(I feel with SLI graphic cards and dual core cpus...that the software to take advantage of it is behind the curve. Same with PCI Express...and that level of 'mainstream' optimisation for technologies that aren't mainstream yet could take another couple of years...)
yeah, i'm sick of hearing IBM "might" do this and "might" do that while apple sees its powermac g5 sales go down the toilet.
bon bon - re: SLI, 3dMark and doom3 and chronicles of riddick shows awesome improvements with SLI but ut2004 and halflife2 do not show much improvement. so it's a question of looking at game developer support in the next 6-12months to see how this moves.
re: PCI express vs AGP, fair enough point. i got a pciexpress sli-ready board with one nvidia 6600gt in it, felt more 'upgradable' to me in the long run. maybe 6 months time i might put in an SLI'ed 6800gt pair for us$200 and i know its gonna scream.
yeah, i'm sick of hearing IBM "might" do this and "might" do that while apple sees its powermac g5 sales go down the toilet.
bon bon - re: SLI, 3dMark and doom3 and chronicles of riddick shows awesome improvements with SLI but ut2004 and halflife2 do not show much improvement. so it's a question of looking at game developer support in the next 6-12months to see how this moves.
re: PCI express vs AGP, fair enough point. i got a pciexpress sli-ready board with one nvidia 6600gt in it, felt more 'upgradable' to me in the long run. maybe 6 months time i might put in an SLI'ed 6800gt pair for us$200 and i know its gonna scream.
I'm getting pretty tired of 'big shot' IBM's promises. Two years and no 3 gig cpu. After making Steve look like an idiot twice no wonder he went Intel. And no G5 low power for the 'Book. And when it does turn up, like the MP, it's late. Meantime, Apple's PowerMac sales begin hitting the 'toilet', something that started with the '500mhz Debacle' under Motorola. The promise of PowerPC? Alot of hot air and broken promise. A great leap followed by years of stasis. And it's right back at square one. After ten years of that and pulling what's left of Steve's hair out of his head or telling Moto execs ('Get the f*ck out of my office!') he finally pulled the plug on this nigh on Still born platform. IBM are all talk. They didn't develop the Cell on their own either. They thought they could stiff Steve...but after the 500mhz debacle, Apple had learned their lesson and the Mac OS 'X' was in place to give IBM teh finger if they did a 'Motorola', 'Just in Case...' Sorry, rant over. I guess I just feel let down after all the initial promise of the WWDC a few years back. Another false dawn on the CPU front. That is set to change now the world's biggest cpu manufacturer has cpus on the desktop and laptop as a priority. Not only that, if you look at the structure of Intel (overall) you'll being to see a company that fits Apple like glove in terms of where Apple wants to be 10 years from now. Steve was right. It's not about the 3 gig or the G5 'Book. It's about something bigger, way bigger.
Your second point. Yes. Some games show improvement. Plenty don't. That's SLi for you? Will Crossfire from ATI offer broader 'general' improvements? They say so. ie that all past and future games will receive some benefit. But I think the fact remains that this technology isn't mainstream yet. Optimisation is only just getting started. Witness: 'some games'. Do I think it will get better? Yes. It may be a mainstream thing in a couple of years. And 'all' or 'most' games might 'scream' because of it. But it's still early days for the technology. To the point where I think the cost of two cards and a 'SLI' motherboard might as well be 'buy the near the top gpu now' and 'buy another gpu in 6 months time'.
Re: your strategy of getting a SLI motherboard and waiting until 'near the top' GPUS like the GT come down in price to bag a brace to update your 6600Gt rig. Okay. I've thought about doing that myself...
or...
You could buy a SLI board. Still a 7800GT now...and wait for the 7800GT to come down in price.
or...
You save your pennies and wait for the ATI520 to drop. And go with the Crossfire solution. Apparently, the 520 is going to eat the G70 for breakfast ie rumour has it (and the rumoured specs seem to confirm this...) that it's designed to compete with the G80!
We shall see. While CPU development has stalled in recent years and has only just begun the 'wait for the apps to be threaded' Dual Core path the brute force, breakneck and articulate development of GPUs remains fascinating to watch.
