to pay $1000 more for a MacPro or go with Windows. But for most people, an iMac suffices, even as the hub of their digital life. Outside of work, my iMac does all I need it to do. Shucks, a MacMini would probably suffice (albeit with an external superdrive).
I think I'm right about the kids, though. I see it all around me, especially girls. They use a netbook/laptop and a smartphone. And they don't like clunky boxes. For them the desktop is utterly irrelevant. There will always be a market for workstations/desktops but I do believe SJ is right in that the vast majority no longer needs them.
Now, in the absence of the xMac, I say that dropping the price of an entry-level MacPro to $1999 would constitute an acceptable compromise.
Good discussion, but to return to the main question, are any of you willing to venture some advice?
I'm currently running a Mac Pro 1,1 (2006) that is rapidly nearing the end of life, and I will need to replace it withing the next few months.
If I had to buy one today, my choice would be the 6-core Westmere with the Radeon 5870- I work in photography and graphic design, so more cores would be unused.
To be frank, after looking at Bare Feats, the high-end iMac appears to rival the Pro in performance, and I'd be tempted, but I'm not going to abandon my investment in displays.
So the question I have is: assuming that a new Mac Pro will be introduced within the next six months or so, how much of an improvement would it be over the current models?
These rumors about the next chips from Intel, integrated graphics, and so on are interesting, but how much of a difference will they mean for someone like me? In other words, if we're talking about cutting times in half, I can definitely wait- but if it's going to be a marginal improvement, I wonder if it might be better to get a 6-core and upgrade the graphics card in a couple of years?
I think I'm right about the kids, though. I see it all around me, especially girls. They use a netbook/laptop and a smartphone. And they don't like clunky boxes. For them the desktop is utterly irrelevant. There will always be a market for workstations/desktops but I do believe SJ is right in that the vast majority no longer needs them.
Now, in the absence of the xMac, I say that dropping the price of an entry-level MacPro to $1999 would constitute an acceptable compromise.
I think you're right there too. The mac pro used to start at $1999 with mid level hardware choices rather than the cheapest in the xeon line. I've gone over that before though. If they did something cool with it, sales potential still exists, just not in the volume of the macbook air and iphones. I almost think a lot of it has to do with brand image, like they don't want the brand to feel too corporate/commercial (note "feel"). They want it to maintain a coolness factor, and really no other brand has ever pulled that off with computers. If I had a nice, reasonably cost effective solution for an up to date workstation that runs OSX (and not a hackintosh) I would take it. Consider that mac pros are used in video capture and still photography quite a lot for their ability to shuffle through and process data. They're even taken on location. If Apple made them a bit lighter without killing the ability for easy expansion entirely, they might see a nice boost in sales from the professional market. As I've said it could even be accomplished by building upward from the mini rather than reverse engineering the mac pro.
These rumors about the next chips from Intel, integrated graphics, and so on are interesting, but how much of a difference will they mean for someone like me? In other words, if we're talking about cutting times in half, I can definitely wait- but if it's going to be a marginal improvement, I wonder if it might be better to get a 6-core and upgrade the graphics card in a couple of years?
While a lot of things are becoming more gpu dependent, much of what I am guessing you rely on is still very cpu dependent (photoshop, illustrator, maybe InDesign and Capture One, Acrobat, etc), and no I don't think the 6 core is a good investment at this point. If it keeps going the way it's going, I may just buy the next imac and plug my eizo cg243w (my point was I'm invested into expensive displays too) into one of the thunderbolt ports. There's nothing stopping you from using a professional grade display there. It should support at least one extra just fine. Regarding graphics, there's no telling what the backward compatibility on future cards from Apple will be. The gpu in the imac is just ok. It's kind of a high end mobile card. The 5870 which is still a $200 upgrade came out on the PC side just over 2 years ago. IT is no longer new and shiny. Really you wouldn't be taking a big loss on graphics going with the imac, and a couple years from now it would be a PCI-E 3.0 card that may or may not work with that 6 core machine.
Something to note here.... the 6 core is a $1200 upgrade to go from a $300 processor to a $600 processor. Really I'd suggest waiting for the next major bump and jumping on whatever gets the best boost at that time. Really i'm trying to tell you I think that build is a waste of $4000.
Now if you post either in the thread or via pm what kind of problems your current computer is having, I will see if I can offer any help. There are ways to diagnose spinning wheel issues and stuff like that. Even kernel panics can sometimes be something as minor as a stick of bad ram (and replacing one in that model was cheap last time I looked).
While a lot of things are becoming more gpu dependent, much of what I am guessing you rely on is still very cpu dependent (photoshop, illustrator, maybe InDesign and Capture One, Acrobat, etc), and no I don't think the 6 core is a good investment at this point.
