Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
IBM's POWER Event
- Thread starter MacRumors
- Start date
- Sort by reaction score
gekko513 said:It means that Apple will use a 3x multiplier with the PowerMac Updates ... Don't be surprised if you see PowerMacs at 2.4, 2.7 and 3.0 GHz ... and eventually 3.3GHz ... When? I don't know?
Actually there is rumor that Apple will move to a 1:1 ratio. The main reason is for SMP systems, allowing each processor to take advantage of their shared caches with very little penalty. On the Power5, IBM already moved t a 1:1 ratio between dual core modules. Of course, on the Power4 the dual core already had 1:1 but 1:2 between dual core modules. It makes sense then that Apple take it to 1:1 eventually. I wouldn't count to much on a marketing brochure. IBM and Apple will not leak such information that way.
However, the memory controller doesn't need that much speed since memory this year will not exceed 1GHz. So Apple can include an on-die memory controller that can handle DDR and DDR2 memory. This way Apple reduces the latency of going through the north bridge and accesses RAM directly (like the Opteron currently does.) It's not that big of a deal for most of the Apps that run on a Mac, since the G5's awesome bandwidth is more important for the professional community. But, it does seem like the next logical evolution and would greatly simplify the North bridge while adding another 10-20% speed boost on heavy memory access apps, like games and database transactions.
Some_Big_Spoon said:
I don't think this is gonna end up in a "980" Power5 derivate. Why should Apple focus on such a feature?
I guess it is something similar to the S/390 architecture (also developed by IBM!), where you can have multiple "virtual" machines on one hardware platform (managed by the system architecture, not like Virtual PC that depends on the multitasking features of Mac OS X/Windows). Linux has been running on this platform for quite a while already. Using one hardware servers, you can run many different OSes with software servers, for example.
Macrumors said:
Wasn't there a rumour recently that had Jobs and the head guy at Sony meeting? Hmmmmm......
"I call as I see 'em and if I don't see 'em - I make 'em up"
It seems it is the PowerPC 400 series that sony has licenced. The ZDnet article says Sony has licenced for system on a chip prcoessors ,and another article on that site mentions the 400 series is a system on a chip and the only one they are licencing so far as it makes sense. The 400 series is used in devices such as pdas...so a new Clio based on the 400 series? Is the 400 series the same as the G3?
Some_Big_Spoon said:
Revolutionary?
You mean like running multiple OS's on x86 using VMWare?
Or multiple instances of the OS concurrently (with full virtualized hardware) like people have been doing on IBM mainframes for 20 years?
You and I have different definitions of 'revolutionary'. Yours seems to mean "doing something everybody else has been doing for a long time".
Likely Mac upgrades
After doing a little math and guesstimates based on the IBM pdf listed earlier, and the information supplied by everymac.com and Motorola's site, here's my predictions (let me know your thoughts):
New PowerMac G5s (soon) running DUAL PPC970fx at 2.0, 2.2, and 2.5 Ghz (given the info provided at the Semiconductor Forum in Jan or Feb).
PowerMac G6s (using a Power5 derivative 980 chip) intro at WWDC running up to 3.2 Ghz and shipping in August.
New iMac G5s running SINGLE PPC970fx at single 1.4 - 1.8Ghz (since iMac is a consumer level PC, they wouldn't make it run faster than a Powerbook as evident from recent configurations).
New Powerbook G5s (no more G4 speed bumps since PPC7447 and 7457 max out at 1.25-1.33 range) running SINGLE PPC970fx at 1.4 - 2.0 Ghz (since the 970fx consumes only 24.5W at 2.0 Ghz, which is less than the approx. 26.0W the current G4 1.33 Ghz runs at).
New iBook G4s still using PPC7457 chips, but maxing out at 1.25 - 1.33 Mhz.
Not sure what G4 chip the new eMacs (if line still continues) will use, but will probably go up to what the current iMacs are running at, 1.25 Ghz?
Don't think we will see a Hybrid 32/64 bit version of OS X until 10.5, because of the massive conversion of all apps to a 64 bit structure, and also when all of Apple's computers are running at least a G5.
After doing a little math and guesstimates based on the IBM pdf listed earlier, and the information supplied by everymac.com and Motorola's site, here's my predictions (let me know your thoughts):
New PowerMac G5s (soon) running DUAL PPC970fx at 2.0, 2.2, and 2.5 Ghz (given the info provided at the Semiconductor Forum in Jan or Feb).
PowerMac G6s (using a Power5 derivative 980 chip) intro at WWDC running up to 3.2 Ghz and shipping in August.
New iMac G5s running SINGLE PPC970fx at single 1.4 - 1.8Ghz (since iMac is a consumer level PC, they wouldn't make it run faster than a Powerbook as evident from recent configurations).
New Powerbook G5s (no more G4 speed bumps since PPC7447 and 7457 max out at 1.25-1.33 range) running SINGLE PPC970fx at 1.4 - 2.0 Ghz (since the 970fx consumes only 24.5W at 2.0 Ghz, which is less than the approx. 26.0W the current G4 1.33 Ghz runs at).
New iBook G4s still using PPC7457 chips, but maxing out at 1.25 - 1.33 Mhz.
Not sure what G4 chip the new eMacs (if line still continues) will use, but will probably go up to what the current iMacs are running at, 1.25 Ghz?
Don't think we will see a Hybrid 32/64 bit version of OS X until 10.5, because of the massive conversion of all apps to a 64 bit structure, and also when all of Apple's computers are running at least a G5.
jayb282 said:
VMWare has had a product like this (i.e., no host OS) for a while, and VirtualPC's big brother is supposed to be just about cooked (both intended for server use). That being said, IBM has had this in their products for so long, that I can only hope they have some new stuff (or at least a new cheesy ad about it).