As a footnote. As GPU development cycles lengthen...then perhaps SLI or Crossfire can give your rig a 'boost' mid-term through said cycle. SLI and Crossfire remain promising...as does PCI Express and Dual Core CPU and also memory speeds on the rise...and hard drive capacities are on the up and up.
Will I buy a dual core PowerMac? Probably. I can't stand it any longer. I want a Mac.
I'd take a dual-core 2.5 PowerMac, GT6800 or GT7800 ATI520 on PCI-Express. We'll see what apple does with the line up. (No 3 gig dual core...Thinksecret got that one wrong...)
Yes. I think one thing when running a business or being in a business environment, is you learn that when you are let down, to not let it happen again.
Check this thing out on immersion lithography in DECEMBER 2004, about IBM being the 'first' to be able to reach 45nm and 65nm using tools that at the time were making 100+nm chips.
Fast forward one year and a half, AMD has a strong 90nm portfolio and establishing its Athlon lines, Intel is chugging away, with impressive stuff in its centrino/pentium M line while pushing Pentium 4 to its limits. Moto? okay, speed bumps of 1ghz to 1.8 (actually overclocked), but better bus speeds and dual core really nowhere to be seen.
IBM? i guess anything more i would say is covered in your rant
bon bon:
i do have an SLI board now, actually the idea was, more of a reasonable upgrade path. i got a 6600GT 128mb now for us$180 in my country, and so maybe in 6 months i can get the exact same card for maybe us$100 ~ almost like an upgrade which doubles your GPU power for half the price ~ plus i don't have to worry about throwing out or ebaying the existing 6600GT.
my intention when putting my system together was a certain level of scalability built in, eg
IDE-->SATA-->SATA RAID
6600GT x1 --> 6600GT x2 --> ?? 7800GT X 2 ??
512mb x 2 kingston value ram --> dual channel 512mbx2 --> dual channel 512x2 low latency pc4000 ram for overclocking
also socket939 gives me athlon 3000+ --> better athlons --> dual core athlon
i had no idea how software (winxp, games) would realistically perform until i tried it myself, and now i feel i have a handle on how to scale along with software (handling dual-core better) and hardware (upgrading from IDE to SATA NCQ for example)
anyway one thing that's a limit is heat and power management, (it's quite atrocious really what you need for like a SLI 2x6800 Ultra, 2 x SATA raid, northbridge/southbridge + CPU cooling, + PowerSuppy cooling and fans itself....)
...so depending on how things pan out over the year maybe i can get a pentium-M or pentium-M derivative MacIntel by June 2006.
also 'insulting' my hardware with win xp1 is fine for now but beyond a certain level maybe macintelX or more suitable-for-noobs linux distros will be around. or maybe microsoft delivers vista ahead of schedule and its fast, secure and reliable holy shit what am i smoking? strike that last line about vista
I've been pondering this IBM dual-core chips scenario upgrades for Mac and frankly, like many have said so far, there will be performance gains with some apps and little to none with many. My point is the following: with Apple going to Intel starting in 2006, how much benefit will Mac buyers really get in the long run by picking up dual-core or dual dual-core systems when most S/W companies may not bother optimizing their apps to take advantage of 4 cores on PPC when Intels are just around the corner and will be the the longer lived platform processor? When all Macs will be Intel-based in 2007, who's going to bother optimizing anything for the PPC systems past that point and perhaps even as early as next summer? Where's the bigger bang for your dollar now and in the future when you factor in S/W support and potential obsolescence?
Check this thing out on immersion lithography in DECEMBER 2004, about IBM being the 'first' to be able to reach 45nm and 65nm using tools that at the time were making 100+nm chips.
I know I must have missed something in what has been posted here, but from December 2004 to August 2005 isn't "one year and a half", more like 9 months. I'm sooooo confused
Pin-compatibility has nothing to do with Apple using this CPU...
Repeat after me noobs, Apple uses daughtercards for their CPUs...
Come on, macro-nitpicker, we're using shorthand here. Would Apple go to the effort of designing a new daughtercard for the 970MP? With the Intel transition on the horizon? How many units would they sell, anyway? It seems unlikely they'd be willing to design anything at all. You don't pour money into a dead-end product. So, then, pin-compatibility is very important.
Frankly, I think even the overhead of the new SKU's, packaging, shipping etc. aren't going to be worth the small number of sales. Apple will simply let the PowerMac languish as-is until the Intel models come out. Yes, he said "great new PowerPC products in the pipeline", but he was probably referring, if to anything, to the 7448.
Come on, macro-nitpicker, we're using shorthand here. Would Apple go to the effort of designing a new daughtercard for the 970MP? With the Intel transition on the horizon? How many units would they sell, anyway? It seems unlikely they'd be willing to design anything at all. You don't pour money into a dead-end product. So, then, pin-compatibility is very important.
Frankly, I think even the overhead of the new SKU's, packaging, shipping etc. aren't going to be worth the small number of sales. Apple will simply let the PowerMac languish as-is until the Intel models come out. Yes, he said "great new PowerPC products in the pipeline", but he was probably referring, if to anything, to the 7448.
If we take the rumor that Apple has had samples of the 970MP for a few months now, then yes, I would fully expect them to have a redesigned daughtercard available...
And the original post that I replied to truly reads as if the poster thinks Apple still has ZIF sockets on their MLBs...
As for sales, the main market for Dual Duals would be the Pro app users...
Final Cut Pro, Motion, Soundtrack Pro, Shake, Maya, RenderMan, Modo, etc. ...
These are the apps that would love to be cranking on "Double Duals", and most of the folks using these apps in a regular day-to-day profit-making business scenario wuld think nothing of popping for a new machineto pump out work until the first (probably second) versions of the Intel PowerMacs come out...
If we take the rumor that Apple has had samples of the 970MP for a few months now, then yes, I would fully expect them to have a redesigned daughtercard available. . .
I agree. I think there is no doubt at all that Apple has 970MP daughter cards, but has even started manufacture of new PowerMac to build up inventory. If not yet, it will begin very soon. Apple had 970MP samples in 2004 I believe.
. . . You don't pour money into a dead-end product. So, then, pin-compatibility is very important. . .
Why do you think it is a dead end product? Developers who design applications to run on a dual-dual PPC will simply recompile for Intel Macs. No lost work, and a big head start on applications that run faster. This will generate many new PowerMac sales.
Quote:
. . . Frankly, I think even the overhead of the new SKU's, packaging, shipping etc. aren't going to be worth the small number of sales.
You picked some of the lowest cost overhead items in developing a new product. Engineering development cost would be the highest, and it's already done I'm sure.
I agree. I think there is no doubt at all that Apple has 970MP daughter cards, but has even started manufacture of new PowerMac to build up inventory. If not yet, it will begin very soon. Apple had 970MP samples in 2004 I believe.
The reality is that upgrading the Powermacs with these 970MPs, now and after the Intel bomb, may just end up being too late. Businesses may just look aside and wait for the Intel systems before putting down any new money on a product whose processor is EOL and software support will no doubt dwindle significantly. It remains to be seen how things will really play out. IMO, the OS advantages will not help Apple much until the Intel Macs start appearing for sale.
The reality is that upgrading the Powermacs with these 970MPs, now and after the Intel bomb, may just end up being too late. Businesses may just look aside and wait for the Intel systems before putting down any new money on a product whose processor is EOL and software support will no doubt dwindle significantly. . .
Even if your analysis is true for today, Apple did most all the engineering work before the Intel bombshell announcement. Since engineering is the largest piece of new product development, there is little reason for Apple not to go ahead with its original plan. When Apple began working on the next generation of PowerMacs, the Intel move was just an idea that had been around for several years. No one knew for sure what would happen.
The reality is that upgrading the Powermacs with these 970MPs, now and after the Intel bomb, may just end up being too late. Businesses may just look aside and wait for the Intel systems before putting down any new money on a product whose processor is EOL and software support will no doubt dwindle significantly. It remains to be seen how things will really play out. IMO, the OS advantages will not help Apple much until the Intel Macs start appearing for sale.
The thing is there is a lot of life left in the PPC processors as far as software support. We know the transition isn't going to be complete until sometime in 2007, so that is at least two years with PPC support. On top of that look at how long Classic is still kicking around, 5 major OS updates later. So I would say at least 6 years before PPC software becomes hard to find and by then you'll be half way to your second Mactel if you upgrade every four or five years. Also the advantages of proven technology is also there where Rev. A Mactels won't necessarily have perfect hardware or software support.