You are correct- it's all photography and print design stuff- no longer considered 'high-end', but still requiring a bit more than consumer-level machines offer. I agree with your evaluation of the 6-core, but remember, I'm running dual 2.66 Ghz Xeons: while the 6-core seems awfully steep to me, it is also a very big upgrade. I'm just trying to figure out whether I should hang on for a few more months for something even better.
If it keeps going the way it's going, I may just buy the next imac and plug my eizo cg243w (my point was I'm invested into expensive displays too) into one of the thunderbolt ports. There's nothing stopping you from using a professional grade display there. It should support at least one extra just fine.
Yes, the iMac seems to be the more prudent choice, but I get ill when I consider that either my Eizo or my NEC would have to gather dust- having the option to install an extra card might end up being a deal-breaker.
Now if you post either in the thread or via pm what kind of problems your current computer is having, I will see if I can offer any help. There are ways to diagnose spinning wheel issues and stuff like that. Even kernel panics can sometimes be something as minor as a stick of bad ram (and replacing one in that model was cheap last time I looked).
Thanks for the offer, but nothing that serious is happening. It's just getting old, and things are wearing out, so I'm trying to figure out which would be better: buying replacement parts (which are becoming scarce), getting a new machine, or hanging on until whatever's next comes out.
You are correct- it's all photography and print design stuff- no longer considered 'high-end', but still requiring a bit more than consumer-level machines offer. I agree with your evaluation of the 6-core, but remember, I'm running dual 2.66 Ghz Xeons: while the 6-core seems awfully steep to me, it is also a very big upgrade. I'm just trying to figure out whether I should hang on for a few more months for something even better.
Yes, the iMac seems to be the more prudent choice, but I get ill when I consider that either my Eizo or my NEC would have to gather dust- having the option to install an extra card might end up being a deal-breaker.
Thanks for the offer, but nothing that serious is happening. It's just getting old, and things are wearing out, so I'm trying to figure out which would be better: buying replacement parts (which are becoming scarce), getting a new machine, or hanging on until whatever's next comes out.
Ahh what NEC are you using? The older ones shifted shifted pretty fast and their calibration software had so many bugs. That and colorcomp was mostly useless compared to Eizo's method of uniformity correction. That's why I switched. I'm not sure what your needs are like but I only end up needing a display of that quality as my primary display so having an imac screen as the secondary wouldn't bother me as much.
Really the mac pro has the same chipset it received in 2009 and it's hardly economical for a machine that will look extremely dated the second an updated model is available. You've held onto the same one for 5 years, so it tells me that you don't change computers very often. It makes it hard for me to suggest spending $4000+ before tax on a configured machine that already looks dated in many aspects (graphics card, logic board, no usb3, no thunderbolt, processor released over a year and a half ago) While I don't feel the mac pros take priority for Apple, there are new processors appropriate to such a machine being debuted in November. It doesn't mean we'll see anything right away, but an updated baseline model may exceed the performance of that 6 core machine at a lower price point. The machine is plenty fast. I just don't think it's a great investment right now if this is going to be the machine you use for the next 5 years or so.
I'm not sure what is wearing out specifically. If it's an issue of a dated hard drive, keep in mind that you could opt to update such a part and pass it on to the new machine later (the only exception being a 3.5" drive in an imac, I almost forgot Apple switched to a proprietary one there). Apple's upgrade prices on such items are often not very competitive.
Last thing just wondering what do you use for storage currently? Internal? eSATA? SAS?
Hi guys,
Good discussion, but to return to the main question, are any of you willing to venture some advice?
I'm currently running a Mac Pro 1,1 (2006) that is rapidly nearing the end of life, and I will need to replace it withing the next few months.
Try to hold out for the next round of bumps or new machines. It is best to buy Apple hardware relatively close to the release date.
If I had to buy one today, my choice would be the 6-core Westmere with the Radeon 5870- I work in photography and graphic design, so more cores would be unused.
I realize every bodies work habits are different but don't assume that more cores won't be useful in the future. At this point I consider four cores to be the minimal for just about anybody, professional users should consider more depending upon their needs.
To be frank, after looking at Bare Feats, the high-end iMac appears to rival the Pro in performance, and I'd be tempted, but I'm not going to abandon my investment in displays.
You don't automatically loose that investment. You will be taking a bit of a step backward in GPU performance potential which "could" be a problem.
So the question I have is: assuming that a new Mac Pro will be introduced within the next six months or so, how much of an improvement would it be over the current models?
That is a really good question and is pretty much what this discussion is about. Well in a sense. The real killer here is the length of time Intel has taken to deliver Xeons based upon Sandy Bridge. Given that they do a proper job the performance increase could be significant. Why? Well if you remove the GPU you then have a different thermal configuration, this could allow Intel to bump base frequency significantly over the iMac. Supposedly suitable chips have been announced, (I'm not following Intel lately) in any event something to consider.
