Simple fact Is tho for a one two man shop you can rig the iMacs to fly.
If you're a straight shooter (little or no editingl - then it's a toss up going forward.
I haven't really tested it in depth. For the things Biscardi mentions, I am not surprised performance was as he described, although cpu performance was never my biggest hangup with it. I've gone into greater detail before. Freelancers in any field usually base purchases on something that can cover 90% of their workload without hiccups. In spite of what I dislike about it, I wonder if the imac would have seen even greater growth with a better range of potential thunderbolt DAS options. I would still say listing Raid 5 support on that box is a mistake as it uses standard drives, and I have no idea whether it has ECC ram. ECC ram isn't critical in all situations, but that changes when we're talking about Raid 5, as a single bit error during rebuild can cause the array to crash. At that point you have more downtime restoring from backup.
Just like if they put a 640M in the current Mac Mini. A $999 2.6 Ghz Quad i7 Mac Mini with a 512MB GT 640M would crater 21" iMac sales.
I have no doubt that Apple believes that, but I wonder if it's true. A mini buyer seems to me to be a much different animal than an iMac buyer. I don't think either is particularly specification-conscious, so I don't think similarity in configuration would be all that likely to sway most buyers of those machines from one to the other.
An iMac buyer wants a complete general-purpose computer, ready to go. I suspect most mini buyers have a specific application in mind (server, media center, cabinet or stand computer), with the display being either superfluous or a disadvantage due to size.
I have no data to back that up, it's just a gut feeling, and obviously it wouldn't apply to AI readers who tend to be much more tech-aware than the average buyer.
The idea that a cheaper Mac Pro will impact iMac slaes is just laughable.
I don't know... for our latest round of capital requests I put together a good/better/best proposal for a simple audio workstation, and for the first time ever I based the cheapest option on an iMac. The performance is good enough, it includes a display and has sufficient I/O to do that job.
If Apple were to offer a Mac Pro in a price category that let me add storage and display without costing a lot more than the iMac (i.e. within $500 or so) I'd almost certainly grab that instead.
I have no doubt that Apple believes that, but I wonder if it's true. A mini buyer seems to me to be a much different animal than an iMac buyer. I don't think either is particularly specification-conscious, so I don't think similarity in configuration would be all that likely to sway most buyers of those machines from one to the other.
An iMac buyer wants a complete general-purpose computer, ready to go. I suspect most mini buyers have a specific application in mind (server, media center, cabinet or stand computer), with the display being either superfluous or a disadvantage due to size.
I have no data to back that up, it's just a gut feeling, and obviously it wouldn't apply to AI readers who tend to be much more tech-aware than the average buyer.
I fully concur with your assumption. I bought a Mini and put it under the TV, as the AppleTV can't do everything. I also use the Mini as a download server (TV shows that I can't rent on iTunes etc). I was surprised the little thing can do quite a lot of stuff concurrently, and perhaps will function as a 'good enough PC' if you put a SSD in it. Even though I've always bought Mac Pro's, I actually think about trying a Mini out if my current 5.1 MP dies. Supposedly it has no real trouble running Aperture with a 100GB DB.
I wouldn't choose an iMac as I don't like the reflection from the glossy screens. I know, it's way less than it used to be, but for me a matte screen rules.
If it fails sales wise you can kiss the machine good by. I really think Apple will dump it like an XServe if they can't improve sales over the old model. Frankly the machines design is a big gamble, it is more of what I expected from an XMac than a Pro, if you remove the workstation processors. That is good and bad. The good part is that they might meet a rational price point with low end XEONs in the base model. Apples marketing of the AIR gives me hope that they will be aggressive pricing wise. The really bad part from my perspective is the lack of slots, this means that some functionality will never be supported on the machine. The funny thing here is that I don't see the AV professionals having a problem here as there is enough demand to make sure the hardware they use transitions to TB.
I agree that the machines design is a gamble - who knows how it will pan out ?
Difficult for Apple to try and accomodate the different requirements of all users.
I was in favour of a half rack style of machine design, something that could be incorporated in with existing hardware - well the new MacPro certainly isn't anything like that lol, wtf do I know.
It's a standalone design and engineering statement.
I can see this new design as being functional for most once users get their head around it and what it means.
One thing Apple has made crystal clear going forward - it's make the change to TB and TB devices or move to another platform.
I don't know... for our latest round of capital requests I put together a good/better/best proposal for a simple audio workstation, and for the first time ever I based the cheapest option on an iMac. The performance is good enough, it includes a display and has sufficient I/O to do that job.