I know I must have missed something in what has been posted here, but from December 2004 to August 2005 isn't "one year and a half", more like 9 months. I'm sooooo confused
oops.... 9 months it is. my bad
so i think this is something IBM could have shown apple, "look we can do 65nm with our existing 100+nm tools" but perhaps it was too costly/experimental/etc for apple.
The reality is that upgrading the Powermacs with these 970MPs, now and after the Intel bomb, may just end up being too late. Businesses may just look aside and wait for the Intel systems before putting down any new money on a product whose processor is EOL and software support will no doubt dwindle significantly. It remains to be seen how things will really play out. IMO, the OS advantages will not help Apple much until the Intel Macs start appearing for sale.
After having designed enough circuit boards over the years, I can say that a daughter card is not a big deal. Cards are designed with software. A card like that could be turned around in a few days. Even after testing, it wouldn't add up to more than a couple of weeks. All it needs is one or two engineers. The question is about the support chips such as the memory controller. If two duals are being used, there might need to be a redesign there. I'm not sure though. It depends on how it works now, and I haven't investigated it.
Businesses have ordering policies in place. If they order in 18 month intervals, then they will continue to do so. You have it backwards here. They would rather continue to buy PPC machines as long as they can. They don't want to move to a new platform until everything is in place, and they have done months of testing on the new product and software.
Comments
Originally posted by Glamingo
Weres the article, or anything about the macmini getting a G5 in it?
Here it is: http://www.macobserver.com/stockwatc.../08/15.1.shtml
Originally posted by Lemon Bon Bon
I hope they just get on with it.
.............
(I feel with SLI graphic cards and dual core cpus...that the software to take advantage of it is behind the curve. Same with PCI Express...and that level of 'mainstream' optimisation for technologies that aren't mainstream yet could take another couple of years...)
yeah, i'm sick of hearing IBM "might" do this and "might" do that while apple sees its powermac g5 sales go down the toilet.
bon bon - re: SLI, 3dMark and doom3 and chronicles of riddick shows awesome improvements with SLI but ut2004 and halflife2 do not show much improvement. so it's a question of looking at game developer support in the next 6-12months to see how this moves.
re: PCI express vs AGP, fair enough point. i got a pciexpress sli-ready board with one nvidia 6600gt in it, felt more 'upgradable' to me in the long run. maybe 6 months time i might put in an SLI'ed 6800gt pair for us$200 and i know its gonna scream.
yeah, i'm sick of hearing IBM "might" do this and "might" do that while apple sees its powermac g5 sales go down the toilet.
bon bon - re: SLI, 3dMark and doom3 and chronicles of riddick shows awesome improvements with SLI but ut2004 and halflife2 do not show much improvement. so it's a question of looking at game developer support in the next 6-12months to see how this moves.
re: PCI express vs AGP, fair enough point. i got a pciexpress sli-ready board with one nvidia 6600gt in it, felt more 'upgradable' to me in the long run. maybe 6 months time i might put in an SLI'ed 6800gt pair for us$200 and i know its gonna scream.
I'm getting pretty tired of 'big shot' IBM's promises. Two years and no 3 gig cpu. After making Steve look like an idiot twice no wonder he went Intel. And no G5 low power for the 'Book. And when it does turn up, like the MP, it's late. Meantime, Apple's PowerMac sales begin hitting the 'toilet', something that started with the '500mhz Debacle' under Motorola. The promise of PowerPC? Alot of hot air and broken promise. A great leap followed by years of stasis. And it's right back at square one. After ten years of that and pulling what's left of Steve's hair out of his head or telling Moto execs ('Get the f*ck out of my office!') he finally pulled the plug on this nigh on Still born platform. IBM are all talk. They didn't develop the Cell on their own either. They thought they could stiff Steve...but after the 500mhz debacle, Apple had learned their lesson and the Mac OS 'X' was in place to give IBM teh finger if they did a 'Motorola', 'Just in Case...' Sorry, rant over. I guess I just feel let down after all the initial promise of the WWDC a few years back. Another false dawn on the CPU front. That is set to change now the world's biggest cpu manufacturer has cpus on the desktop and laptop as a priority. Not only that, if you look at the structure of Intel (overall) you'll being to see a company that fits Apple like glove in terms of where Apple wants to be 10 years from now. Steve was right. It's not about the 3 gig or the G5 'Book. It's about something bigger, way bigger.
