the new intel-powermacs will probably (?) use a chipset with the ich8 southbridge.
i dont know much about that one, but ich7 found in for example the 975x chipset got support for one 16x-lane pci-e or two 8x-lane, six 1x-lane pci-e and six 'regular' pci slots.
(and you'll get a free tpm too, no extra cost!)
i dont think the ich8 will be much worse...
edit: now (after writing all that) i think i get what you're asking.
i guess you could 'route' some of the pci-e lanes in some kind of extension cable to a break-out slot-box, but i dont think it'll be pretty...
offering two sizes to begin with would probably save apple a lot of headache...
Of course, properly, Apple will never offer a machine that includes the old PCI specification.
The current machine is very highly spec'd. It has only one 16 lane slot, but unlike other Express machines, it also has an 8 lane slot, and two 4 lane slots That's pretty good.
If they did offer a 6 slot machine, the question is how they would allocate them. Would they offer 2 board SLI or Crossfire? Will the continue to offer 8 and 4 lane slots.
Will they even offer SLI and Crossfire on another 4 slot machine?
One reason why more slots is important is because all of the high end graphics cards take up the space of the next slot as well. So a 4 slot machine efectively becomes a 3 slot one. If you are using either SLI or CF with two high end boards, you might find that you have NO slots left.
That's where the question of 6 slots comes into play. If you need a firewire/USB board, there might be no where to put one.
A 2 slot machine would become a 1 slot machine with a high end board.
This subject comes up on Mac sites because Mac sites are filled with computer enthusiasts.
I'm not a computer enthusiast and I don't know if this statement can be proved.
Quote:
Originally posted by hmurchison
A low cost expandable headless Mac isn't going to increase Marketshare. Apple has copious amounts of data and that data is likely going to show that no matter what you do for a desktop it's not going to affect much sales change because PEOPLE WANT LAPTOPS.
I haven't a clue how much market research Apple has done and this can not be proved. However, I have read on the internet(re: veracity to be questioned here, we all know how accurate the internet is) numerous times that Apple performs a minimal amount of market research.
But we do know that there are/were AIO Windows computers offered and virtually no one bought them. This alone is indirect evidence that the majority of computer buyers, for whatever reasons, valid or not, do not want AIO computers.
Quote:
Originally posted by hmurchison
The people that want desktops generally have an ideal selection of specs they want. If an AIO meets those specs then they will generally have no apprehension towards buying it. Once I told potential iMac purchasers that they weren't locked out of better monitor choices they were more at ease.
Anecdotal evidence, while interesting and important won't convince me that the largest segment of computer buyers want or would buy an AIO.
Quote:
Originally posted by hmurchison
I really think the days of a big roomy box with loud fans is coming to an end. Workstations will always exist but then again people who need workstations aren't really complaining vociferously about price.
I, at least, am not talking about " a big roomy box with loud fans".
Quote:
Originally posted by melgross
Don't forget that Apple had a single cpu tower G5 for $1,500. Just making the changes I suggested would have brought that down to $999. This didn't have anything to do with the iMac. And, it would have kept a shorter aluminum case. I did some costing at the time
Yes they did. However, as you pointed out, Apple neutered it. I was the same processer, slower bus, limited ram capable and PCI instead of PCI-X as the iMac and for the exact same cost of the iMac that came with a 17" LCD screen. If Apple had made it a dual processor, with the faster bus, and PCI-X and same ram capabilities as the other towers it might have sold better. But who in their right mind would pay for this machine when they could either buy the iMac or spend the extra money for a dual processor machine?
I'm not saying that Apple didn't make the right choices in the past. Maybe they needed to protect iMac sales or didn't want to sell the low end tower and force people to decide between the iMac and the dual processor tower.
Seriously though, if Apple truly wants to increase market share and now we have them on record twice saying that increasing market share is a goal, then it is IMHO that they will have to add a computer to their line up that targets the arguably the largest segment of computer buyers.
edit cleaned up some grammer, if I left more they ain't getting fixed.
While I may offer anecdotal evidence my anecdotes are supported by the sales figures(which I shall not divulge) for a Top 5 reseller thus they carry a wee bit more weight that the typical mac fans conjecture about what people want.
I think it's clear that,if given the choice between having more computing power for your dollar versus less, most of us would choose the former.
However Apple has decided on their current structure and frankly we vote with our wallets. If they were doing the wrong things sales would be declining rapidly.
I have no faith in the notion that a somewhat expandable headless Mac would increase marketshare at all. This is based on my own experiences and others are free to disagree.
I have no faith in the notion that a somewhat expandable headless Mac would increase marketshare at all. This is based on my own experiences and others are free to disagree.
I haven't read it (but I will) Probably because Conroe isn't a dual socket processor. The Opteron, and Woodcrest are. Or they are just testing the highest spec'd processors from both vendors. OK, I'll go read it.
[edit] OK, read it. It's all about server workstation processors really. The part I liked best were these few quoted sentences.
"The Clovertown dual-socket, 2x4-core workstation worked well rendering the Cinema4D"
That looks like an Octo PowerMac.