Doesn't this sort of prove some of my points though, you aren't considering that iMac to be anything more than an entry level machine.
If Apple were to offer a Mac Pro in a price category that let me add storage and display without costing a lot more than the iMac (i.e. within $500 or so) I'd almost certainly grab that instead.
This is my point, most professionals wold be far better off with the new Mac Pro than an iMac Further cost is a factor here. The interesting thing with the new iMac is its support for cheap monitors right out of the box. The HDMI port means that very cheap monitors are possible though maybe not ideal. These won't be 4K monitors obviously but certainly a low end choice.
The thing is the Mac Pro can grow in ways the iMac really can't, that due to more I/O and dramatically better processing capabilities. I could easily see people out growing an iMac or worst having the machine become a problem with a simple software update where an app becomes too much for the machine. In the end an entry level solution at best.
Don't get me wrong there is nothing wrong with entry level, I just don't see most professionals seeing such a machine as a long term investment. At least not this year. Every year the equations change which you pretty much highlight here - is it fair to say this is the first year you have even considered an iMac? This is an important consideration because in 2014 we should see 14 nm chips from Intel which has the potential to redefine the iMac yet again. Ideally Apple would see too more TB ports in the iMac if they really wanted to set it off as a pro machine.
On the flips side 14 nm could make the Mac Pro far more interesting than it will be at release this fall. The advantage will stay with the pro for a very long time. While it is true that the iMac can pick up more and more work traditionally done on a Mac Pro, the Mac Pro just continues to pick up more capability support more advance app functions.
I fully concur with your assumption. I bought a Mini and put it under the TV, as the AppleTV can't do everything. I also use the Mini as a download server (TV shows that I can't rent on iTunes etc). I was surprised the little thing can do quite a lot of stuff concurrently, and perhaps will function as a 'good enough PC' if you put a SSD in it.
The Mini for a long time was a terrible machine value wise, it has gotten a lot better in the last couple of revisions in that respect. Much of that is due to dramatically lower power requirements in new processors. Today it is a very passable light duty workstation as TB gives it exceptional capabilities considering its size.
Even though I've always bought Mac Pro's, I actually think about trying a Mini out if my current 5.1 MP dies. Supposedly it has no real trouble running Aperture with a 100GB DB.
If you are considering a Mini I might suggest that you didn't need a Mac Pro in the first place. A Mini shouldn't be a problem as I've had Aperture running on old laptops just fine. Of course that doesn't mean that the current Minis do much for you in the way of OpenCL support.
For that matter we don't know how the next version of Aperture will be able to leverage the Mini. For me this is the biggest issue with going extreme low end. You either loose features completely in an app or the app reverts to very low performance implementations of a feature. So you have to ask your self what will the next version of Aperture do with respect to the Mini. Also any plug ins that might leverage GPU computing become a problem.
The way I look at it is that we have entry level hardware which is fine, nut you have to be careful when you take advice like "supposedly it has no real trouble running Aperture". It could very well be perfect or it could crap out on you with long run times if a pet feature isn't supported well. You need to investigate these things carefully and not take anyone opinion as gospel.
As a side note the same thing applies to the Mini that applies to the other Macs, each hardware update changes the equation. That is Haswell could make for far better Aperture performance even if Apple never gets around to supporting OpenCL on the GPU. The new multi media instructions, once put to use, should have a very good impact on multimedia apps. In effect what is a passable machine is influx on a yearly basis.
I wouldn't choose an iMac as I don't like the reflection from the glossy screens. I know, it's way less than it used to be, but for me a matte screen rules.
Interesting because I reject iMacs based largely on service ability which is pretty bad. I've had far to many hard drives fail over the years so access to secondary store is important.
I agree that the machines design is a gamble - who knows how it will pan out ?
I have good feelings as far as success goes. It might be an uphill battle but the machine does have some very good qualities. Almost everyone agrees price will be a factor and as a result it needs to actually be cheaper than the old models. If Apple tries to price it in the same slots as the old machine I see it getting rejected out of hand due to the expense of adding external storage.
Difficult for Apple to try and accomodate the different requirements of all users.
Actually it is impossible. For one you have a large number of users that think they know what they need but in reality they don't know anything. Instead they look to the past and what worked for them.
While I'm not 100% with the new architecture the one thing I do know is that it puts a lot of power in a small box that should lead to lots of happy users. I'm really hoping that Apples design goal was to lower the cost of the base Mac Pro significantly. If not Apple will have failed to accommodate enough users to matter.