Your second point. Yes. Some games show improvement. Plenty don't. That's SLi for you? Will Crossfire from ATI offer broader 'general' improvements? They say so. ie that all past and future games will receive some benefit. But I think the fact remains that this technology isn't mainstream yet. Optimisation is only just getting started. Witness: 'some games'. Do I think it will get better? Yes. It may be a mainstream thing in a couple of years. And 'all' or 'most' games might 'scream' because of it. But it's still early days for the technology. To the point where I think the cost of two cards and a 'SLI' motherboard might as well be 'buy the near the top gpu now' and 'buy another gpu in 6 months time'.
Re: your strategy of getting a SLI motherboard and waiting until 'near the top' GPUS like the GT come down in price to bag a brace to update your 6600Gt rig. Okay. I've thought about doing that myself...
or...
You could buy a SLI board. Still a 7800GT now...and wait for the 7800GT to come down in price.
or...
You save your pennies and wait for the ATI520 to drop. And go with the Crossfire solution. Apparently, the 520 is going to eat the G70 for breakfast ie rumour has it (and the rumoured specs seem to confirm this...) that it's designed to compete with the G80!
We shall see. While CPU development has stalled in recent years and has only just begun the 'wait for the apps to be threaded' Dual Core path the brute force, breakneck and articulate development of GPUs remains fascinating to watch.
As a footnote. As GPU development cycles lengthen...then perhaps SLI or Crossfire can give your rig a 'boost' mid-term through said cycle. SLI and Crossfire remain promising...as does PCI Express and Dual Core CPU and also memory speeds on the rise...and hard drive capacities are on the up and up.
Will I buy a dual core PowerMac? Probably. I can't stand it any longer. I want a Mac.
I'd take a dual-core 2.5 PowerMac, GT6800 or GT7800 ATI520 on PCI-Express. We'll see what apple does with the line up. (No 3 gig dual core...Thinksecret got that one wrong...)
Lemon Bon Bon
Check this thing out on immersion lithography in DECEMBER 2004, about IBM being the 'first' to be able to reach 45nm and 65nm using tools that at the time were making 100+nm chips.
http://www.geek.com/news/geeknews/20...1203028085.htm
Fast forward one year and a half, AMD has a strong 90nm portfolio and establishing its Athlon lines, Intel is chugging away, with impressive stuff in its centrino/pentium M line while pushing Pentium 4 to its limits. Moto? okay, speed bumps of 1ghz to 1.8 (actually overclocked), but better bus speeds and dual core really nowhere to be seen.
IBM? i guess anything more i would say is covered in your rant
bon bon:
i do have an SLI board now, actually the idea was, more of a reasonable upgrade path. i got a 6600GT 128mb now for us$180 in my country, and so maybe in 6 months i can get the exact same card for maybe us$100 ~ almost like an upgrade which doubles your GPU power for half the price ~ plus i don't have to worry about throwing out or ebaying the existing 6600GT.
my intention when putting my system together was a certain level of scalability built in, eg
IDE-->SATA-->SATA RAID
6600GT x1 --> 6600GT x2 --> ?? 7800GT X 2 ??
512mb x 2 kingston value ram --> dual channel 512mbx2 --> dual channel 512x2 low latency pc4000 ram for overclocking
also socket939 gives me athlon 3000+ --> better athlons --> dual core athlon
i had no idea how software (winxp, games) would realistically perform until i tried it myself, and now i feel i have a handle on how to scale along with software (handling dual-core better) and hardware (upgrading from IDE to SATA NCQ for example)
anyway one thing that's a limit is heat and power management, (it's quite atrocious really what you need for like a SLI 2x6800 Ultra, 2 x SATA raid, northbridge/southbridge + CPU cooling, + PowerSuppy cooling and fans itself....)
...so depending on how things pan out over the year maybe i can get a pentium-M or pentium-M derivative MacIntel by June 2006.
also 'insulting' my hardware with win xp1 is fine for now but beyond a certain level maybe macintelX or more suitable-for-noobs linux distros will be around. or maybe microsoft delivers vista ahead of schedule and its fast, secure and reliable
Originally posted by Lemon Bon Bon
The 'next gen' G5 rumours have been floating around like a Cow in labour.
Umm...am I the only one that doesn't get the metaphor? Cow? float?
Originally posted by sunilraman
...<snip>....
Check this thing out on immersion lithography in DECEMBER 2004, about IBM being the 'first' to be able to reach 45nm and 65nm using tools that at the time were making 100+nm chips.
http://www.geek.com/news/geeknews/20...1203028085.htm
Fast forward one year and a half, ...<snip>...
Not making excuses for IBM mind you, but....
I know I must have missed something in what has been posted here, but from December 2004 to August 2005 isn't "one year and a half", more like 9 months. I'm sooooo confused
Not in Paris.
Originally posted by Harald
Guys,
Not in Paris.
So you think next Tuesday 8/23/05 maybe?
Originally posted by MacRonin
WTF is wrong with you people...?!?
Pin-compatibility has nothing to do with Apple using this CPU...
Repeat after me noobs, Apple uses daughtercards for their CPUs...
Come on, macro-nitpicker, we're using shorthand here. Would Apple go to the effort of designing a new daughtercard for the 970MP? With the Intel transition on the horizon? How many units would they sell, anyway? It seems unlikely they'd be willing to design anything at all. You don't pour money into a dead-end product. So, then, pin-compatibility is very important.
Frankly, I think even the overhead of the new SKU's, packaging, shipping etc. aren't going to be worth the small number of sales. Apple will simply let the PowerMac languish as-is until the Intel models come out. Yes, he said "great new PowerPC products in the pipeline", but he was probably referring, if to anything, to the 7448.
Originally posted by Harald
Guys,
Not in Paris.
I tend to agree. Not that I would mind... But I still believe that it's PowerBook time!
Apple Expo Paris: PowerBooks ( 7448 ) and iPod minis/shuffle updates
MacWorld SF: PowerMacs and iMacs updates ( 970MP )
Spring '06: Intel minis and iBooks ( Yonah )
Summer '06: Intel PowerBooks ( DualCore Yonah )
Fall '06: Intel PowerMacs and iMacs ( Conroe )
Winter '06: Intel minis and iBooks ( Merom )
Spring '07: Intel PowerBooks ( Merom )
Summer '07: Intel PowerMacs, iMacs & Leopard ( Conroe+ )
Fall '07: Intel XServes
Originally posted by cubist
Come on, macro-nitpicker, we're using shorthand here. Would Apple go to the effort of designing a new daughtercard for the 970MP? With the Intel transition on the horizon? How many units would they sell, anyway? It seems unlikely they'd be willing to design anything at all. You don't pour money into a dead-end product. So, then, pin-compatibility is very important.
Frankly, I think even the overhead of the new SKU's, packaging, shipping etc. aren't going to be worth the small number of sales. Apple will simply let the PowerMac languish as-is until the Intel models come out. Yes, he said "great new PowerPC products in the pipeline", but he was probably referring, if to anything, to the 7448.
If we take the rumor that Apple has had samples of the 970MP for a few months now, then yes, I would fully expect them to have a redesigned daughtercard available...
And the original post that I replied to truly reads as if the poster thinks Apple still has ZIF sockets on their MLBs...
As for sales, the main market for Dual Duals would be the Pro app users...
Final Cut Pro, Motion, Soundtrack Pro, Shake, Maya, RenderMan, Modo, etc. ...
These are the apps that would love to be cranking on "Double Duals", and most of the folks using these apps in a regular day-to-day profit-making business scenario wuld think nothing of popping for a new machineto pump out work until the first (probably second) versions of the Intel PowerMacs come out...
Originally posted by MacRonin
If we take the rumor that Apple has had samples of the 970MP for a few months now, then yes, I would fully expect them to have a redesigned daughtercard available. . .
I agree. I think there is no doubt at all that Apple has 970MP daughter cards, but has even started manufacture of new PowerMac to build up inventory. If not yet, it will begin very soon. Apple had 970MP samples in 2004 I believe.
