Intel to Dump Pentium 4 in Favor of Pentium M

Opinion writes "According to The Register, Intel is to dump its Pentium 4 plans in favour of the new Pentium M architecture. The scrapped Tejas and Jayhawk processors represented Intel's next-gen 90nm P4 CPUs, due to arrive in 2005."

Comments

Where are they going to dump them?

[Anonymous Coward]

I'll pick up as many as i can carry..

Re:Where are they going to dump them?

[Rick.C]

I'll pick up as many as i can carry..

Wear an oven mitt... they tend to run HOT!

End of an era?

[swordboy]

So... What's the deal with Moore's law? It appears that heat density vs. pricing trends are now causing microprocessors to compete with charcoal (very hot, very cheap).

Is this the end? Or is Intel just trying to squeeze every last drop of cost out of a deal with IBM on their silicon-on-insulator patents?

Re:End of an era?

[croddy]

or are they just feeling the heat from AMD?

...or is that heat coming from their current products? ;-)

Re:End of an era?

[Paladin128]

It's an unexpected turn that makes a LOT of sense. If you read the article, the real catalyst for this change is the decision to go with 2 or more cores on one die that share the same L2 cache. The P4 is a poor architecture to do this with. Yes, nothing can really beat it at simple integer math, but it's got lots of problems:

1) The core is fscking big!
2) high frequency == draws lots of juice == runs way too hot
3) 20 stage pipeline (or like 30 in case of Prescott) makes penalties way too high on a branch mis-prediction, and requires more cache to minimize the impact.

The Pentium M architecture has a relatively high IPC, and lack of int throughput that is lost from lower overall clockspeed can be overcome by paralellism that multicore will bring. It also is rather efficient as far as power goes, and a much smaller core overall.

Re:End of an era?

[anderm7]

Yeah, I agree. I this is a great move. I'm an EE in Microelctronics, and I had been very dissapointed in Intel's tricks to get MHz up. For instance, those overly long pipelines. I'm glad they finally decided to come around and realized that both MHz and CPI(cycles per instruction) matter.

To first order, a chip is only limited by the setup & hold time of a latch, but that may not be a very good chip. It may run at 50 GHz, but its not going to do much more than heat up your case.

Re:End of an era?

[dasmegabyte]

I can't wait to see how they spin this is marketing.

"Say, guys? When we told you that clock speed was the only factor in speed, and that a long pipeline was the only way to go, we were joking! NO NO NO, wait, don't go buy AMD chips or G5s, because -- uh -- they're the suck. Look, a dancing guy in a shiny suit uses our chips!"

Re:End of an era?

[timeOday]

I can't believe this, to me it seems like a big deal. The Pentium-M is a modded Pentium-3. If this is true, Intel is abandoning the whole direction they took with the P4 = MHz at all cost. Also IIRC the Pentium-M was designed in Israel, and was their first-ever chip delivered on time, so this isn't a great development for US computer engineering.

I don't think we can count on quiet, low-power desktops though. I bet Intel will just ramp up the Pentium-M until it's a hot as the P4 (but by then it will be faster than todays P4 due to higher IPC).

Good Idea

[derphilipp]

This seems like a really smart idea. Dont go an get the Ultra-Gigaherz-Processor but a descend, processor that consumes only a low amount of power -> Longer batterylife for laptops -> Silent PCs -> Longer lifetime of the processor (?)

Pentium mm

[Mad Man]

due to arrive in 2005

Shouldn't that be Pentium MMV?

Re:Pentium mm

[mrogers]

Your megahertz may vary?

Power consumption is important

[cpghost]

The more laptops out there, the more important are power saving CPUs. Pentium-M's are a good step in the right direction after the P4 90nm debacle.

Even in the server market, cutting on power consumption is getting more and more important. If you have a park of 1000+ machines in a data center, power consumption matters.

Re:Power consumption is important

[getch();]

The issue of power consumption is rapidly becoming much more significant than even the parent poster realizes.

The general dynamic power (operating power) equation for CMOS circuits has switching frequency as a squared term. Voltage and junction capacitance (think die size here) are also present, but are not squared.

If Intel were to take the P-IV architecture as far as it had planned, an extra few bucks for electricity would be the least of its worries. Without some unforseen advancement, power per unit area would become a (relatively) intractable problem. Even though voltage and die size would probably decrease, the increase in frequency coupled with the reduced area would likely provide a serious problem for cooling. I've read papers that have estimated that air cooling won't be able to dissipate much more heat than it's already required to. Taken far enough, the head produced could just vaporize the silicon (obviously that's not occurring in the near future).

In short: good move, Intel.

Re:Power consumption is important

[ballpoint]

Not only for laptops or server rooms. My power consumption at home has increased by 25% in three years due to increasing computer use by kids & wife.

I'd like to install still more always-on equipment like webcams, video servers and such. But, with energy prices that will probably triple over the next 10 years, I'm not going to be able to afford these increases much longer.

Change of ideas

[SavedLinuXgeeK]

As stated in a previous article, I think, Intel has been running the PIV name for a long, long time,(in computer years), and now with AMD64 coming out, people will see the PIV as old, and the AMD as new, even if things are comparable. Consumers are extremely superficial (Speaking from sales experience). I think this may just help Intel get some more umph into their line, before 64-bit hits critical mass.

No Suprise

[paitre]

This really shouldn't be a suprise to -anyone- who's been paying attention to what's been going on.
Prescott is disturbingly hot, and the next-gen chips had no real hope of being much cooler. At most 10-15%, which wouldn't have gotten near their MHz goals.
P-M, on the other hand, is a damned good chip in its own right, has better IPC, and is a better CPU, all around, than the P4 line.

Now, what does this mean for those of us in the enterprise space? Are we -really- going to have to wait until 2006 for a new chip iteration from Intel? If that's the case (and I -really- doubt it), AMD would have a disturbingly large (and long) opening in which to pitch its wares...Intel would definately lose marker share in the server arena at that point.

So, multi-core P-M chips for the desktop next year-ish. So we're stuck with the hotplate known as Prescott until then. Guess I'll be sticking with AMD for a while yet :)

Re:No Suprise

[ionpro]

...consuming only 70% of the power of the AMD Athlon 64 mobile chip with about as much CPU power.

Eh? I don't think so. The Athlon 64 mobile chip (at least the 3000+) achieves near performance parity with the desktop chip. Near as I can tell, the IPC of the Athlon and Pentium-M archetectures is nearly the same per clock (and I own one of each). For instance, my 1.3Ghz Pentium-M performs equivilantly in benchmarks to my Athlon 1800+ desktop (which is handicapped by PC133 memory). But the clock speed of the Pentium Ms is far below the clock of the Athlon mobile, and it wasn't designed to ramp in clock speed at all. The Athlon, on the other hand, has shown itself to be quite the clock speed maven, going from a 550Mhz Duron Slot-A to the short-lived Throughbred-B Athlon 2800+ at 2.25Ghz (and the soon-to-be-released Athlon 64 3700+ at 2.4Ghz)

I agree, though; it will be interesting to see AMD as the overly-hot higher clockspeed contender in this new processor race. AMD has never done particularly well with mobile chips (as witnessed by Intel's 85% market share in the mobile sector); perhaps this will force their hand on that front somewhat.

Religious Nomenclature?

[Himring]

Dothan is in due course expected form the basis for 'Jonah', Intel's first two-core Pentium M, due to ship during H2 2005, possibly at 65nm. To date, Jonah has been scheduled to be succeeded by 'Merom' and 'Conroe', two chips based on the same architecture, during H1 2006. While Merom is to be pitched at notebooks, Conroe - crucially - is a desktop chip.

Dothan: Meaning: two wells. A famous pasture-ground where Joseph found his brethren watching their flocks. Here, at the suggestion of Judah, they sold him to the Ishmaelite merchants (Gen. 37:17). It is mentioned on monuments in B.C. 1600.

Jonah (We all know who Jonah was and/or you need to back to sunday school....)

Merom (WebBible Encyclopedia) - christianAnswers.Net. Merom. Meaning: height. a lake in Northern Palestine through which the Jordan flows

Looks like Intel got some religion....

Re:Religious Nomenclature?

[kmcmartin]

iirc, the Pentium M was designed at Intel's Israeli division, so this makes some sense compared to the old Washington/Oregon naming scheme.

Re:Religious Nomenclature?

[CodeMonkey4Hire]

Most of the products that come out of Chandler, AZ are named after Arizona geologic features. Their codenames are chosen so that there is no possible infringement. Since it is not possible to trademark geological features (or Biblical/historical features), they are safe bets. As the previous poster noted, there is some work going on at their [relatively] new site in Isreal.

Faster Pentium M?

[strictnein]

Have they been able to ramp up the speeds for this architecture? None of the articles that I've read even speculate on what speeds these would be introduced at. I know Intel was planning on releasing a 2.0GHz Pentium M in the near future, but what about for desktops?

Side note: "Whitefield" a new processor in the Xeon line based somewhat aroudn the Pentium M, was created in India.

Is this surprising ?

[data1]

It has been known for quite some time that the Pentium-M processors would outperform desktop chips even when clocked at a higher frequency.
Seems that Intel finally wised up and is exploiting the technology in the Pentium-M Chips to lower its development costs even though that isnt explicitly stated in the article.
Yes, I did RTFA.

About the P-M architecture

[Anonymous Coward]

Ars has an in-depth article on the Pentium-M architecture. A quote from the conclusion:

The PM takes one of the P4's strengths--its branch prediction capabilities--and improves on it, adding its advantages to the strengths of the P6 architecture. The PM also deepens the P6's pipeline a bit, allowing for better clockspeed scaling, but without making clockspeed the central factor driving performance. In short, the PM looks like what the P4 might have been, had Intel not been so obsessed with the MHz race--it's a kind of alternate past, but one that may provide a glimpse of Intel's future.

Return of the son of the revenge of the P6

[zsazsa]

Wow. This is amazing. The P6 (PPro, PII, PIII) architecture is coming back to the desktop. This does make pretty good sense. The P6 has high IPC, and by applying some Pentium 4 tricks (Quad-pumped FSB, longer pipeline), this can make for a killer CPU. For more information, check out this Ars Technica Article on the Pentium-M's P6 heritage. The chip doesn't even lie about it - its CPUID reports a P6 family CPU.

Re:hrm

[twbecker]

It's the fan in my pc that's loud not the processor. even small fans make noise.

Uhh, yeah. And the reason you NEED that big fan is because of all the heat that CPU is generating. Smaller fans = less noise.

I still have a couple Pentium I with MMX running and without a hitch. How much longer are you talking about?

And what kind of temps do they run at? Much cooler than a 3.2Ghz Prescott, I can promise you.

Not to sound like an Apple zealot (I'm far from it), but it seems like you've bought into the "Mhz myth" hook, line and sinker. Lower power and lower speeds does not need to equal lower performance.

You are groping for....

[Groo Wanderer]

"So, like, where's the Hexium, Heptium, Octium?"

I believe the term you are groping for is 'Opteron'.

-Charlie

(Apologies, I couldn't resist such low hanging fruit).

Re:hrm

[w3weasel]

I still have a couple Pentium I with MMX running and without a hitch. How much longer are you talking about?

That's not a fair comparison... you can hardly even fry an egg on those old processors.
Today's processors let you fry that egg with ease, while simultaneously calculating whether the egg preceeded the chicken, render the resulting proto-fowl in stunning 3D, with time left over to disprove your own existance.

Re:hrm

[mobiux]

"but it seems like you've bought into the "Mhz myth" hook, line and sinker"

I think that this is going to be a HUGE problem for intel. For years this has been the major marketing tool they've used. So now, they are just going to say, "Um, oh yeah, all that stuff we told you, about mhz, ghz, and stuff, that doesn't really apply to us anymore."

AMD and Apple really should take advantage of this and do a little "we were right all along" ad.

Re:braniac vs. speed demon

[Yoda's Mum]

The long pipeline approach was sustainable for a while, but with their newer processes (like the 90nm process used for the Prescotts) the heat costs of having the longer pipelines have proven too high. Their long pipeline design worked quite well for the Pentium 4's - give them enough cache, and they perform spectacularly, but the even longer pipelines required to keep cranking the clock speeds up, as with Prescott, are starting to be quite detrimental to the design. The Prescott architecture may be able to run at much higher clock speeds than the previous Northwood P4's, but they do so at the cost of requiring an even larger cache, and a much improved branch predictor. Had the improved branch predictor and increased cache simply been implemented on an existing Northwood core, and if Intel manufactured the chip on their 90nm process, it's quite likely that they'd have an even better performing chip than what Prescotts are capable of at higher clock speeds. That's all conjecture though - Intel didn't go that way, they let their marketing people decide on what the Prescott was going to be, and are now paying for it.

Re:Intel is so far behind anyway

[Chanc_Gorkon]

80186 was not a failure. I just was not used in PC's. Once upon a time, there was no real difference in the embedded field between the desktop processor and the processor used in say a traffic light. The 80186 was used in lots of embedded solutions. Checkout Wikipedia

Re:Lemme guess...

[mcbridematt]

Current breed of Pentium M's are pin-for-pin compatible with Socket 478 Pentium 4's, but appear to use a different type of GTL+ signalling. I guess that Intel will release a Pentium M version for LGA 778 (the new socket).

Finally, Intel realises that some long pipeline design with zero decent hardware rotation (up to Prescott), requiring huge cache and big clockspeed isn't that good.

Re:hrm

[devaldez]

Gotta love revisionist history...AMD was the king of MHz marketing back in the day...oh yeah, don't worry that we're AMD cuz our K6 is clocked at the same speed, so we're just as fast...except that it didn't work that way for them, either.

Feel free to hate marketing (I'm in marketing and I hate it), but keep the right sense of perspective and history or you simply lose credibility. If you want to hate a company, simply hate them, don't seek ways to justify the hate.

Re:where does that leave performance freaks like m

[Gannoc]

No AMD suggestions please. I've never found them to be as stable as Intel CPUs .

I have an Intel machine myself, but have had AMDs in the past. The reason people think that AMD chips are unstable is that many people buy AMD processors when they're trying to get a cheap computer, and also use cheap memory, cheap motherboards, etc.

Get an AMD with a good motherboard, and it is as stable as an Intel.

Re:Intel is so far behind anyway

[LocoBurger]

The Pentium M is based on the same P6 core as the Pentium Pro through the Pentium III.

The really interesting part about this story is that Intel is going from their seventh generation architecture (Pentium 4) back to their sixth generation architecture (Pentium Pro/II/III/M).

We all knew this Pentium 4 thing would go nowhere.. :) except for the millions and millions of dollars it got Intel. Now they're trying to gracefully back out. It seems like a sound technical decision. I say good for them.

Increasing number of cores...

[Spoing]

Here's a guess about what Intel is up to. From the brief article -- and keeping in mind this is The Register -- it looks like Intel is going for two targets;

1. Smaller; faster with lower power.
2. Multiple logical and physical cores; multi-threaded apps/OS will do well.

These two basic tagets seem to be a good idea;

1. Processing speed is really damn good now for single tasks.
2. More tasks are becoming standard, so having the 'extras' run without delay on a seperate core makes the system seem more snappy.
3. Some tasks can be CPU intensive and benifit from the extra cores; a reason to upgrade for Intel customers.
4. Speed per-core will increase, though the raw speed in MHZ is costly (in power and because it costs more to make the fab plants) so the fewer fab changes the better or being able to spread the operating life out for more years would be good.
5. With the speed and core # increase, lower-end devices become practical; disable the cores not needed just like other parts were disabled in the past (FPU, cache, ...).
6. Power savings; some cores can be throttled down when hybernating without taking down the whole processor.
7. Multi-processing; if you need the extra umph, plugging another set of cores in might be an easy upgrade or for use in a cluster. (Though CPU speed is not typically the main issue even for many complex problems.)

Another version 4 failure

[yngv]

Chalk it up as yet another product whose fourth version was bloated and disappointing.

Think DOS 4, Netscape 4, IE 4... any others?

Interesting how Windows skipped version 4...

Re:braniac vs. speed demon

[bhtooefr]

About 5GHz is the upper limit for processors running on EXPERIMENTAL technology available currently. The Pentium-M performs as well as the P4, but at MUCH lower clock speeds, which also means lower heat.

The long pipelines allow higher clock speeds (shorter paths for current to flow down) at the expense of Instructions Per Clock (for a very rough estimate of the efficiency of a CPU, multiply clock speed * IPC).

Re:hrm

[Fex303]

Ever wonder about the reason why the fan is sitting there right on top of the processor?

It's like a hat for it, right? So it looks stylish. Otherwise the RAM will laugh at it...

Re:hrm

[Cyno01]

Not smaller fans, slower fans. A smaller fan runnign the same RPM as a large fan will move less aire with the same amount of noise, a large fan can move the same amount of air at much smaller RPMs than the small fan, and in doing so generate less noise. My case has 1 120mm fan and 1 92mm fan, it probably pushes the same amount of air as my friends setup which has 4 80mm case fans, but with substantially less noise.

Re:hrm

[Doomdark]

AMD and Apple really should take advantage of this and do a little "we were right all along" ad.

Hardly. That'd only be relevant for part of geek population, or Apple loyalists, because:

• General populations attention span (half a Friends episode or so) prevents them from even remembering Intel's earlier claims
• People never heard anyone claiming Mhz thing was a myth (it was only used by non-Intel companies, minor players for many computer illiterates); for them "Intel inside" and "this goes to eleven" sales speeches were all the "facts" needed.
• Intel wouldn't have to explain lower clock frequencies, just wait few months for new designs to catch up. It's not like they stopped speeding up frequencies, just that M design _currently_ uses lower than what 4 series was currently designed to use

Any decent marketing department should be able to fairly easily sell change like this. If they anticipated significant trouble, this decision wasn't announced at this point, rest assured. It's not like design decisions for longer-running production lines didn't radically change fairly often. That's their job, to explain and spin it appropriately. And in this case there's enough positive spin to go around. Just imply these are the "wireless chips" (idiotic term, for sure, but only for people who spend few seconds to think about it), and extend from there.