I was in favour of a half rack style of machine design, something that could be incorporated in with existing hardware - well the new MacPro certainly isn't anything like that lol, wtf do I know.
If you look back at my posts I was of a similar mind. Such a machine could have been made to serve the needs of many users.
As for the new design it has grown on me a bit. It is a compact design that eliminates many parts to allow for high performance. It unfortunately leaves out many features that I'd consider important
It's a standalone design and engineering statement.
Certainly interesting and a machine that uniquely solves some of the issues with high performance machines. Possibly number one here being how do you cool hot running GPU cards. Most machines of this class would have around seven fans if not more to try to keep everything within running temperatures.
Given all of that though I see this machine almost as a proof of concept like the first iPad. Sure it will work but the next generation running on 14nm technology will likely be even more impressive.
I can see this new design as being functional for most once users get their head around it and what it means.
Yeah it does take a few minutes of thought.
One thing Apple has made crystal clear going forward - it's make the change to TB and TB devices or move to another platform.
What if your vendor of choice decides that going forward only TB will be supported? For some products TB could end up being the better technical solution.
What if your vendor of choice decides that going forward only TB will be supported? For some products TB could end up being the better technical solution.
Yep - Ive arrived at that point. I'd better start allocating some money for TB externals, no matter which machine I end up buying.
A year ago or so I was skeptical of TB and what timeframe Intel would move forward with it. Well they've surprised me with TB 2 - the main third parties are more or less with it. AJA and BMD and the others are to be commended at their rate of adoption. Those companies are really responsive to new tech and how it might be applied.
The only stumbling block to wide adoption I can see is the pc guys and their seeming unwillingness to change. Remember how long it took them to offer fw ? Many of the comments on their boards don't seem to give it a chance, "The next FireWire, ho ho ho " etc They seem to think usb3 is it for the moment. Their arguments are largely based around the price and cost of cables. oh well, that pretty well sums up the pc world all over doesn't it ?
cheers, r
The benefits of TB are amazing over USB3, even considering cost IMO.
Speed. Speed. Speed. Even more speed with TB2. Daisy chaining. Display.
I found in my tests that USB3 drives don't work well with USB3 hubs (the latter are tough to find). Transfer rates dropped in half using the hub, even with only one device attached to it. That eliminated USB3 drives as a cheaper alternative to TB external drives.
I still use USB3 drives for Time Machine and off-site storage as well as transferring huge files in-house... through direct connect only. But I will stick with TB for regular use and any large set-up.
Yep - Ive arrived at that point. I'd better start allocating some money for TB externals, no matter which machine I end up buying.
A year ago or so I was skeptical of TB and what timeframe Intel would move forward with it. Well they've surprised me with TB 2 - the main third parties are more or less with it. AJA and BMD and the others are to be commended at their rate of adoption. Those companies are really responsive to new tech and how it might be applied.
The only stumbling block to wide adoption I can see is the pc guys and their seeming unwillingness to change. Remember how long it took them to offer fw ? Many of the comments on their boards don't seem to give it a chance, "The next FireWire, ho ho ho " etc They seem to think usb3 is it for the moment. Their arguments are largely based around the price and cost of cables. oh well, that pretty well sums up the pc world all over doesn't it ?
cheers, r
USB3 is a lot cheaper to implement. Consider that you get it with the chipset in most cases (not workstations or servers). Thunderbolt is really aimed at machines with embedded and especially integrated graphics. There isn't a standard reference configuration from intel to implement it on everything, and if they really wanted ubiquity, it would have become part of the chipset.
hmm, I don't go along with the idea that its aimed at integrated graphics and embedded chipsets only. The tech has far wider potential as Im sure you know. And as can be seen already.
Intel has to make its intentions clear and fairly soon, I agree. At the moment they're are straddling the fence for sure. Apple, as far as I can see, are backing it 100%
And this is pretty much the discussion in a nutshell.
One reason for that position is that over the years the vast majority of computer failure I've had to deal with have been secondary store related probably followed closely by power supply failures. I just think it is asinine to design the iMac in the way it is today. It wouldn't bE that difficult to allow easy access to the hard drive and if they go to SSD blades only, it would be even easier. In my book it is a fatal flaw design wise.
hmm, I don't go along with the idea that its aimed at integrated graphics and embedded chipsets only.
The tech has far wider potential as Im sure you know. And as can be seen already.
Intel has to make its intentions clear and fairly soon, I agree. At the moment they're are straddling the fence for sure. Apple, as far as I can see, are backing it 100%
If you look at the way it's designed, it essentially bets on some form of embedded graphics. The mac pro design is somewhat of a rube goldberg one there. It has 3 thunderbolt chips hooked up to two cards. I still don't understand why everyone bought into the thunderbolt 2 kool-aid. It seems as if they merely added a half duplex mode to deal with higher bandwidth requirements inherent to 4K streams.
Quote:
Originally Posted by wizard69
]
One reason for that position is that over the years the vast majority of computer failure I've had to deal with have been secondary store related probably followed closely by power supply failures. I just think it is asinine to design the iMac in the way it is today. It wouldn't bE that difficult to allow easy access to the hard drive and if they go to SSD blades only, it would be even easier. In my book it is a fatal flaw design wise.
The ssd blades they have used thus far aren't really standard implementations like mSATA or SATA express types. They're whatever Apple concocted. Even if you could easily access them, you might not be able to do much. Personally I think Apple's design team really hates visible seams.
damn you, hmm. Now Im going to have to research further ! Thanks, I think As I have said before - for many of us it's as much about third party devices as it is about the computer. External equipment can monster the cost of the computer easily. Might let this shake down for a little while longer. cheers
If you look at the way it's designed, it essentially bets on some form of embedded graphics. The mac pro design is somewhat of a rube goldberg one there. It has 3 thunderbolt chips hooked up to two cards.
It isn't anymore Rube Goldberg than the iMac. There is nothing about TB that bets on embedded graphics.
I still don't understand why everyone bought into the thunderbolt 2 kool-aid. It seems as if they merely added a half duplex mode to deal with higher bandwidth requirements inherent to 4K streams.
Because the bandwidth isn't suitable for many uses in the TB 1 design? Video is one example but TB 1 is slow for many uses including storage arrays. TB2 gives users and designers sim breathing room. I do wonder if TB 2 forces a port for video usage to be video only though when hooked up to a display 4 K or greater.
I really don't get the KoolAid statement, without TB 2 the new Mac Pro would be far less of a machine.
The ssd blades they have used thus far aren't really standard implementations like mSATA or SATA express types. They're whatever Apple concocted. Even if you could easily access them, you might not be able to do much. Personally I think Apple's design team really hates visible seams.
I realize there are mental issues at Apple with seams, I think it is time for them to see the doctor and get over this defect! Personally I wouldn't want nor recommend an iMac to anybody until they address this problem.
As to the blade implementations you bring up a good question. The physical interface is obviously Apples own as I don't see it documented with any standards body. At least not yet. The virtual interface is an even bigger mystery, I'm wondering if NVMe is being used. Apple does not appear to be a member of the standards group driving NVMe which frankly saddens me. I can understand Apple going its own way with the physical profile, but it would be nice to know that they are supporting a standard industry command set. This simply to allow for lots of third party competition.
In any event it looks lke SATA is dead at Apple. This is a very good thing in my estimation, hopefully the iMac and Mini complete the transition this year.
damn you, hmm.
Now Im going to have to research further !
Unfortunately information is sparse at the moment. If you find some good documentation on TB 2 post a link.
Thanks, I think
As I have said before - for many of us it's as much about third party devices as it is about the computer.
External equipment can monster the cost of the computer easily.
This is the case in many industries, buy a milling machine for $x and add another $1.5x for tooling. The problem here is the mystery of what the real capabilities of the new Mac Pro are. Yeah Apple has some specs up but we don't know how high bandwidth things like 4K displays impact the TB multi drop capability. Or for that matter the companion port on each chip.
Might let this shake down for a little while longer.
cheers
It is always good to allow for some shake down time. Even the AIRs had a few glitches that needed ironing out.
It isn't anymore Rube Goldberg than the iMac. There is nothing about TB that bets on embedded graphics.
Because the bandwidth isn't suitable for many uses in the TB 1 design? Video is one example but TB 1 is slow for many uses including storage arrays. TB2 gives users and designers sim breathing room. I do wonder if TB 2 forces a port for video usage to be video only though when hooked up to a display 4 K or greater.
The i7 3770 has integrated graphics included. The chip itself was designed as part of the logic board. I'm not sure whether 4K completely saturates it. I haven't done the math at 60hz, although I would assume it may. As I mentioned the increase in bandwidth is accomplished by a half duplex mode and channel bonding. I could be wrong on this. My bet was on falcon ridge. Your favorite news site suggests redwood ridge could be out sooner than I expected, and that is the one that bumps the per channel bandwidth. Falcon ridge to me is more of a stop gap to add in 4K support. It could also be a cost thing. Falcon ridge is the first that I really consider truly impressive. I could definitely see it appearing on higher end PC notebooks as well, and it does bet on notebooks and integrated graphics. Intel came up with these to align with areas where they see continued growth.