"Clovertown (as well as its desktop cousin Kentsfield) as early as this Christmas..."
That looks like it could be ready for MWSF in January.
"Now, the Intel show had a HP dual-socket Woodcrest at 3 GHz / FSB 1333 vs a Sun dual-socket Opteron 285 2.6 GHz / 1 GHz HT, both with 2 GB RAM etc, running SunGard credit analysis application. The Woodcrest completed the job 35% faster, while the power meter also showed roughly 6% less power consumed - 307 W vs 325 W.?"
"So, my feel is: Intel wins the 1-socket and the 2-socket soccer match this year - next year, anything can happen..."
Well we all new the woodcrest would kick ass. Which is what I have been trying to tell all the people that said Apple should have used AMD.
I haven't read it (but I will) Probably because Conroe isn't a dual socket processor. The Opteron, and Woodcrest are. Or they are just testing the highest spec'd processors from both vendors. OK, I'll go read it.
[edit] OK, read it. It's all about server workstation processors really. The part I liked best were these few quoted sentences.
"The Clovertown dual-socket, 2x4-core workstation worked well rendering the Cinema4D"
That looks like an Octo PowerMac.
"Clovertown (as well as its desktop cousin Kentsfield) as early as this Christmas..."
That looks like it could be ready for MWSF in January.
"Now, the Intel show had a HP dual-socket Woodcrest at 3 GHz / FSB 1333 vs a Sun dual-socket Opteron 285 2.6 GHz / 1 GHz HT, both with 2 GB RAM etc, running SunGard credit analysis application. The Woodcrest completed the job 35% faster, while the power meter also showed roughly 6% less power consumed - 307 W vs 325 W.?"
"So, my feel is: Intel wins the 1-socket and the 2-socket soccer match this year - next year, anything can happen..."
Well we all new the woodcrest would kick ass. Which is what I have been trying to tell all the people that said Apple should have used AMD.
People who thought that Intel was down for the count weren't really undersding the amount of R&D Intel can bring to bear. Once Intel understood that the Prescott design wasn't leading anywhere they turned their aircraft carrier around pretty quickly.
People who thought that Intel was down for the count weren't really undersding the amount of R&D Intel can bring to bear. Once Intel understood that the Prescott design wasn't leading anywhere they turned their aircraft carrier around pretty quickly.
Hell that Israeli team they have must but the companies golden boys/girls. They've laid golden eggs with most of their projects (Pentium M)
AMD isn't going to lay down and raise the white flag but they'll have to compete very hard now that Intel is bringing out the products that killed Tejas.
The prospect of 8 cores in a future Powermac cause me to salivate like Pavlov's dogs. Then you toss in stuff like GPUs with h.264 acceleration and huge framebuffers and I have no complaints other than I'd like to see 10G ethernet at an affordable price<smile>
Hell that Israeli team they have must but the companies golden boys/girls. They've laid golden eggs with most of their projects (Pentium M)
AMD isn't going to lay down and raise the white flag but they'll have to compete very hard now that Intel is bringing out the products that killed Tejas.
The prospect of 8 cores in a future Powermac cause me to salivate like Pavlov's dogs. Then you toss in stuff like GPUs with h.264 acceleration and huge framebuffers and I have no complaints other than I'd like to see 10G ethernet at an affordable price<smile>
Something else interesting as well. Parallels software works well because of the virtualization built into Intel's new chips. But graphics performance sucks.
However, both ATI and Nvidia are working on GPU virtualization. The estimate is that either, or both, will have it out anywhere from end of 2007 to beginning of 2008. while performance won't equal what the chip can do with just one system, it could give 75% of its performance to the virtualized OS.
I did read something like that today. If Nvidia and ATI can get GPU virtualization down by end of 2007 that's great. By then Intel and AMD should be moving virtualization to not only the CPU but PCI-Express cards and other I/O.
I imagine that as soon as 2010 we'll look back and laugh about how archaic 2006 really was with Boot Camp dual booting and Parallels just getting started.
The computing Power User at that time will likely have at least 3 OS running simultaneously at all times. Fibre Broadband connections, 8 cores of processing and Terabytes of storage likely on the network.
Today we had a laugh about the old Trash 80s and Commodore 64s at work today. I guess someday we'll look at today's Powermacs with the same nostalgia.
I did read something like that today. If Nvidia and ATI can get GPU virtualization down by end of 2007 that's great. By then Intel and AMD should be moving virtualization to not only the CPU but PCI-Express cards and other I/O.
I imagine that as soon as 2010 we'll look back and laugh about how archaic 2006 really was with Boot Camp dual booting and Parallels just getting started.
The computing Power User at that time will likely have at least 3 OS running simultaneously at all times. Fibre Broadband connections, 8 cores of processing and Terabytes of storage likely on the network.
Today we had a laugh about the old Trash 80s and Commodore 64s at work today. I guess someday we'll look at today's Powermacs with the same nostalgia.
