Intel "East Fork" Technology Migration

Hack Jandy writes "When Intel's Centrino platform first unveiled, industry experts were surprised to see such great performance of the Pentium M, based off Intel's P6 (Pentium III) architecture. According to sources in the industry, Intel has officially adopted the approach to migrating Pentium M to the desktop (hence, "East Fork") to offset some of its Pentium 4 processor sales. Cheaper, slower, cooler, but higher performing processors are on the way to an Intel desktop near you!"

Comments

So Intel is basically saying...

[XNormal]

"So perhaps this Pentium 4 architecture with its ridiculously deep pipeline wasn't such a great idea after all?"

Re:So Intel is basically saying...

[swordboy]

I don't think that you are seeing the whole story. Basically, Intel has been holding out for IBM's silicon-on-insulator technology because it reduces power requirements a good deal. Unfortunately for Intel, IBM is pretty sneaky when it comes to licensing and often prefer to swap technology rather than accept cash. I'd imagine that IBM is holding out for an x86 cross-license agreement while Intel does not want to give that up.

What you've seen in the past couple years is a game of chess. With each move, the other hopes that they have positioned themselves to better reach a licensing deal. Intel's move to non-clock processor ratings was a big move in this game.

From what I've seen at Intel's developer forums, they're working on some radically different architecture. Something that isn't von Neumann at all. They're calling it "massively parallel" but the industry seems to think that this means multiple cores on one chip. I think that it means thousands or millions of "processing elements" on one chip (think really small processing elements). Their claim is that they'll be able to apply this architecture to everything from mobile to high-end servers simply by adding or subtracting elements as power constraints allow.

Re:So Intel is basically saying...

[qbwiz]

They better tell programmers and compiler-writers about this soon. Any chip like this is would be very hard to program for - I suspect that any attempted move to this architecture would end up like the Itanic.

Re:So Intel is basically saying...

[gadget junkie]

"So perhaps this Pentium 4 architecture with its ridiculously deep pipeline wasn't such a great idea after all?"

It is not that deep pipeline is bad in itself; the point is, the decision to build the pIV that way was slaved to the use of MHZ as a marketing tool. That, in itself, drove the chip design in a way that essentially banned it from the laptop market, which in turn drove the design of the pentium-m , a.k.a. Centrino.

Now Intel itself is at a fork in the road, because Prescott is also geared towards higher frequencies, which means it will probably be hotter still.
Now, I do not know how much money Intel sunk in the prescott design, but if it is serious in building this new Centrino derivative processor, all this money will be washed away; and if Intel tries to keep this processor one step behind Prescott in performance, it risks a royal Chewing up by AMD.

Obscene

[BabyJaysus]

Intel employee: "Shall I try migrating Pentium M to the desktop?"

Intel boss: "Fork off!"

</shame>

Re:The pentium that should been

[eobanb]

slower, cooler, but higher performing processors

.

Slower, higher performance. Only from Intel.

Re:Why do this?

[Slack3r78]

In what way is the Pentium M "dumbed down?" Quite frankly, I'm firmly of the opinion that it's the best processor that Intel has produced to date, and I'm not alone in that view point.

The Pentium M is based on the old P6 core, with things like SSE added it to bring it up to current standards, and power saving circuitry of its own added in to suit the mobile role. The one major complaint about the chip is the fact that it's somewhat bottlenecked by a 400MHz FSB, but there's speculation that that's partly related to it currently being a mobile part. Even so, a relatively low clocked Pentium M compares very favorably to a much higher clocked P4.

Basically, the Pentium M is a move back to a P3 type design philosophy, away from the 30-stage pipeline madness Intel's gotten themselves into with Prescott. I fail to see how going with a more intelligent design is going with a dumbed down processor.

Big changes at Intel?

[data1]

It seems the company is trying to go in a significantly different direction to retain its market dominance.

1) New Non Engineer CEO :
http://www.itweb.co.za/sections/business/2004/0 411 151128.asp?S=Career%20Moves&A=MOV&O=FRGN

2) GHz No longer a big deal after marketing it for so many years as the only major thing you need to know about the performance of a computer.
http://www.theregister.co.uk/2004/10/14 /intel_kill s_4gh/

3) Shift to Better if not necessarily newer technology - see article above: oh who am I kidding....
http://www.xbitlabs.com/news/chipsets /display/2004 1111133206.html

Re:Architecture

[Gr8Apes]

Imagine AMD's speed with AMD's architectural benefits. Wait....

How high can it climb?

[ceeam]

I noticed that every x86 CPU architecture in the past decade climbed 4-5 times in MHz from inception to the "end of the line" model: 486 - 25..100(???, 133 is AMD's version and those started higher than 25), Pentium - 50..200, Pentium4 - 1200..3600 now and still has a tad in reserve as shown by extreme overclockers; similarly for AMD, K6 - 166..550; Athlon - 500..2.x(?). And now Pentium2/3 - started at 233 and climbed until around 1300, which is higher than 4/5x. But maybe there's been some really notable arch changes since P2? What're your thoughts?

What do you do when Itanic sinks?