Quote:
I really don't get the KoolAid statement, without TB 2 the new Mac Pro would be far less of a machine.
I realize there are mental issues at Apple with seams, I think it is time for them to see the doctor and get over this defect! Personally I wouldn't want nor recommend an iMac to anybody until they address this problem.
It's not really harvesting more bandwidth so much as it is boning upstream and downstream channels. 4K might completely saturate a chip. Further the "thunderbolt 2" designation is really more of a rebranding than anything. It's more of a 1.2 kind of revision. A year ago what people considered thunderbolt 2 was a chip that won't be out for another 1-2 years at this point. It brings bandwidth up to PCIe 3.0 levels.
Quote:
As to the blade implementations you bring up a good question. The physical interface is obviously Apples own as I don't see it documented with any standards body. At least not yet. The virtual interface is an even bigger mystery, I'm wondering if NVMe is being used. Apple does not appear to be a member of the standards group driving NVMe which frankly saddens me. I can understand Apple going its own way with the physical profile, but it would be nice to know that they are supporting a standard industry command set. This simply to allow for lots of third party competition.
Standard is generally good.
Quote:
In any event it looks lke SATA is dead at Apple. This is a very good thing in my estimation, hopefully the iMac and Mini complete the transition this year.
The rmbp still uses SATA, just with its different connector. They may adopt something like sata express (sata over pcie). Personally I hate proprietary solutions. They tend to be very expensive if something fails.
The i7 3770 has integrated graphics included. The chip itself was designed as part of the logic board. I'm not sure whether 4K completely saturates it. I haven't done the math at 60hz, although I would assume it may. As I mentioned the increase in bandwidth is accomplished by a half duplex mode and channel bonding. I could be wrong on this. My bet was on falcon ridge. Your favorite news site:devil: suggests redwood ridge could be out sooner than I expected, and that is the one that bumps the per channel bandwidth.
I'm not sure they increased the bandwidth per channel, it looks more like they bonded channels to get one pipe in each direction. If I read the graphic right we still get full duplex that is a good thing.
Falcon ridge to me is more of a stop gap to add in 4K support. It could also be a cost thing. Falcon ridge is the first that I really consider truly impressive. I could definitely see it appearing on higher end PC notebooks as well, and it does bet on notebooks and integrated graphics. Intel came up with these to align with areas where they see continued growth.
As I've said before I think Apple has already gotten 90% of what they want out of TB, it is an ideal docking solution for them. That it can be exploited on the new Mac Pro is just icing on the cake so to speak. Contrary to popular opinion I don't think Apple ever intends to replace USB with TB
It's not really harvesting more bandwidth so much as it is boning upstream and downstream channels. 4K might completely saturate a chip. Further the "thunderbolt 2" designation is really more of a rebranding than anything. It's more of a 1.2 kind of revision. A year ago what people considered thunderbolt 2 was a chip that won't be out for another 1-2 years at this point. It brings bandwidth up to PCIe 3.0 levels.
I haven't had a real chance to do some reading on TB 2 but it see it as more than a rebranding. The 1.2 designation may be correct though as it looks like that was the intention all along.
Standard is generally good.
This is one problem I do have with Apple, public documentation sucks.
The rmbp still uses SATA, just with its different connector. They may adopt something like sata express (sata over pcie). Personally I hate proprietary solutions. They tend to be very expensive if something fails.
True but after the PCI Express solution in the AIR I can't see Apple staying with SATA.
Comments
Quote:
Originally Posted by RobM
Simple fact Is tho for a one two man shop you can rig the iMacs to fly.
If you're a straight shooter (little or no editingl - then it's a toss up going forward.
I haven't really tested it in depth. For the things Biscardi mentions, I am not surprised performance was as he described, although cpu performance was never my biggest hangup with it. I've gone into greater detail before. Freelancers in any field usually base purchases on something that can cover 90% of their workload without hiccups. In spite of what I dislike about it, I wonder if the imac would have seen even greater growth with a better range of potential thunderbolt DAS options. I would still say listing Raid 5 support on that box is a mistake as it uses standard drives, and I have no idea whether it has ECC ram. ECC ram isn't critical in all situations, but that changes when we're talking about Raid 5, as a single bit error during rebuild can cause the array to crash. At that point you have more downtime restoring from backup.