I can still remember my heart pounding from the Altair 8080. While I never bought one (damn!), I do still have the literature, as well as my old Atari 800, my wifes 400, my ST's, etc.
Yeah, those were the days. It really was exciting back then.
But, in a few years, we'll likely think the same about now.
I can still remember my heart pounding from the Altair 8080. While I never bought one (damn!), I do still have the literature, as well as my old Atari 800, my wifes 400, my ST's, etc.
Yeah, those were the days. It really was exciting back then.
But, in a few years, we'll likely think the same about now.
Back then it was exciting because there was stuffing coming out that could do things that simply no consumer product could do before. It was bleeding edge, exciting stuff and people were putting out software that nobody had seen before. The last 15 years or so can really just be categorized as "more". More memory, more pixels, more megahertz, more disk space, more (of the same kind of) software.
Right now we are on the cusp of a new kind of transition. Multi-processors have been in the dual range for a while, but it hasn't been a major change since there is usually 2 things in existing software to be done at the same time, even if the OS is "the other one". The first commodity 4-8 way systems are about to appear and that will encourage the software developers to start thinking outside their simple single-processor box. By 2010 we're going to be seeing far more cores far more commonly... I've seen predictions of 20-50 cores in a system, depending on how simple or complex each one is (imagine the 9 processor Cell moved to a process 4x smaller). The developers who haven't adapted will see no real change to what they can do. The developers who adapted and figure out how to leverage the potential will deliver stuff that won't just make you salivate... sphincter control will be in jeapordy. Exciting times are ahead.
Back then it was exciting because there was stuffing coming out that could do things that simply no consumer product could do before. It was bleeding edge, exciting stuff and people were putting out software that nobody had seen before. The last 15 years or so can really just be categorized as "more". More memory, more pixels, more megahertz, more disk space, more (of the same kind of) software.
Right now we are on the cusp of a new kind of transition. Multi-processors have been in the dual range for a while, but it hasn't been a major change since there is usually 2 things in existing software to be done at the same time, even if the OS is "the other one". The first commodity 4-8 way systems are about to appear and that will encourage the software developers to start thinking outside their simple single-processor box. By 2010 we're going to be seeing far more cores far more commonly... I've seen predictions of 20-50 cores in a system, depending on how simple or complex each one is (imagine the 9 processor Cell moved to a process 4x smaller). The developers who haven't adapted will see no real change to what they can do. The developers who adapted and figure out how to leverage the potential will deliver stuff that won't just make you salivate... sphincter control will be in jeapordy. Exciting times are ahead.
That's certainly true. But, you know, this is an emotional response, not a logical progression. New generations of people see what they see, and not what we've seen in the past.
Wherever you get in on the ride is exciting. I'm excited about the future.
My thought's have always been that no matter where we are, no matter how advanced we think we are, we are always on the primitive end of the curve, always at the beginning. 100 years from now they will be laughing at just how backward we are. 100 years after that, they will be having the same laugh about what was being used in 2100. And on and on. Of course, that's if people are still people as we know ourselves to be, and are thus still capable of laughing at all.
And that's assuming that civilization will be around for some time after we're gone.
While I may offer anecdotal evidence my anecdotes are supported by the sales figures(which I shall not divulge) for a Top 5 reseller thus they carry a wee bit more weight that the typical mac fans conjecture about what people want.
point taken.
Quote:
Originally posted by hmurchison
I think it's clear that,if given the choice between having more computing power for your dollar versus less, most of us would choose the former.
Not relevant in discussing Apple's options to fill a void in their line-up.
Quote:
Originally posted by hmurchison
However Apple has decided on their current structure and frankly we vote with our wallets. If they were doing the wrong things sales would be declining rapidly.
Yes, but in the last 2 years they are on record that market share matters.
Few people thought they would introduce an inexpensive computer like the mini Mac. Very few thought they would ever switch to Intel and even fewer ever dreamt they would introduce software like Bootcamp. Times change, so does Apple.
Quote:
Originally posted by hmurchison
I have no faith in the notion that a somewhat expandable headless Mac would increase marketshare at all. This is based on my own experiences and others are free to disagree.
I disagree, but respect your opinion.
and I guess I should stop posting on this subject in this thread, kind of a thread hijack and I apologize.
I haven't seen this mentioned(here or in "current hardware" so I'll link to the article site), but AnandTech has a good review of the MacBook Pro and in their conclusion kind of recommended that if you could wait for Memrom you might consider waiting. They seem impressed with coming Intel offerings. Still, good to see such an overall good review about the current MacBook Pro from what was previously an all Windows site.
True, but to increase market share where does Apple have to pull from?
Quote:
I have no faith in the notion that a somewhat expandable headless Mac would increase marketshare at all. This is based on my own experiences and others are free to disagree.
No of course this computer would not raise Apple's market share all by itself. But yes it would play a part in the team.
The Mac mini is not the hottest selling computer for Apple, of course Apple does not really want it to be, but the mini plays its part in bringing new people to the Mac.
Quote:
The prospect of 8 cores in a future Powermac cause me to salivate like Pavlov's dogs.
