The only way that will happen is if they get the hell away from motorola!Originally posted by Chaszmyr
Apple might be at 3ghz by then
I have faith that Apple will manage to keep up somehow...
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.
Intel to reach 10Ghz in 2005!
- Thread starter gaomay
- Start date
- Sort by reaction score
Intel's definitely between a rock and a hard place. They got to 3.06G by making a long, long, long cache. The P4's not going to go much faster without liquid nitrogen. Now they want to move server people to Itanium, but surprise! It's only 1GHz, and it's completely incompatible. 32-bit machine code runs like a 33MHz 486. You can recompile all your apps and run on HP's Unix...
Meanwhile AMD is working on the Hammer, a 64-bit CPU with 32-bit code compatibility. It can run 32-bit machine code fast. But surprise! It's only 1.2GHz (1.8GHz? Who knows, it's not out yet). Hence the return of the old Cyrix "PR" numbers.
Intel wants to steal the Hammer's thunder (hm, sounds like Thor) but they'll probably end up having to swallow their pride and license it.
Meantime, GHz numbers will go the "PR" route and become completely mythical.
Solution for Apple? Hey, we can play the "PR" game too! That dual 1.4, let's call it "PR-3300"!
Meanwhile AMD is working on the Hammer, a 64-bit CPU with 32-bit code compatibility. It can run 32-bit machine code fast. But surprise! It's only 1.2GHz (1.8GHz? Who knows, it's not out yet). Hence the return of the old Cyrix "PR" numbers.
Intel wants to steal the Hammer's thunder (hm, sounds like Thor) but they'll probably end up having to swallow their pride and license it.
Meantime, GHz numbers will go the "PR" route and become completely mythical.
Solution for Apple? Hey, we can play the "PR" game too! That dual 1.4, let's call it "PR-3300"!
Originally posted by cubist
Intel's definitely between a rock and a hard place. They got to 3.06G by making a long, long, long cache. The P4's not going to go much faster without liquid nitrogen. Now they want to move server people to Itanium, but surprise! It's only 1GHz, and it's completely incompatible. 32-bit machine code runs like a 33MHz 486. You can recompile all your apps and run on HP's Unix...
Meanwhile AMD is working on the Hammer, a 64-bit CPU with 32-bit code compatibility. It can run 32-bit machine code fast. But surprise! It's only 1.2GHz (1.8GHz? Who knows, it's not out yet). Hence the return of the old Cyrix "PR" numbers.
Intel wants to steal the Hammer's thunder (hm, sounds like Thor) but they'll probably end up having to swallow their pride and license it.
Meantime, GHz numbers will go the "PR" route and become completely mythical.
Solution for Apple? Hey, we can play the "PR" game too! That dual 1.4, let's call it "PR-3300"!
Nah we need to repackage everything like... Apple Currently shipping 2000+ Dual Processor PowerMacs *
* Does not equal to GHZ.
Re: Re: Re: Who cares
Booooooo and lame,,,,sure you did,,,,but where did he get HIS info? My grandmother told my mom who told my sister,,,who told me that Santa Claus is for real,,,,really. Bla Bla Bla. please note that this not an attack on YOU. Have a great day!!
Intel unlikely to reach 10GHz in 2005!
The Intel roadmap has a 32-bit chip: Code Name Nehalem , running at 6 GHz in 2005. Of course, at that point, a 32-bit chip will be simply slower because it won't be capable of running the 64-bit version of Windows now in development for AMD's 64-bit offering (AMD's chip is scheduled for Q1 2003). Some of Intel's other offerings will actually regress a bit in speed as they are redesigned for better efficiency.
It's hard to conceive of at this point, but a 32-bit chip at 6GHz is bound to have a long-ass pipeline and need all of those ticks to keep up with much slower speed, short pipeline chips. It will - as noted - be less efficient than a 64-bit competitor running 64-bit optimized software. It - in any case - will be 4 GHz slower than 10 GHz.
Look to Intel's high end (or IBM for that matter) for true state-of-the-art, you'll see they abandon clock-speed advertising with the Itanium 2 - a 1GHz chip with up-to 3mb of L3 cache (moving to 64-bit and 6mb L3 cache in the third quarter of 2003, the release season for our long-awaited IBM and Motorola unveilings).
The Intel roadmap has a 32-bit chip: Code Name Nehalem , running at 6 GHz in 2005. Of course, at that point, a 32-bit chip will be simply slower because it won't be capable of running the 64-bit version of Windows now in development for AMD's 64-bit offering (AMD's chip is scheduled for Q1 2003). Some of Intel's other offerings will actually regress a bit in speed as they are redesigned for better efficiency.
It's hard to conceive of at this point, but a 32-bit chip at 6GHz is bound to have a long-ass pipeline and need all of those ticks to keep up with much slower speed, short pipeline chips. It will - as noted - be less efficient than a 64-bit competitor running 64-bit optimized software. It - in any case - will be 4 GHz slower than 10 GHz.
Look to Intel's high end (or IBM for that matter) for true state-of-the-art, you'll see they abandon clock-speed advertising with the Itanium 2 - a 1GHz chip with up-to 3mb of L3 cache (moving to 64-bit and 6mb L3 cache in the third quarter of 2003, the release season for our long-awaited IBM and Motorola unveilings).
Originally posted by Backtothemac
Agreenster,
my iMac G3 500 runs photoshop faster than the wifes PIII 1GHZ. 1/2 the speed and it is faster at loading, running filters, etc.
So, don't believe the hype
that is the worst line anyone can come up with to defend an apple CPU, the fact of the matter is who runs photoshop on a daily basis, how many philters do you want or need to do. For the average joe who wants a computer for chat, internet and games a PC is by far a better choice by price and speed, saying duh my mac runs photoshop better than a PC is plain dumb, if mac want to catch up, they need to jump on IBMs new PPC 970 or an intel/AMD platform, I don't mind if my macs got a pentium in it. F it does what I want with speed, then I'll be happy
Something to remember is that Intel is having trouble cooling the P4 on .13 micron. At .09, or .065 (which iirc, is what Nehalem should be, perhaps it was .045), they'll be able to get a good deal higher. Of course, a dual core, 2MB on chip cache 4-5GHz 970 derivative should do just find too . Or maybe a CELL derivative with like 12 cores or something.
. Or maybe a CELL derivative with like 12 cores or something.