Quote:
Originally Posted by nht
Just like if they put a 640M in the current Mac Mini. A $999 2.6 Ghz Quad i7 Mac Mini with a 512MB GT 640M would crater 21" iMac sales.
I have no doubt that Apple believes that, but I wonder if it's true. A mini buyer seems to me to be a much different animal than an iMac buyer. I don't think either is particularly specification-conscious, so I don't think similarity in configuration would be all that likely to sway most buyers of those machines from one to the other.
An iMac buyer wants a complete general-purpose computer, ready to go. I suspect most mini buyers have a specific application in mind (server, media center, cabinet or stand computer), with the display being either superfluous or a disadvantage due to size.
I have no data to back that up, it's just a gut feeling, and obviously it wouldn't apply to AI readers who tend to be much more tech-aware than the average buyer.
Quote:
Originally Posted by wizard69
The idea that a cheaper Mac Pro will impact iMac slaes is just laughable.
I don't know... for our latest round of capital requests I put together a good/better/best proposal for a simple audio workstation, and for the first time ever I based the cheapest option on an iMac. The performance is good enough, it includes a display and has sufficient I/O to do that job.
If Apple were to offer a Mac Pro in a price category that let me add storage and display without costing a lot more than the iMac (i.e. within $500 or so) I'd almost certainly grab that instead.
I fully concur with your assumption. I bought a Mini and put it under the TV, as the AppleTV can't do everything. I also use the Mini as a download server (TV shows that I can't rent on iTunes etc). I was surprised the little thing can do quite a lot of stuff concurrently, and perhaps will function as a 'good enough PC' if you put a SSD in it. Even though I've always bought Mac Pro's, I actually think about trying a Mini out if my current 5.1 MP dies. Supposedly it has no real trouble running Aperture with a 100GB DB.
I wouldn't choose an iMac as I don't like the reflection from the glossy screens. I know, it's way less than it used to be, but for me a matte screen rules.
I agree that the machines design is a gamble - who knows how it will pan out ?
Difficult for Apple to try and accomodate the different requirements of all users.
I was in favour of a half rack style of machine design, something that could be incorporated in with existing hardware - well the new MacPro certainly isn't anything like that lol, wtf do I know.
It's a standalone design and engineering statement.
I can see this new design as being functional for most once users get their head around it and what it means.
One thing Apple has made crystal clear going forward - it's make the change to TB and TB devices or move to another platform.
This is my point, most professionals wold be far better off with the new Mac Pro than an iMac Further cost is a factor here. The interesting thing with the new iMac is its support for cheap monitors right out of the box. The HDMI port means that very cheap monitors are possible though maybe not ideal. These won't be 4K monitors obviously but certainly a low end choice.
The thing is the Mac Pro can grow in ways the iMac really can't, that due to more I/O and dramatically better processing capabilities. I could easily see people out growing an iMac or worst having the machine become a problem with a simple software update where an app becomes too much for the machine. In the end an entry level solution at best.
Don't get me wrong there is nothing wrong with entry level, I just don't see most professionals seeing such a machine as a long term investment. At least not this year. Every year the equations change which you pretty much highlight here - is it fair to say this is the first year you have even considered an iMac? This is an important consideration because in 2014 we should see 14 nm chips from Intel which has the potential to redefine the iMac yet again. Ideally Apple would see too more TB ports in the iMac if they really wanted to set it off as a pro machine.
On the flips side 14 nm could make the Mac Pro far more interesting than it will be at release this fall. The advantage will stay with the pro for a very long time. While it is true that the iMac can pick up more and more work traditionally done on a Mac Pro, the Mac Pro just continues to pick up more capability support more advance app functions.
For that matter we don't know how the next version of Aperture will be able to leverage the Mini. For me this is the biggest issue with going extreme low end. You either loose features completely in an app or the app reverts to very low performance implementations of a feature. So you have to ask your self what will the next version of Aperture do with respect to the Mini. Also any plug ins that might leverage GPU computing become a problem.
The way I look at it is that we have entry level hardware which is fine, nut you have to be careful when you take advice like "supposedly it has no real trouble running Aperture". It could very well be perfect or it could crap out on you with long run times if a pet feature isn't supported well. You need to investigate these things carefully and not take anyone opinion as gospel.
As a side note the same thing applies to the Mini that applies to the other Macs, each hardware update changes the equation. That is Haswell could make for far better Aperture performance even if Apple never gets around to supporting OpenCL on the GPU. The new multi media instructions, once put to use, should have a very good impact on multimedia apps. In effect what is a passable machine is influx on a yearly basis. Interesting because I reject iMacs based largely on service ability which is pretty bad. I've had far to many hard drives fail over the years so access to secondary store is important.
While I'm not 100% with the new architecture the one thing I do know is that it puts a lot of power in a small box that should lead to lots of happy users. I'm really hoping that Apples design goal was to lower the cost of the base Mac Pro significantly. If not Apple will have failed to accommodate enough users to matter. If you look back at my posts I was of a similar mind. Such a machine could have been made to serve the needs of many users.
As for the new design it has grown on me a bit. It is a compact design that eliminates many parts to allow for high performance. It unfortunately leaves out many features that I'd consider important Certainly interesting and a machine that uniquely solves some of the issues with high performance machines. Possibly number one here being how do you cool hot running GPU cards. Most machines of this class would have around seven fans if not more to try to keep everything within running temperatures.
Given all of that though I see this machine almost as a proof of concept like the first iPad. Sure it will work but the next generation running on 14nm technology will likely be even more impressive. Yeah it does take a few minutes of thought.
What if your vendor of choice decides that going forward only TB will be supported? For some products TB could end up being the better technical solution.
Yep - Ive arrived at that point. I'd better start allocating some money for TB externals, no matter which machine I end up buying.
A year ago or so I was skeptical of TB and what timeframe Intel would move forward with it. Well they've surprised me with TB 2 - the main third parties are more or less with it. AJA and BMD and the others are to be commended at their rate of adoption. Those companies are really responsive to new tech and how it might be applied.
The only stumbling block to wide adoption I can see is the pc guys and their seeming unwillingness to change. Remember how long it took them to offer fw ? Many of the comments on their boards don't seem to give it a chance, "The next FireWire, ho ho ho " etc They seem to think usb3 is it for the moment. Their arguments are largely based around the price and cost of cables. oh well, that pretty well sums up the pc world all over doesn't it ?
cheers, r
The benefits of TB are amazing over USB3, even considering cost IMO.
Speed. Speed. Speed. Even more speed with TB2. Daisy chaining. Display.
I found in my tests that USB3 drives don't work well with USB3 hubs (the latter are tough to find). Transfer rates dropped in half using the hub, even with only one device attached to it. That eliminated USB3 drives as a cheaper alternative to TB external drives.
I still use USB3 drives for Time Machine and off-site storage as well as transferring huge files in-house... through direct connect only. But I will stick with TB for regular use and any large set-up.
Quote:
Originally Posted by RobM
Yep - Ive arrived at that point. I'd better start allocating some money for TB externals, no matter which machine I end up buying.
A year ago or so I was skeptical of TB and what timeframe Intel would move forward with it. Well they've surprised me with TB 2 - the main third parties are more or less with it. AJA and BMD and the others are to be commended at their rate of adoption. Those companies are really responsive to new tech and how it might be applied.
The only stumbling block to wide adoption I can see is the pc guys and their seeming unwillingness to change. Remember how long it took them to offer fw ? Many of the comments on their boards don't seem to give it a chance, "The next FireWire, ho ho ho " etc They seem to think usb3 is it for the moment. Their arguments are largely based around the price and cost of cables. oh well, that pretty well sums up the pc world all over doesn't it ?
cheers, r
USB3 is a lot cheaper to implement. Consider that you get it with the chipset in most cases (not workstations or servers). Thunderbolt is really aimed at machines with embedded and especially integrated graphics. There isn't a standard reference configuration from intel to implement it on everything, and if they really wanted ubiquity, it would have become part of the chipset.
The tech has far wider potential as Im sure you know. And as can be seen already.
Intel has to make its intentions clear and fairly soon, I agree. At the moment they're are straddling the fence for sure. Apple, as far as I can see, are backing it 100%
Quote:
Originally Posted by wizard69
Maybe they work well in some applications but I dismiss the iMacs out of hand...
And this is pretty much the discussion in a nutshell.
One reason for that position is that over the years the vast majority of computer failure I've had to deal with have been secondary store related probably followed closely by power supply failures. I just think it is asinine to design the iMac in the way it is today. It wouldn't bE that difficult to allow easy access to the hard drive and if they go to SSD blades only, it would be even easier. In my book it is a fatal flaw design wise.
Quote:
Originally Posted by RobM
hmm, I don't go along with the idea that its aimed at integrated graphics and embedded chipsets only.
The tech has far wider potential as Im sure you know. And as can be seen already.