The prospect of this actually reinforces the need for Apple to produce a small cheaper tower.
An extremely small segment of the user base could afford or even need an 8 core PowerMac.
At this point 4 cores is complete overkill for most users.
As far as we know, the iMac is Apple's best seller. The Mac mini is apparently not a hot seller, but as said above, brings new customers. The PowerMac is the ''butter'', Apple may not sell a lot of them (because of the price?) but the margin on each is big.
I don't see why introducing a new design would not bring more ''average'' customers, those who are not used to AIOs and who could see the Mini more as a ''toy'' (although I am NOT thinking that) and who think that the current PowerMacs are overpriced. There are lots of offerings in the $999-$1499 range of PC desktops and different people want different screens: small, big, standard or widescreen, with speakers or without, analog and/or digital...
I know a lot of people that find the iMac beautiful, but wouldn't want one for themselves, go figure!
Like someone wrote above a dual-woodcrest computer has a non-negligeable power consumption while a single (dual-core) Conroe is rated at about 65W.
So why not keep the current PowerMac enclosure for the dual-woodcrest models (with good cooling and some modifications to support, lets say 6 slots, in a 16-8-2-2-2-2 configuration, and 4 HD spaces etc...), and design a new one for Conroe-based PowerMacs (standard cooling, 2 slots, 2 HD spaces...)?
Offering two configurations for each model wouldn't hurt, and with the price of the Conroe CPUs compared to the Woodcrest ones, I think Apple can release another good margin computer line. Example:
As far as we know, the iMac is Apple's best seller. The Mac mini is apparently not a hot seller, but as said above, brings new customers. The PowerMac is the ''butter'', Apple may not sell a lot of them (because of the price?) but the margin on each is big.
I don't see why introducing a new design would not bring more ''average'' customers, those who are not used to AIOs and who could see the Mini more as a ''toy'' (although I am NOT thinking that) and who think that the current PowerMacs are overpriced. There are lots of offerings in the $999-$1499 range of PC desktops and different people want different screens: small, big, standard or widescreen, with speakers or without, analog and/or digital...
I know a lot of people that find the iMac beautiful, but wouldn't want one for themselves, go figure!
Like someone wrote above a dual-woodcrest computer has a non-negligeable power consumption while a single (dual-core) Conroe is rated at about 65W.
So why not keep the current PowerMac enclosure for the dual-woodcrest models (with good cooling and some modifications to support, lets say 6 slots, in a 16-8-2-2-2-2 configuration, and 4 HD spaces etc...), and design a new one for Conroe-based PowerMacs (standard cooling, 2 slots, 2 HD spaces...)?
Offering two configurations for each model wouldn't hurt, and with the price of the Conroe CPUs compared to the Woodcrest ones, I think Apple can release another good margin computer line. Example:
$2499 Woodcrest (quad) 2.33GHz + more RAM, even better GPU
$3299 Woodcrest (quad) 3.00GHz (+/- same config)
...
This also leaves room for a Conroe Extreme Edition model (X-Mas 2006 or MWSF 2007) at $1999!!!
Every model will have some sales, and bring in more customers. Dell has many models, and they all sell, to a certain extent.
So, if Apple had ten more models, they would certainly pick up more sales. The question is whether these models would sell in great enough numbers to be profitable. Dell can afford to play around, Apple can't. If Dell doesn't sell much of a couple of models, and discontinues them, no one hears about it, or cares. But when Apple makes a model that fails, it's spoken about for years afterwards. That's a danger for Apple. In their position, publicity like that can scare away potential customers.
Insofar as power usage for Woodcrest goes, it's 85 watts for the 3GHz version. The low voltage models use less, but I don't know the number offhand.
I don't remember the exact number, but about a year, or so ago when I thought the powerMac line was not selling at all because it was:
Old, and not been updated in forever.
The last of the powerMacs before the new G6, and a huge PM update that everyone was waiting for.
Long overdue for an update. (moving to 3GHz IBM)
Apple had a conference call, and in it they said they had sold over 400,000 PowerMac units in that quarter.
The ones that were not selling was the iMac, and PowerBook which both had been updated before the last powerMac update at that time.
So saying the PowerMac is a not as big a seller than the iMac for Apple may be just a what you think, because you really have no idea about pro users, or what kind of sales apple has with pro users.
I was suprised at the numbers. I thought that everybody was pretty much waiting on the revision. I know a lot of people that were. Which is actually what I'm doing right now. I want an intel Dual Woodcrest machine with a Quadro FX (if not two of them).
Comments
Originally posted by tubgirl
support what, six slots?
the new intel-powermacs will probably (?) use a chipset with the ich8 southbridge.
i dont know much about that one, but ich7 found in for example the 975x chipset got support for one 16x-lane pci-e or two 8x-lane, six 1x-lane pci-e and six 'regular' pci slots.
(and you'll get a free tpm too, no extra cost!)
i dont think the ich8 will be much worse...
edit: now (after writing all that) i think i get what you're asking.
i guess you could 'route' some of the pci-e lanes in some kind of extension cable to a break-out slot-box, but i dont think it'll be pretty...
offering two sizes to begin with would probably save apple a lot of headache...