These rumors about the next chips from Intel, integrated graphics, and so on are interesting, but how much of a difference will they mean for someone like me? In other words, if we're talking about cutting times in half, I can definitely wait- but if it's going to be a marginal improvement, I wonder if it might be better to get a 6-core and upgrade the graphics card in a couple of years?
Well that is the question. I suspect a significant improvement over the current machine. Not that I've been wrong here before.
Ahh what NEC are you using? The older ones shifted shifted pretty fast and their calibration software had so many bugs. That and colorcomp was mostly useless compared to Eizo's method of uniformity correction. That's why I switched. I'm not sure what your needs are like but I only end up needing a display of that quality as my primary display so having an imac screen as the secondary wouldn't bother me as much.
I've been using the NEC PA30 as my layout display (or, I did until my Radeon XT1900 died) and the Eizo (same as yours) for detail work. Neither have given me a spot of trouble, I'd swear by both of them. Unfortunately, he Eizo has to make do with the old GForce 7300 that came standard on that model. As you can imagine, the gpu would be the first priority regardless of what I bought.
I've heard that the 5770 can be used in this machine with some workaround- and it's only $250 or so- that might be another option to tide me over.
While I don't feel the mac pros take priority for Apple, there are new processors appropriate to such a machine being debuted in November. It doesn't mean we'll see anything right away, but an updated baseline model may exceed the performance of that 6 core machine at a lower price point.
Yes, this is what I'm trying to determine- I'd be disappointed if they simply added more cores, since that seemed to be the direction for a while, and wouldn't help me much at all. I haven't heard much about the new processors, other than they're due in November, and really I just want to know what sort of boost they will give over the current model. If it's substantial, and will show up in the next three or four months, I can make do until then; if it's not, I'll need to try something else (like the iMac), and budget for something else in a year or two.
The machine is plenty fast. I just don't think it's a great investment right now if this is going to be the machine you use for the next 5 years or so.
That's what's holding me back- besides the possibilities of new processors, I'd like to see how other technologies like Thunderbolt pan out.
Last thing just wondering what do you use for storage currently? Internal? eSATA? SAS?
I've recently replaced most of my HDs with an eye towards a new machine. Internally, I have a 250 Gb WD for my system, and 1Tb raid-0 stripe (two 500 GB Hitachis) for data (all SATA). Those are backed up by external firewire drives. Whenever I do get a new machine, I plan on swapping the system to an SSD.
I realize every bodies work habits are different but don't assume that more cores won't be useful in the future. At this point I consider four cores to be the minimal for just about anybody, professional users should consider more depending upon their needs.
My personal benchmark is Photoshop, which according to what I've read, actually performs better on 6-8 cores than on 12. Apparently it is Adobe's problem to fix- and I haven't heard anything to indicate that it will be happening. So, if there's a big difference in price, I'll settle for the midrange.
The real killer here is the length of time Intel has taken to deliver Xeons based upon Sandy Bridge. Given that they do a proper job the performance increase could be significant.? Supposedly suitable chips have been announced, (I'm not following Intel lately) in any event something to consider.
Exactly: when is the big question, immediately followed by how suitable? Personally, I want to believe the talk that it will be sooner rather than later, but know better than to put any stock in that.
Xeon E5-1620\t4\t8\t3.6 GHz\t10 MB\t130 Watt\t$294
Xeon E5-1650\t6\t12\t3.2 GHz\t12 MB\t130 Watt\t$583
Xeon E5-1660\t6\t12\t3.3 GHz\t15 MB\t130 Watt\t$1080
It means the base model has 4 physical cores and 8 virtual. You're looking at a significantly higher clock speed and a newer architecture, so it should be a nice boost to performance.
I've been using the NEC PA30 as my layout display (or, I did until my Radeon XT1900 died) and the Eizo (same as yours) for detail work. Neither have given me a spot of trouble, I'd swear by both of them. Unfortunately, he Eizo has to make do with the old GForce 7300 that came standard on that model. As you can imagine, the gpu would be the first priority regardless of what I bought.
That sucks that your card died
I've heard that the 5770 can be used in this machine with some workaround- and it's only $250 or so- that might be another option to tide me over.
That's not a terrible option. It's still a bit expensive for an older computer, but it's not terrible.
Yes, this is what I'm trying to determine- I'd be disappointed if they simply added more cores, since that seemed to be the direction for a while, and wouldn't help me much at all. I haven't heard much about the new processors, other than they're due in November, and really I just want to know what sort of boost they will give over the current model. If it's substantial, and will show up in the next three or four months, I can make do until then; if it's not, I'll need to try something else (like the iMac), and budget for something else in a year or two.
That's what's holding me back- besides the possibilities of new processors, I'd like to see how other technologies like Thunderbolt pan out.
How thunderbolt will be implemented here is undetermined. It's possible they won't all be routed through integrated graphics. This seems like more of a workaround for mobile devices with constrained space, and Intel has announced plans for thunderbolt on the PC side as well and they have confirmed that the way we see it on the current Apple machines isn't the only possible style of implementation. Thunderbolt is known to support PCIe protocols but I don't know the exact details of it will be included, only that they've stated it will be included in chipsets coming in 2012. Apple may wait for this and usb3 to be available at the chipset level before releasing an updated mac pro.
Regarding processors they do seem to have retained a quad core configuration. I'd have to find the SKU's but you must remember the current 6 core machine uses an older architecture. It's still on Westmere (die shrink of Nehalem) rather than the Sandy Bridge generation. Overall there are a lot of newer IO standards right on the horizon so it is somewhat of an awkward time to buy. PCIe 3.0 is not yet implemented, and we're waiting on thunderbolt (on the mac pros) and usb3. The potential for a huge step up from the current lineup is there. It's just a matter of what Apple does in the end.
I've recently replaced most of my HDs with an eye towards a new machine. Internally, I have a 250 Gb WD for my system, and 1Tb raid-0 stripe (two 500 GB Hitachis) for data (all SATA). Those are backed up by external firewire drives. Whenever I do get a new machine, I plan on swapping the system to an SSD.
If hard drive speed and SSDs are a concern of yours, the mac pro does not seem to have 6Gb/s SATA implemented yet. This feeds into what I was saying before. An updated machine "could" make the older ones feel very dated very fast just because of where we're at in terms of newer IO standards. USB2 has been the norm for quite a few years. SATA2 has lasted for a few years. PCIe 2.0 has been in mac pros since 2008? There are a lot of things they could implement, and this machine is still largely comprised of a 2009 layout internally. That's why I'm saying it's kind of a bad time to buy if you're on a lengthy repurchasing cycle.
If hard drive speed and SSDs are a concern of yours, the mac pro does not seem to have 6Gb/s SATA implemented yet. This feeds into what I was saying before. An updated machine "could" make the older ones feel very dated very fast just because of where we're at in terms of newer IO standards. USB2 has been the norm for quite a few years. SATA2 has lasted for a few years. PCIe 2.0 has been in mac pros since 2008? There are a lot of things they could implement, and this machine is still largely comprised of a 2009 layout internally. That's why I'm saying it's kind of a bad time to buy if you're on a lengthy repurchasing cycle.
I was waiting for the new MacPro but it didn't come and I had to pull the trigger. That's why I got a pre-owned '09 quad. Basically the same and with the firmware tweak you can drop in a hex and 1333 memory. Saved myself $1000 that way. I actually consider it to be my own personal xMac!
I thought about getting a new one but I felt like I was gonna pay $1000 more for a 140 Mhz cpu speedbump, a nicer gpu (that I don't need) and 360GB of extra hdd space. Combined with the prospect of a replacement coming this year, it just felt like a raw deal.
Now I'm happy with my '09 and will hold out for IvyBridge. If Apple hasn't EOL'd the MacPro by then.
Edit: This is what I found googling the new processors.
Xeon E5-1620\t4\t8\t3.6 GHz\t10 MB\t130 Watt\t$294
Xeon E5-1650\t6\t12\t3.2 GHz\t12 MB\t130 Watt\t$583
Xeon E5-1660\t6\t12\t3.3 GHz\t15 MB\t130 Watt\t$1080
It means the base model has 4 physical cores and 8 virtual. You're looking at a significantly higher clock speed and a newer architecture, so it should be a nice boost to performance.
Sorry, these numbers are for which chips- the current Mac Pros, or the upcoming release?
That sucks that your card died
To be honest, I wasn't surprised. In my experience, the Radeon XT1900 was a dog from from the start- it ran hot and sounded like a hair dryer, and while nominally an upgrade (the only one available at the time), it wasn't much of an improvement at all. Frankly, it's put me off ATI- if they're the only option available, I'll take it, but I'd rather try something else.
I don't map the tablet to it though. Freehand drawing requires something close to a 1:1 mapping ratio and I do a lot of detail work.
I miss Freehand! What tablet do you use? I don't do much vector work, but I think I might get a tablet to take advantage of the gesture support- anything to take the strain from the wrist.
PCIe 3.0 is not yet implemented, and we're waiting on thunderbolt (on the mac pros and usb3. The potential for a huge step up from the current lineup is there. It's just a matter of what Apple does in the end.
Yes, this is what I'm curious about- while I'm all set regarding storage for the next few years (I hedged my bets and made sure that all my external drives are both Firewire and USB 3), I wonder which of these will take off, and which will slowly fade away... Faster storage would be nice- it's not bad now, but 6Gb/s SATA would considerably speed up Photoshop.