Intel to deliver dual-core, hyper-threaded chips earlier than expected
With development ahead of schedule, Intel Corporation today announced it is accelerating the availability of its dual-core, hyper-threaded Xeon and Xeon MP processors.
The new processors will help improve server responsiveness, speed and multi-tasking by allowing software to manage information from up to four ?brains? per Intel processor. In addition, Intel said today that it has begun a broad evaluation program of thousands of dual-core platforms for software developers and enterprise customers.
?As they did with dual-core PC processors earlier this year, Intel engineers have executed exquisitely, and because of that we?ll bring our dual-core Intel Xeon processor platforms to the marketplace well ahead of schedule,? said Kirk Skaugen, general manager of Intel's Server Platforms Group.
Originally due in 2006, Intel plans to introduce the dual-core Xeon processor MP, codenamed ?Paxville,? for servers with four or more processors later in 2005. Paxville will provide more than 60 percent better performance over previous generations and will use the Intel E8500 chipset, which has been architected for dual-core performance and was introduced earlier this year.
For dual processor servers, the company in 2005 plans to ship a premium dual-core Xeon processor, codenamed ?Paxville DP.? The processor will deliver up to 50 percent improved performance over previous generations through use the Intel E7520 chipset.
Paxville DP will be targeted at early adopters and evaluators of dual-core technology and is to be followed in 2006 by a broader family of dual-core Xeon processor-based platforms, codenamed ?Bensley? for servers and ?Glidewell? for workstations. Both processors are targeted to complete an extremely aggressive transition to dual-core top to bottom in Intel?s entire server and workstation line-up.
The 64-bit Paxville and Paxville DP processors will utilize Intel's Hyper-Threading Technology, allowing a single dual-core processor to run four threads simultaneously. The platforms will also include enhanced security features such as Execute Disable Bit and improved power management with Demand Based Switching.
Intel said it has 17 multi-core projects under development and expects more than 85 percent of its server volume exiting 2006 to be multi-core processors. In addition to the Xeon processors due in 2005, Intel began shipping the dual-core Intel Pentium D processor for uni-processor servers in July 2005 and remains on track to begin shipping dual-core Itanium processors by the end of the year, the company said.
Intel?s evaluation program, which began on Monday, will ultimately deliver thousands of dual-core platforms based on Pentium D processors, Xeon processors, Xeon MP processors and Itanium processors to early adopter customers and software developers through 2005 and into 2006.
The new processors will help improve server responsiveness, speed and multi-tasking by allowing software to manage information from up to four ?brains? per Intel processor. In addition, Intel said today that it has begun a broad evaluation program of thousands of dual-core platforms for software developers and enterprise customers.
?As they did with dual-core PC processors earlier this year, Intel engineers have executed exquisitely, and because of that we?ll bring our dual-core Intel Xeon processor platforms to the marketplace well ahead of schedule,? said Kirk Skaugen, general manager of Intel's Server Platforms Group.
Originally due in 2006, Intel plans to introduce the dual-core Xeon processor MP, codenamed ?Paxville,? for servers with four or more processors later in 2005. Paxville will provide more than 60 percent better performance over previous generations and will use the Intel E8500 chipset, which has been architected for dual-core performance and was introduced earlier this year.
For dual processor servers, the company in 2005 plans to ship a premium dual-core Xeon processor, codenamed ?Paxville DP.? The processor will deliver up to 50 percent improved performance over previous generations through use the Intel E7520 chipset.
Paxville DP will be targeted at early adopters and evaluators of dual-core technology and is to be followed in 2006 by a broader family of dual-core Xeon processor-based platforms, codenamed ?Bensley? for servers and ?Glidewell? for workstations. Both processors are targeted to complete an extremely aggressive transition to dual-core top to bottom in Intel?s entire server and workstation line-up.
The 64-bit Paxville and Paxville DP processors will utilize Intel's Hyper-Threading Technology, allowing a single dual-core processor to run four threads simultaneously. The platforms will also include enhanced security features such as Execute Disable Bit and improved power management with Demand Based Switching.
Intel said it has 17 multi-core projects under development and expects more than 85 percent of its server volume exiting 2006 to be multi-core processors. In addition to the Xeon processors due in 2005, Intel began shipping the dual-core Intel Pentium D processor for uni-processor servers in July 2005 and remains on track to begin shipping dual-core Itanium processors by the end of the year, the company said.
Intel?s evaluation program, which began on Monday, will ultimately deliver thousands of dual-core platforms based on Pentium D processors, Xeon processors, Xeon MP processors and Itanium processors to early adopter customers and software developers through 2005 and into 2006.
Comments
The chips that Apple is interested in is the Pentium-M. The only place that we are likely to see a Xeon is on a desktop, but there is very little info on what Apple's desktop plans are.
Homies, I'm on the jonze for a new powerbook. I would wish on a friggin star if I thought it would help...
Originally posted by ajmas
You won't see an Xeon on a PowerBook any time soon, unless you are wanting to go beyond melting your hands and having a battery life of two seconds. The xeon is a high performance processor, which just pulls the watts out of the mains at a ridiculous rate.
The chips that Apple is interested in is the Pentium-M. The only place that we are likely to see a Xeon is on a desktop, but there is very little info on what Apple's desktop plans are.
Yep... more likely to see them in the PowerMac and the Xserve than anything else.
But thanks for clearing that up... I can wait till I get my hands a 1.8ghz Pentium M PB!
Originally posted by popmetal
does this make it likelier that we'll see Intel based PowerMacs sooner?
Nah, not really. Intel hardware will have to follow the availability of ported/recompined software. Not the other way around. So I think it'll be the same waiting period.
Originally posted by nowayout11
Nah, not really. Intel hardware will have to follow the availability of ported/recompined software. Not the other way around. So I think it'll be the same waiting period.
In addition, Apple doesn't want this first generation of chips. Not as much of a performance swing as would be required to get decent performance out of Rosetta.
With portables, the MIPS/watts ratio won't be there until Yonah in 2006 sometime around the middle of the year. Esp. if they are looking for 64 bit PB's.
But I think Intel is waking up. AMD might not have much time left to move in if their lawsuits don't do Intel much damage.
Originally posted by Mike Moscow
The new powerbook is coming... ITS COMING!
Originally posted by Glamingo
Crazy, i never would have thought that we would all be going towards dual core proccesors.
Crazy? I never thought we'd have personal computers.
Originally posted by macFanDave
Anybody remember the last time a vendor of Apple processors delivered a chip early?
Uhmmmm that would be a uhm... a... uhm.... a [B A BIG FAT NO! NO! ]NO![/B]
Ike
Blogging with Ike
I know it's fashionable to think otherwise these days.
Originally posted by melgross
You know, people, Apple didn't have this problem until the G4's intro in 1999. Before that they were delivering ahead of Intel, both in speed and power.
I know it's fashionable to think otherwise these days.
It's fashionable because it's been like this for half of the PPC's lifetime.
Originally posted by melgross
You know, people, Apple didn't have this problem until the G4's intro in 1999. Before that they were delivering ahead of Intel, both in speed and power.
I know it's fashionable to think otherwise these days.
Actually it wasn't the G4 intro that was the problem, it was that the G4 stayed where it was introd for like 18 months.
And they might have been ahead of intel , but that doesn't mean any of those chips were on time. And some chips never came (wasn't there a partnership working on something called Taligent, or was that just an OS? And didn't some company proclaim that they could make the G3 better then before, better, stronger, faster, sort of a bionic G3 chip, only to have it turn out they could barely squeeze a couple of extra % in the speed category?)
Originally posted by nowayout11
Nah, not really. Intel hardware will have to follow the availability of ported/recompined software. Not the other way around. So I think it'll be the same waiting period.
Actually, I think it will be the other way around.. we'll have some Intel software out, but when Apple has Intel versions of their stuff ready, they'll release it.. essentially turning up the head on 3rd party software developers to get Intel native versions out the door.
I imagine if Apple has the chips, and has the motherboard/case designs, and then has their software done, you'll see them asap on apple.com
Originally posted by ajmas
IBM Had been providing high power CPUs, just not the type Apple needed for it products.
The IBM PPC processor has seen the end of their day in mainstream computers. In 1999 it was apparent that the PPC was in trouble. The PPC obviously has it's place; which is to be engineered for specific purposes, and applications like high range GFLOP simulations, and such for scientific computing. They (IBM) are now harnessing that potential in another constrained specifically designed environment by producing processors for console gaming machines. They blew Apple away with the #'s they could achieve in certain controlled situations, but translating those #'s into something that isn't highly constrained, and controlled; like having a varying bunch of Applications under a ever changing environment is more than the PPC can handle. IBM would never admit that to Apple, but Apple finally realized it, and got the hell out of dodge. Better late than never I say.