Someone mentioned that if more of the G5's end up getting out there because of Microsoft or Sony or whoever, the price will go down. This probably true. They were however skeptical about Apple passing on the savings. Since Apple's profit margins are 2% he does have a valid point, they certainly could use the money. (Compare this to Microsoft's 25%. Ouch.) However, since people do take price into account when buying a computer, if Apple can keep good margins and bring the cost down, I doubt they will hesitate. They are at roughly 5% of the market, anything that can work in their favor I expect they would consider.
jyvin said:
IBM lists the original 970 at a top speed of 1.8 GHz, with up to a 900 MHz bus. It seems the 2 GHz frequency for the 970 in the PowerMac and it's 1 GHz bus are exclusive to Apple, similar to the frequencies that Apple with the G4 that were well beyond the listed 1 Ghz topend speed for the chip on Motorola's website.
I would not be surprised to see Apple using a 3 GHz 970FX, with a 1.5 GHz bus, in the next few months. The effective bus speed would actually be less than 1.5 GHz due to its inherent design, so Apple could potentially keep that bus speed busy by using 667 MHz DDR2 memory that is just recently went into production. Because this memory is now being produced in very low volumes, (due to the X86 chip manufacturers not yet using it) there will be a considerable price premium if Apple does go ahead and use it. Considering the price of the topend PowerMac computers, the additional costs of 667 MHz DDR2 memory would not add substantially to the overall price of the computer. Although the costs for adding memory will come as a sticker shock to many potential purchasers.
amyhre said:
Where do you get the information that Apple's profit margins are 2%? And are you referring to hardware or software? I've heard that Apple sells their OS at a loss to move their hardware, but I'm reasonable certain that the margin is above 2%.
If IBM can't get the bus to run faster than 1.1GHz reliably then the fastest one would be on a 3:1 multiplier, i.e., an 833MHz bus for a 2.5GHz PPC 970.squatch said:
This would be slower than the 1.1GHz bus on the 2.2GHz model, so Apple might decide to go 3:1 across the line for this upcoming speed bump.
Apple might be risking 1.2GHz bus speeds however for 2.4GHz processors.
So it is either:
2:1 Dual: 2.0, 2.2, 2.4 GHz
3:1 Dual: 1.8, 2.1, 2.4 GHz
As no rumours are putting 970FX's at 2.7GHz at the moment. This could be a pleasant surprise though couldn't it? Dual 2.1, 2.4 and 2.7, with a bump to Dual 3.0 in September without any system modification.
Yeah, I'm assuming that Apple won't care to run the bus at non-hundred-rounded speeds here.
Next year I reckon. Unless IBM really do release POWER5 and the PowerPC 900 equivalent at the same time. Regardless this chip will eat more power because it will be capable of doing a lot more.
I'd hope for this as well, actually 1.6, 1.8 and 2.0 GHz I would hope. Why dumb down the iMac speed unnecessarily?
Mr Maui said:
Only place I could say is his a$$. 2%??? Apple enjoys some of the highest, if not the highest, gross margins in the computer industry. Take a look at their latest 10-K filing.
http://www.hoovers.com/free/co/secdoc.xhtml?ipage=2485511&doc=1&num=31
squatch said:
The 970fx should easily reach 3 Ghz. The die shrink and IBM adding up to a 20-30% boost in speed with strained silicon gives the 970fx the potential to reach speeds of at least 3 GHz.
Just because IBM does not mention speeds beyond 2.5 GHz for the 970fx does not mean that Apple will not get 3 GHz 970fx chips. Apple may have exclusive rights to use higher chip speeds than what IBM officially lists.
Odd no debate concerning AltiVec
It stands to reason Apple with POWER 5 Core chips will modify the design to include an Apple designed AltiVec branch to keep its G5/G6 line in check.
When the Power6 Core chips arrive they'll modify once more with a newer version of AltiVec.
Since neither the POWER 4 or POWER 5 has has AltiVec it stands to reason the parallel development involves Apple augmenting this into their own modified POWER 5+ as has been mentioned prior.
Whatever the FSB ratios to actual clock speed turn out to be I'm sure it will be a competetive compromise between Server Process Scalability and Workstation Number Crunching.
It stands to reason Apple with POWER 5 Core chips will modify the design to include an Apple designed AltiVec branch to keep its G5/G6 line in check.
When the Power6 Core chips arrive they'll modify once more with a newer version of AltiVec.
Since neither the POWER 4 or POWER 5 has has AltiVec it stands to reason the parallel development involves Apple augmenting this into their own modified POWER 5+ as has been mentioned prior.
Whatever the FSB ratios to actual clock speed turn out to be I'm sure it will be a competetive compromise between Server Process Scalability and Workstation Number Crunching.
That was profit margins, not gross margins
He stated profit margins, not gross margins. Apple has lost money on sales, minus expenses, for about half of the last 12 fiscal quarters. The company made a profit in some of those quarters off of interest and investments.
rdowns said:
He stated profit margins, not gross margins. Apple has lost money on sales, minus expenses, for about half of the last 12 fiscal quarters. The company made a profit in some of those quarters off of interest and investments.
The IBM VP at WWDC laid out the roadmap for the PPC and said that the development of the Power5 and the 97x equivalent would be carried out as a parallel development effort. If you don't care to believe me, that's fine. I was following the event closely at the time and I know what I heard and read. If you search the MR archives, I'm sure you can dig this info up.jyvin said:
Or if he still doesn't believe you he can become an ADC Member and download the QuickTime conferences and sift through the session archives.
daveL said: