Phil Of Mac
macrumors 68020
I'm concerned.
If Power5 is at "Hot Chips", does this mean we can't design a PowerBook based on it? 😀
If Power5 is at "Hot Chips", does this mean we can't design a PowerBook based on it? 😀
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.
Power5 at Hot Chips
- Thread starter MacRumors
- Start date
- Sort by reaction score
Re: Re: Why use the 970 on a 90-nm process?
post deleted as another poster covered everything and more. 🙂
post deleted as another poster covered everything and more. 🙂
AidenShaw
macrumors P6
Not correct - Pentium M (Centrino) 1.7 shipping now
The 1.7 GHz Centrino has been shipping since July - I've received 3 of them so far.
http://www.dell.com/us/en/biz/products/model_latit_2_latit_d600.htm
"Intel® Pentium® M processors with speeds up to 1.7 GHz"
The 1.7 GHz Centrino has been shipping since July - I've received 3 of them so far.
http://www.dell.com/us/en/biz/products/model_latit_2_latit_d600.htm
"Intel® Pentium® M processors with speeds up to 1.7 GHz"
Re: Re: Why use the 970 on a 90-nm process?
Did you ever stop to think that IBM has almost certainly been working on a PowerPC version of the Power5 core at the same time that the Power5 chip was being developed? If IBM waited until the Power5 chip was in production before starting on it's PowerPC version, then the PowerPC development would move at a snails pace and Apple would fall further behind Intel.
That's why IBM has almost certainly been working on the next version of the 970 before the 970 was even in full production.
It remains to be seen when the PowerPC version of the Power5 core will be in production, but I'd bet it will be a lot sooner than you believe and I would be surprised if it wasn't the next version of the 9XX chip.
Did you ever stop to think that IBM has almost certainly been working on a PowerPC version of the Power5 core at the same time that the Power5 chip was being developed? If IBM waited until the Power5 chip was in production before starting on it's PowerPC version, then the PowerPC development would move at a snails pace and Apple would fall further behind Intel.
That's why IBM has almost certainly been working on the next version of the 970 before the 970 was even in full production.
It remains to be seen when the PowerPC version of the Power5 core will be in production, but I'd bet it will be a lot sooner than you believe and I would be surprised if it wasn't the next version of the 9XX chip.
Re: Re: Why use the 970 on a 90-nm process?
Well, looking at Ace's Hardware Dataminehttp://www.aceshardware.com/SPECmine/index.jsp?b=2&s=0&v=2&if=0&r1f=2&r2f=0&m1f=0&m2f=2&o=0&o=1 SPECfp numbers, the 1.5 GHz Itanium 2 is about 24% faster than the 1.7 GHz Power4.
On SPECinthttp://www.aceshardware.com/SPECmine/index.jsp?b=0&s=0&v=2&if=0&r1f=2&r2f=0&m1f=0&m2f=2&o=0&o=1, the 1.5 GHz Itanium 2 is about 18% faster than the 1.7 GHz Power4. The 1 GHz Pentium-M scores a 687, multiply that by 170% to get the 1.7 GHz P-M and you have a score that is roughly the equivilant of a 1.8 Ghz 970 processor.
IBM seems to have been speaking about system performance when they mention that the Power5 will be 4X faster than a Power4. That would include doubling the amount of processors in a system. To quadruple the performance of the chip would also entail a large increase in the power use, which would make it prohibitive. In other words the chip would use say 400 watts instead of 100 watts.
Well, looking at Ace's Hardware Dataminehttp://www.aceshardware.com/SPECmine/index.jsp?b=2&s=0&v=2&if=0&r1f=2&r2f=0&m1f=0&m2f=2&o=0&o=1 SPECfp numbers, the 1.5 GHz Itanium 2 is about 24% faster than the 1.7 GHz Power4.
On SPECinthttp://www.aceshardware.com/SPECmine/index.jsp?b=0&s=0&v=2&if=0&r1f=2&r2f=0&m1f=0&m2f=2&o=0&o=1, the 1.5 GHz Itanium 2 is about 18% faster than the 1.7 GHz Power4. The 1 GHz Pentium-M scores a 687, multiply that by 170% to get the 1.7 GHz P-M and you have a score that is roughly the equivilant of a 1.8 Ghz 970 processor.
IBM seems to have been speaking about system performance when they mention that the Power5 will be 4X faster than a Power4. That would include doubling the amount of processors in a system. To quadruple the performance of the chip would also entail a large increase in the power use, which would make it prohibitive. In other words the chip would use say 400 watts instead of 100 watts.
Note that the 4x claim is based off of the "original" Power4 which probably refers to the "original" 1 Ghz Power4, not the "current" 1.7 GHz Power4++.