Of course, properly, Apple will never offer a machine that includes the old PCI specification.
The current machine is very highly spec'd. It has only one 16 lane slot, but unlike other Express machines, it also has an 8 lane slot, and two 4 lane slots That's pretty good.
If they did offer a 6 slot machine, the question is how they would allocate them. Would they offer 2 board SLI or Crossfire? Will the continue to offer 8 and 4 lane slots.
Will they even offer SLI and Crossfire on another 4 slot machine?
One reason why more slots is important is because all of the high end graphics cards take up the space of the next slot as well. So a 4 slot machine efectively becomes a 3 slot one. If you are using either SLI or CF with two high end boards, you might find that you have NO slots left.
That's where the question of 6 slots comes into play. If you need a firewire/USB board, there might be no where to put one.
A 2 slot machine would become a 1 slot machine with a high end board.
Originally posted by hmurchison
This subject comes up on Mac sites because Mac sites are filled with computer enthusiasts.
I'm not a computer enthusiast and I don't know if this statement can be proved.
Originally posted by hmurchison
A low cost expandable headless Mac isn't going to increase Marketshare. Apple has copious amounts of data and that data is likely going to show that no matter what you do for a desktop it's not going to affect much sales change because PEOPLE WANT LAPTOPS.
I haven't a clue how much market research Apple has done and this can not be proved. However, I have read on the internet(re: veracity to be questioned here, we all know how accurate the internet is) numerous times that Apple performs a minimal amount of market research.
But we do know that there are/were AIO Windows computers offered and virtually no one bought them. This alone is indirect evidence that the majority of computer buyers, for whatever reasons, valid or not, do not want AIO computers.
Originally posted by hmurchison
The people that want desktops generally have an ideal selection of specs they want. If an AIO meets those specs then they will generally have no apprehension towards buying it. Once I told potential iMac purchasers that they weren't locked out of better monitor choices they were more at ease.
Anecdotal evidence, while interesting and important won't convince me that the largest segment of computer buyers want or would buy an AIO.
Originally posted by hmurchison
I really think the days of a big roomy box with loud fans is coming to an end. Workstations will always exist but then again people who need workstations aren't really complaining vociferously about price.
I, at least, am not talking about " a big roomy box with loud fans".
Originally posted by melgross
Don't forget that Apple had a single cpu tower G5 for $1,500. Just making the changes I suggested would have brought that down to $999. This didn't have anything to do with the iMac. And, it would have kept a shorter aluminum case. I did some costing at the time
Yes they did. However, as you pointed out, Apple neutered it. I was the same processer, slower bus, limited ram capable and PCI instead of PCI-X as the iMac and for the exact same cost of the iMac that came with a 17" LCD screen. If Apple had made it a dual processor, with the faster bus, and PCI-X and same ram capabilities as the other towers it might have sold better. But who in their right mind would pay for this machine when they could either buy the iMac or spend the extra money for a dual processor machine?
I'm not saying that Apple didn't make the right choices in the past. Maybe they needed to protect iMac sales or didn't want to sell the low end tower and force people to decide between the iMac and the dual processor tower.
Seriously though, if Apple truly wants to increase market share and now we have them on record twice saying that increasing market share is a goal, then it is IMHO that they will have to add a computer to their line up that targets the arguably the largest segment of computer buyers.
edit cleaned up some grammer, if I left more they ain't getting fixed.
While I may offer anecdotal evidence my anecdotes are supported by the sales figures(which I shall not divulge) for a Top 5 reseller thus they carry a wee bit more weight that the typical mac fans conjecture about what people want.
I think it's clear that,if given the choice between having more computing power for your dollar versus less, most of us would choose the former.
However Apple has decided on their current structure and frankly we vote with our wallets. If they were doing the wrong things sales would be declining rapidly.
I have no faith in the notion that a somewhat expandable headless Mac would increase marketshare at all. This is based on my own experiences and others are free to disagree.
Originally posted by hmurchison
I have no faith in the notion that a somewhat expandable headless Mac would increase marketshare at all. This is based on my own experiences and others are free to disagree.
I totally agree with that.
http://www.theinquirer.net/?article=30963
Originally posted by melgross
Back to the Woodcrest/Opteron contest for a moment (notice that Conroe isn't considered in this matchup).
http://www.theinquirer.net/?article=30963
I haven't read it (but I will) Probably because Conroe isn't a dual socket processor. The Opteron, and Woodcrest are. Or they are just testing the highest spec'd processors from both vendors. OK, I'll go read it.
[edit] OK, read it. It's all about server workstation processors really. The part I liked best were these few quoted sentences.
"The Clovertown dual-socket, 2x4-core workstation worked well rendering the Cinema4D"
That looks like an Octo PowerMac.
"Clovertown (as well as its desktop cousin Kentsfield) as early as this Christmas..."
That looks like it could be ready for MWSF in January.
"Now, the Intel show had a HP dual-socket Woodcrest at 3 GHz / FSB 1333 vs a Sun dual-socket Opteron 285 2.6 GHz / 1 GHz HT, both with 2 GB RAM etc, running SunGard credit analysis application. The Woodcrest completed the job 35% faster, while the power meter also showed roughly 6% less power consumed - 307 W vs 325 W.?"
"So, my feel is: Intel wins the 1-socket and the 2-socket soccer match this year - next year, anything can happen..."
Well we all new the woodcrest would kick ass. Which is what I have been trying to tell all the people that said Apple should have used AMD.
Originally posted by onlooker
I haven't read it (but I will) Probably because Conroe isn't a dual socket processor. The Opteron, and Woodcrest are. Or they are just testing the highest spec'd processors from both vendors. OK, I'll go read it.
[edit] OK, read it. It's all about server workstation processors really. The part I liked best were these few quoted sentences.
"The Clovertown dual-socket, 2x4-core workstation worked well rendering the Cinema4D"
That looks like an Octo PowerMac.
"Clovertown (as well as its desktop cousin Kentsfield) as early as this Christmas..."
That looks like it could be ready for MWSF in January.
"Now, the Intel show had a HP dual-socket Woodcrest at 3 GHz / FSB 1333 vs a Sun dual-socket Opteron 285 2.6 GHz / 1 GHz HT, both with 2 GB RAM etc, running SunGard credit analysis application. The Woodcrest completed the job 35% faster, while the power meter also showed roughly 6% less power consumed - 307 W vs 325 W.?"
"So, my feel is: Intel wins the 1-socket and the 2-socket soccer match this year - next year, anything can happen..."
Well we all new the woodcrest would kick ass. Which is what I have been trying to tell all the people that said Apple should have used AMD.
People who thought that Intel was down for the count weren't really undersding the amount of R&D Intel can bring to bear. Once Intel understood that the Prescott design wasn't leading anywhere they turned their aircraft carrier around pretty quickly.
Originally posted by melgross
People who thought that Intel was down for the count weren't really undersding the amount of R&D Intel can bring to bear. Once Intel understood that the Prescott design wasn't leading anywhere they turned their aircraft carrier around pretty quickly.
Hell that Israeli team they have must but the companies golden boys/girls. They've laid golden eggs with most of their projects (Pentium M)
AMD isn't going to lay down and raise the white flag but they'll have to compete very hard now that Intel is bringing out the products that killed Tejas.
The prospect of 8 cores in a future Powermac cause me to salivate like Pavlov's dogs. Then you toss in stuff like GPUs with h.264 acceleration and huge framebuffers and I have no complaints other than I'd like to see 10G ethernet at an affordable price<smile>
Originally posted by hmurchison
Hell that Israeli team they have must but the companies golden boys/girls. They've laid golden eggs with most of their projects (Pentium M)
AMD isn't going to lay down and raise the white flag but they'll have to compete very hard now that Intel is bringing out the products that killed Tejas.
The prospect of 8 cores in a future Powermac cause me to salivate like Pavlov's dogs. Then you toss in stuff like GPUs with h.264 acceleration and huge framebuffers and I have no complaints other than I'd like to see 10G ethernet at an affordable price<smile>
Something else interesting as well. Parallels software works well because of the virtualization built into Intel's new chips. But graphics performance sucks.
However, both ATI and Nvidia are working on GPU virtualization. The estimate is that either, or both, will have it out anywhere from end of 2007 to beginning of 2008. while performance won't equal what the chip can do with just one system, it could give 75% of its performance to the virtualized OS.
That will be a humdinger.
I did read something like that today. If Nvidia and ATI can get GPU virtualization down by end of 2007 that's great. By then Intel and AMD should be moving virtualization to not only the CPU but PCI-Express cards and other I/O.
I imagine that as soon as 2010 we'll look back and laugh about how archaic 2006 really was with Boot Camp dual booting and Parallels just getting started.
The computing Power User at that time will likely have at least 3 OS running simultaneously at all times. Fibre Broadband connections, 8 cores of processing and Terabytes of storage likely on the network.
Today we had a laugh about the old Trash 80s and Commodore 64s at work today. I guess someday we'll look at today's Powermacs with the same nostalgia.
Originally posted by hmurchison
Melgross
I did read something like that today. If Nvidia and ATI can get GPU virtualization down by end of 2007 that's great. By then Intel and AMD should be moving virtualization to not only the CPU but PCI-Express cards and other I/O.
I imagine that as soon as 2010 we'll look back and laugh about how archaic 2006 really was with Boot Camp dual booting and Parallels just getting started.
The computing Power User at that time will likely have at least 3 OS running simultaneously at all times. Fibre Broadband connections, 8 cores of processing and Terabytes of storage likely on the network.
Today we had a laugh about the old Trash 80s and Commodore 64s at work today. I guess someday we'll look at today's Powermacs with the same nostalgia.
I can still remember my heart pounding from the Altair 8080. While I never bought one (damn!), I do still have the literature, as well as my old Atari 800, my wifes 400, my ST's, etc.
Yeah, those were the days. It really was exciting back then.
But, in a few years, we'll likely think the same about now.
Originally posted by melgross
I can still remember my heart pounding from the Altair 8080. While I never bought one (damn!), I do still have the literature, as well as my old Atari 800, my wifes 400, my ST's, etc.
Yeah, those were the days. It really was exciting back then.
But, in a few years, we'll likely think the same about now.
Back then it was exciting because there was stuffing coming out that could do things that simply no consumer product could do before. It was bleeding edge, exciting stuff and people were putting out software that nobody had seen before. The last 15 years or so can really just be categorized as "more". More memory, more pixels, more megahertz, more disk space, more (of the same kind of) software.
Right now we are on the cusp of a new kind of transition. Multi-processors have been in the dual range for a while, but it hasn't been a major change since there is usually 2 things in existing software to be done at the same time, even if the OS is "the other one". The first commodity 4-8 way systems are about to appear and that will encourage the software developers to start thinking outside their simple single-processor box. By 2010 we're going to be seeing far more cores far more commonly... I've seen predictions of 20-50 cores in a system, depending on how simple or complex each one is (imagine the 9 processor Cell moved to a process 4x smaller). The developers who haven't adapted will see no real change to what they can do. The developers who adapted and figure out how to leverage the potential will deliver stuff that won't just make you salivate... sphincter control will be in jeapordy. Exciting times are ahead.
Originally posted by Programmer
Back then it was exciting because there was stuffing coming out that could do things that simply no consumer product could do before. It was bleeding edge, exciting stuff and people were putting out software that nobody had seen before. The last 15 years or so can really just be categorized as "more". More memory, more pixels, more megahertz, more disk space, more (of the same kind of) software.
Right now we are on the cusp of a new kind of transition. Multi-processors have been in the dual range for a while, but it hasn't been a major change since there is usually 2 things in existing software to be done at the same time, even if the OS is "the other one". The first commodity 4-8 way systems are about to appear and that will encourage the software developers to start thinking outside their simple single-processor box. By 2010 we're going to be seeing far more cores far more commonly... I've seen predictions of 20-50 cores in a system, depending on how simple or complex each one is (imagine the 9 processor Cell moved to a process 4x smaller). The developers who haven't adapted will see no real change to what they can do. The developers who adapted and figure out how to leverage the potential will deliver stuff that won't just make you salivate... sphincter control will be in jeapordy. Exciting times are ahead.
That's certainly true. But, you know, this is an emotional response, not a logical progression. New generations of people see what they see, and not what we've seen in the past.
Wherever you get in on the ride is exciting. I'm excited about the future.
My thought's have always been that no matter where we are, no matter how advanced we think we are, we are always on the primitive end of the curve, always at the beginning. 100 years from now they will be laughing at just how backward we are. 100 years after that, they will be having the same laugh about what was being used in 2100. And on and on. Of course, that's if people are still people as we know ourselves to be, and are thus still capable of laughing at all.
And that's assuming that civilization will be around for some time after we're gone.
Originally posted by hmurchison
PC users and Mac users are different.
While I may offer anecdotal evidence my anecdotes are supported by the sales figures(which I shall not divulge) for a Top 5 reseller thus they carry a wee bit more weight that the typical mac fans conjecture about what people want.
point taken.
Originally posted by hmurchison
I think it's clear that,if given the choice between having more computing power for your dollar versus less, most of us would choose the former.
Not relevant in discussing Apple's options to fill a void in their line-up.
Originally posted by hmurchison
However Apple has decided on their current structure and frankly we vote with our wallets. If they were doing the wrong things sales would be declining rapidly.
Yes, but in the last 2 years they are on record that market share matters.
Few people thought they would introduce an inexpensive computer like the mini Mac. Very few thought they would ever switch to Intel and even fewer ever dreamt they would introduce software like Bootcamp. Times change, so does Apple.
Originally posted by hmurchison
I have no faith in the notion that a somewhat expandable headless Mac would increase marketshare at all. This is based on my own experiences and others are free to disagree.
I disagree, but respect your opinion.
and I guess I should stop posting on this subject in this thread, kind of a thread hijack and I apologize.
I haven't seen this mentioned(here or in "current hardware" so I'll link to the article site), but AnandTech has a good review of the MacBook Pro and in their conclusion kind of recommended that if you could wait for Memrom you might consider waiting. They seem impressed with coming Intel offerings. Still, good to see such an overall good review about the current MacBook Pro from what was previously an all Windows site.
(and I was being generous. Two months ago I would have paid $1800 for G5; today I wouldn?t even pay $500.)
PC users and Mac users are different.
True, but to increase market share where does Apple have to pull from?
I have no faith in the notion that a somewhat expandable headless Mac would increase marketshare at all. This is based on my own experiences and others are free to disagree.
No of course this computer would not raise Apple's market share all by itself. But yes it would play a part in the team.
The Mac mini is not the hottest selling computer for Apple, of course Apple does not really want it to be, but the mini plays its part in bringing new people to the Mac.
The prospect of 8 cores in a future Powermac cause me to salivate like Pavlov's dogs.
The prospect of this actually reinforces the need for Apple to produce a small cheaper tower.
An extremely small segment of the user base could afford or even need an 8 core PowerMac.
At this point 4 cores is complete overkill for most users.
Originally posted by TenoBell
At this point 4 cores is complete overkill for most users.
Yes... but we're only 1 killer app away from it being essential.
I don't see why introducing a new design would not bring more ''average'' customers, those who are not used to AIOs and who could see the Mini more as a ''toy'' (although I am NOT thinking that) and who think that the current PowerMacs are overpriced. There are lots of offerings in the $999-$1499 range of PC desktops and different people want different screens: small, big, standard or widescreen, with speakers or without, analog and/or digital...
I know a lot of people that find the iMac beautiful, but wouldn't want one for themselves, go figure!
Like someone wrote above a dual-woodcrest computer has a non-negligeable power consumption while a single (dual-core) Conroe is rated at about 65W.
So why not keep the current PowerMac enclosure for the dual-woodcrest models (with good cooling and some modifications to support, lets say 6 slots, in a 16-8-2-2-2-2 configuration, and 4 HD spaces etc...), and design a new one for Conroe-based PowerMacs (standard cooling, 2 slots, 2 HD spaces...)?
Offering two configurations for each model wouldn't hurt, and with the price of the Conroe CPUs compared to the Woodcrest ones, I think Apple can release another good margin computer line. Example:
$1099 Conroe dual-core 2.13GHz, 512MB, 160HD...
$1499 Conroe dual-core 2.67GHz + bigger HD, better GPU
$2499 Woodcrest (quad) 2.33GHz + more RAM, even better GPU
$3299 Woodcrest (quad) 3.00GHz (+/- same config)
...
This also leaves room for a Conroe Extreme Edition model (X-Mas 2006 or MWSF 2007) at $1999!!!
Originally posted by mjteix
As far as we know, the iMac is Apple's best seller. The Mac mini is apparently not a hot seller, but as said above, brings new customers. The PowerMac is the ''butter'', Apple may not sell a lot of them (because of the price?) but the margin on each is big.
I don't see why introducing a new design would not bring more ''average'' customers, those who are not used to AIOs and who could see the Mini more as a ''toy'' (although I am NOT thinking that) and who think that the current PowerMacs are overpriced. There are lots of offerings in the $999-$1499 range of PC desktops and different people want different screens: small, big, standard or widescreen, with speakers or without, analog and/or digital...
I know a lot of people that find the iMac beautiful, but wouldn't want one for themselves, go figure!
Like someone wrote above a dual-woodcrest computer has a non-negligeable power consumption while a single (dual-core) Conroe is rated at about 65W.
So why not keep the current PowerMac enclosure for the dual-woodcrest models (with good cooling and some modifications to support, lets say 6 slots, in a 16-8-2-2-2-2 configuration, and 4 HD spaces etc...), and design a new one for Conroe-based PowerMacs (standard cooling, 2 slots, 2 HD spaces...)?
Offering two configurations for each model wouldn't hurt, and with the price of the Conroe CPUs compared to the Woodcrest ones, I think Apple can release another good margin computer line. Example:
$1099 Conroe dual-core 2.13GHz, 512MB, 160HD...
$1499 Conroe dual-core 2.67GHz + bigger HD, better GPU
$2499 Woodcrest (quad) 2.33GHz + more RAM, even better GPU
$3299 Woodcrest (quad) 3.00GHz (+/- same config)
...
This also leaves room for a Conroe Extreme Edition model (X-Mas 2006 or MWSF 2007) at $1999!!!
Every model will have some sales, and bring in more customers. Dell has many models, and they all sell, to a certain extent.
So, if Apple had ten more models, they would certainly pick up more sales. The question is whether these models would sell in great enough numbers to be profitable. Dell can afford to play around, Apple can't. If Dell doesn't sell much of a couple of models, and discontinues them, no one hears about it, or cares. But when Apple makes a model that fails, it's spoken about for years afterwards. That's a danger for Apple. In their position, publicity like that can scare away potential customers.
Insofar as power usage for Woodcrest goes, it's 85 watts for the 3GHz version. The low voltage models use less, but I don't know the number offhand.
- Old, and not been updated in forever.
- The last of the powerMacs before the new G6, and a huge PM update that everyone was waiting for.
- Long overdue for an update. (moving to 3GHz IBM)
Apple had a conference call, and in it they said they had sold over 400,000 PowerMac units in that quarter.The ones that were not selling was the iMac, and PowerBook which both had been updated before the last powerMac update at that time.
So saying the PowerMac is a not as big a seller than the iMac for Apple may be just a what you think, because you really have no idea about pro users, or what kind of sales apple has with pro users.
I was suprised at the numbers. I thought that everybody was pretty much waiting on the revision. I know a lot of people that were. Which is actually what I'm doing right now. I want an intel Dual Woodcrest machine with a Quadro FX (if not two of them).