Originally posted by cubist
. . . You don't pour money into a dead-end product. So, then, pin-compatibility is very important. . .
Why do you think it is a dead end product? Developers who design applications to run on a dual-dual PPC will simply recompile for Intel Macs. No lost work, and a big head start on applications that run faster. This will generate many new PowerMac sales.
. . . Frankly, I think even the overhead of the new SKU's, packaging, shipping etc. aren't going to be worth the small number of sales.
You picked some of the lowest cost overhead items in developing a new product. Engineering development cost would be the highest, and it's already done I'm sure.
Originally posted by snoopy
I agree. I think there is no doubt at all that Apple has 970MP daughter cards, but has even started manufacture of new PowerMac to build up inventory. If not yet, it will begin very soon. Apple had 970MP samples in 2004 I believe.
The reality is that upgrading the Powermacs with these 970MPs, now and after the Intel bomb, may just end up being too late. Businesses may just look aside and wait for the Intel systems before putting down any new money on a product whose processor is EOL and software support will no doubt dwindle significantly. It remains to be seen how things will really play out. IMO, the OS advantages will not help Apple much until the Intel Macs start appearing for sale.
Originally posted by DVD_Junkie
The reality is that upgrading the Powermacs with these 970MPs, now and after the Intel bomb, may just end up being too late. Businesses may just look aside and wait for the Intel systems before putting down any new money on a product whose processor is EOL and software support will no doubt dwindle significantly. . .
Even if your analysis is true for today, Apple did most all the engineering work before the Intel bombshell announcement. Since engineering is the largest piece of new product development, there is little reason for Apple not to go ahead with its original plan. When Apple began working on the next generation of PowerMacs, the Intel move was just an idea that had been around for several years. No one knew for sure what would happen.
Originally posted by DVD_Junkie
The reality is that upgrading the Powermacs with these 970MPs, now and after the Intel bomb, may just end up being too late. Businesses may just look aside and wait for the Intel systems before putting down any new money on a product whose processor is EOL and software support will no doubt dwindle significantly. It remains to be seen how things will really play out. IMO, the OS advantages will not help Apple much until the Intel Macs start appearing for sale.
The thing is there is a lot of life left in the PPC processors as far as software support. We know the transition isn't going to be complete until sometime in 2007, so that is at least two years with PPC support. On top of that look at how long Classic is still kicking around, 5 major OS updates later. So I would say at least 6 years before PPC software becomes hard to find and by then you'll be half way to your second Mactel if you upgrade every four or five years. Also the advantages of proven technology is also there where Rev. A Mactels won't necessarily have perfect hardware or software support.
Originally posted by rickag
Not making excuses for IBM mind you, but....
I know I must have missed something in what has been posted here, but from December 2004 to August 2005 isn't "one year and a half", more like 9 months. I'm sooooo confused
oops.... 9 months it is. my bad
so i think this is something IBM could have shown apple, "look we can do 65nm with our existing 100+nm tools" but perhaps it was too costly/experimental/etc for apple.
Originally posted by DVD_Junkie
The reality is that upgrading the Powermacs with these 970MPs, now and after the Intel bomb, may just end up being too late. Businesses may just look aside and wait for the Intel systems before putting down any new money on a product whose processor is EOL and software support will no doubt dwindle significantly. It remains to be seen how things will really play out. IMO, the OS advantages will not help Apple much until the Intel Macs start appearing for sale.
After having designed enough circuit boards over the years, I can say that a daughter card is not a big deal. Cards are designed with software. A card like that could be turned around in a few days. Even after testing, it wouldn't add up to more than a couple of weeks. All it needs is one or two engineers. The question is about the support chips such as the memory controller. If two duals are being used, there might need to be a redesign there. I'm not sure though. It depends on how it works now, and I haven't investigated it.
Businesses have ordering policies in place. If they order in 18 month intervals, then they will continue to do so. You have it backwards here. They would rather continue to buy PPC machines as long as they can. They don't want to move to a new platform until everything is in place, and they have done months of testing on the new product and software.
Originally posted by onlooker
They said All Mac's would be shipping with intel processors in 2007.
For accuracy's sake, does anyone remember for sure whether Steve said in 2007 during the keynote or by 2007?