Intel has to make its intentions clear and fairly soon, I agree. At the moment they're are straddling the fence for sure. Apple, as far as I can see, are backing it 100%
If you look at the way it's designed, it essentially bets on some form of embedded graphics. The mac pro design is somewhat of a rube goldberg one there. It has 3 thunderbolt chips hooked up to two cards. I still don't understand why everyone bought into the thunderbolt 2 kool-aid. It seems as if they merely added a half duplex mode to deal with higher bandwidth requirements inherent to 4K streams.
Quote:
Originally Posted by wizard69
]
One reason for that position is that over the years the vast majority of computer failure I've had to deal with have been secondary store related probably followed closely by power supply failures. I just think it is asinine to design the iMac in the way it is today. It wouldn't bE that difficult to allow easy access to the hard drive and if they go to SSD blades only, it would be even easier. In my book it is a fatal flaw design wise.
The ssd blades they have used thus far aren't really standard implementations like mSATA or SATA express types. They're whatever Apple concocted. Even if you could easily access them, you might not be able to do much. Personally I think Apple's design team really hates visible seams.
Now Im going to have to research further !
Thanks, I think
As I have said before - for many of us it's as much about third party devices as it is about the computer.
External equipment can monster the cost of the computer easily.
Might let this shake down for a little while longer.
cheers
I really don't get the KoolAid statement, without TB 2 the new Mac Pro would be far less of a machine.
I realize there are mental issues at Apple with seams, I think it is time for them to see the doctor and get over this defect! Personally I wouldn't want nor recommend an iMac to anybody until they address this problem.
As to the blade implementations you bring up a good question. The physical interface is obviously Apples own as I don't see it documented with any standards body. At least not yet. The virtual interface is an even bigger mystery, I'm wondering if NVMe is being used. Apple does not appear to be a member of the standards group driving NVMe which frankly saddens me. I can understand Apple going its own way with the physical profile, but it would be nice to know that they are supporting a standard industry command set. This simply to allow for lots of third party competition.
In any event it looks lke SATA is dead at Apple. This is a very good thing in my estimation, hopefully the iMac and Mini complete the transition this year.
It is always good to allow for some shake down time. Even the AIRs had a few glitches that needed ironing out.
Quote:
Originally Posted by wizard69
It isn't anymore Rube Goldberg than the iMac. There is nothing about TB that bets on embedded graphics.
Because the bandwidth isn't suitable for many uses in the TB 1 design? Video is one example but TB 1 is slow for many uses including storage arrays. TB2 gives users and designers sim breathing room. I do wonder if TB 2 forces a port for video usage to be video only though when hooked up to a display 4 K or greater.
The i7 3770 has integrated graphics included. The chip itself was designed as part of the logic board. I'm not sure whether 4K completely saturates it. I haven't done the math at 60hz, although I would assume it may. As I mentioned the increase in bandwidth is accomplished by a half duplex mode and channel bonding. I could be wrong on this. My bet was on falcon ridge. Your favorite news site
Quote:
I really don't get the KoolAid statement, without TB 2 the new Mac Pro would be far less of a machine.
I realize there are mental issues at Apple with seams, I think it is time for them to see the doctor and get over this defect! Personally I wouldn't want nor recommend an iMac to anybody until they address this problem.
It's not really harvesting more bandwidth so much as it is boning upstream and downstream channels. 4K might completely saturate a chip. Further the "thunderbolt 2" designation is really more of a rebranding than anything. It's more of a 1.2 kind of revision. A year ago what people considered thunderbolt 2 was a chip that won't be out for another 1-2 years at this point. It brings bandwidth up to PCIe 3.0 levels.
Quote:
As to the blade implementations you bring up a good question. The physical interface is obviously Apples own as I don't see it documented with any standards body. At least not yet. The virtual interface is an even bigger mystery, I'm wondering if NVMe is being used. Apple does not appear to be a member of the standards group driving NVMe which frankly saddens me. I can understand Apple going its own way with the physical profile, but it would be nice to know that they are supporting a standard industry command set. This simply to allow for lots of third party competition.
Standard is generally good.
Quote:
In any event it looks lke SATA is dead at Apple. This is a very good thing in my estimation, hopefully the iMac and Mini complete the transition this year.
The rmbp still uses SATA, just with its different connector. They may adopt something like sata express (sata over pcie). Personally I hate proprietary solutions. They tend to be very expensive if something fails.
True but after the PCI Express solution in the AIR I can't see Apple staying with SATA.