[IGnatius T Foobar]

This is really about Intel finally coming to terms with the fact that nobody wants to buy Itanium chips. That's where Intel was headed, and Intel assumed that everyone would follow along. Unfortunately, Itanium's future depended on technology advancements that never happened, and a rate of adoption that nobody was willing to pursue.

This is why Xeon became an architectural dead end: Intel wasn't willing to move the technology forward, because Xeon was supposed to be superseded by Itanium.

Did you know that "Pentium M" is actually based on the same technology they originally called Pentium Pro? It's true. It was a good design. It didn't do all that well initially because its 16-bit performance was abysmal, and people were still running a lot of 16-bit software at the time. Now that everything is 32-bit, Pentium Pro (now Pentium M) is just fine. The fact that it gets used in laptops is a testament to its ratio of performance to power consumption.

Intel would be wise to move forward with this. They ought to ditch Xeon entirely, and perhaps even graft the AMD64 instruction set onto this chip.

Re:What do you do when Itanic sinks?

[pertinax18]

Did you know that "Pentium M" is actually based on the same technology they originally called Pentium Pro?

So are the Pentium II and Pentium III, what's your point? The article clearly states (and it is common knowlegde) that the "M" is based on the PIII, this is no secret or some massive Intel conspiracy... Yes the Pentium Pro was a great design; it really has legs to go from 166MHz to 2GHz or whatever the "M" runs at these days. But it has been a long evolutionary process, not a direct jump from the Pro to "M".

Re:What do you do when Itanic sinks?

[LWATCDR]

"Well... some folks would disagree with this. The 8051 (and followons) were huge in the embedded world.'
They still are extermly popular but not really an inovative design. But very successful but mainly for other companies Intel left the 8085 bussines a long time ago.
" The i860 wasn't intended to be a "home PC" type processor and saw good use in the HPC world (Intel Paragons, iPSC860s, etc.) and in the graphics world (high end SGI graphics cards were based on i860s - RealityEngine, etc.)" Actually the i860 was going to be a major new family of CPUs for workstations and the like. It never really lived up to it's billing. The worst problem with it was context switching was dog slow and the "smart" compilers never got smart enough. Running really tight code writen by hand running a single task they proved very fast and as you pointed out ended up in graphics cards and the like.

" Likewise, the i960 family was huge in embedded systems. They were big in printers and all sorts of other devices. The i960s were phased out for newer/better technology in the XScales. The i960 was getting pretty old :)
"
The i960 is no older than the ARM. In fact it came out a year after the first of the ARMs did. I would have to say that Intel except for the HUGE Wintel market really has not been all that successful. Frankly the have not had to since the x86 has been a huge money pump for them. I mean if you are going to win only one market that was the right one to win.
I do wonder what type of perfromance you could squeeze out of an ARM or an Alpha if you put as much money into them as Intel has with the x86.

"Well... If one thing has been proven in the past it is that software is the driving force, not hardware. It will still take some time for the near 30 years of x86 software to be replaced by "platform independent" stuff (like Java and .NET).
" You have forgoten the stealth platfrom independent stuff" Linux and c. For the server market anyway things like Samba, Apache, PHP, Perl, Postgres, and MySQL are all available to run on none Intel platforms. Linux and c are bringing write once compiler everywhere to the server world. Think of all the companies that are already porting stuff to Linux from old unix systems. Do you think they care if they are moving from a Sun or Vax to a linux box if they recompile for x86 or PPC? For the desktop you are right but even that is changing now. OpenOffice and Firebird/Thunderbird are bigger changes than anyone really wants to admit.

Not quite.

[glrotate]

The Inquirer article concerned the 1st Gen P4 Xeons with 1MB L3 vs P3 Xeons with 2MB L2. The article is 2 years old.

Bit late for me

[Spacejock]

I went through an upgrade about 2 months ago. Looked around to see whether I could get a Pentium-M motherboard and CPU (in Perth, Western Australia - hah.)

I liked the idea of throttling the CPU back when it wasn't busy. We get daytime temps of 100+ degrees (40 deg centigrade) fairly regularly in summer, keeping a hot CPU cool isn't fun.

Before I wasted too much time looking, I read about the Athlon64 3400+ and that was that. Mind you, cool 'n' quiet locked up hard on my Gigabyte K8NSNXP bios revisions F5 and F6. (Whether I was running Win Xp or Linux) Rev. F7 came out about 3 weeks after I got the board, and it's been rock solid at 1ghz to 2.4 ghz ever s--

Re:I guess.

[stevelinton]

I read this as "more heat per acre than most smelters". This piqued my curiousity.

A Pentium 4 seems to run around 217 mm^2 and produce about 100W of heat. This is quickly converted to almost exactly 2.5 million horsepower/acre. Leaving aside the livestock management problems of fitting 2.5 million horses into your 1 acre field, we now turn to a smelter, running, according to ask Jeeves at about 1400K. Radiated heat output per unit area is sigma*T^4 for a black body, less for a real material (where sigma is the Stefan Boltzman contstant), although there will also be quite a bit of convection and so on, which we ignore because it's too hard.

So, thanks to the magic of the units program, we find that the Smelter puts out about 1.18 million hp/acre, or about half the power output of the PIV.

So parent was right, P4s really do put out more heat per area (or acre) than most smelters!