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!

Re:Where are they going to dump them?

[KrisCowboy]

I guess they'll ship them to India or any other developing African country. Better throw them to India, I'll pick 'em up at discount prices. Considering the fact that I've been using a 600MHz Celeron for the past 4 years without any major upgrades (well, just one, added 128MB RAM to my original 64MB ), I should get a biiiiig discount.

Re:Where are they going to dump them?

[Patrik_AKA_RedX]

600Mhz? And you complain? I would give my neighbours right arm for that!

Re:Where are they going to dump them?

[KrisCowboy]

And I'd give my right arm for a better box. I'm not complaining about 600MHz processor. My only problem is that the newer versions of Linux distros fail to detect my vintage hardware. A few days back, Fedora was refusing to detect my 14" monitor. So, I borrowed my friend's monitor and I had to do a lot of "trial and error" configuration to make it work. It finally worked for "Generic LCD display - 1024x768" and it won't work with Generic CRT settings. All I want is new hardware. If Linux can detect my hardware, problem solved. I'd work on my 600MHz for another 10 years :-)

Re:Where are they going to dump them?

[Anonymous Coward]

Hey, I'd give up just about any ARM for something @ 600 MHz.. My PDA would love it..

Re:Where are they going to dump them?

[Anonymous Coward]

Holy shit. Aieee. My linux box won't detect my monitor. I don't know anything about display timing, and I'm too lazy to read..... Whaaa.

Geez, man. There were monitors made *before* hardware detection, live with it.

Re:Where are they going to dump them?

[OxyFrog]

1999 called, they want their jokes about AM... oh.

Re:Where are they going to dump them?

[Thing 1]

"I'd give my right arm to be ambidextrous!" - Crowe T. Robot, MST3K

Re:Where are they going to dump them?

[obeythefist]

There's an alternative. An operating system that will run quite nicely on an old 14" monitor, and it will detect it without fail.

You don't need to browse the net to find and compile drivers for it. You don't need to understand monitor refresh timings or resolution settings to make it just work. You don't even need to do any microconfiguration if you don't want to.

It's called "Microsoft Windows".

And yes it will run on a 600MHz CPU (maybe not well, but it will run).

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?

[Anonymous Coward]

The 4 rules of karmawhoredom:
1. Write a joke and then write "It's Funny, Laugh!".
2. Post an unrelated comment high up in a discussion so everyone will see it and moderate it up.
3. Write "MOD UP" as AC after posting.
4. Ask a question that is explained in the article (guess why this is in bold).

Re:End of an era?

[Anonymous Coward]

I spoke to an Intel guy recently, they have moore's law covered for about the next ten years. If you're prepared to pay. Basically they can produce insanely fast processors, they just can't mass produce them. I suspect this is the case with AMD and IBM as well, it's just a case of getting the fab costs down.

Re:End of an era?

[AviLazar]

When are the processor companies goign to invest in the diamond wafers? - Diamond Age
Now that will be a great way to get that processor speed up, and not worry about the silicon melting.
[Yes I expect this to be modded down for redundency] ;)

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?

[/ASCII]

Moore's law states that the transistor count per dollar on standard sized chips increase exponentially. Seeing as how the cost is factored in to Moores law, your statement doesn't make a lot of sense...

Re:End of an era?

[CKW]

> When are the processor companies goign to invest in the diamond wafers?

Oh, if only the big huge ultra-competition-paranoid hyper advanced ultra-tech companies full of PhD Phycisics and Electrical Engineers would adopt this ground-breaking technology that I a lowly luser know about!!!!! The fools!!! They're missing the boat!! Holding back progress!! How could they possibly be so stupid!!!?!??!?? It's it all so obvious!!!!

Re:End of an era?

[AviLazar]

In the article they talk about why they are not using the technology. Something to the effect of that they haven't finished getting their returns on silicon...So they are holding progress to "recoup" costs on technology that should be legacy material.

Re:End of an era?

[dasmegabyte]

I'd like to know who the grandparent poster thinks is paying for all this diamond research.

If I recall, most of it is taking place in small research companies funded by capital from corporations like Intel, who then would have a percentage stake in the technology if and when it comes around.

There's also plenty of interneal research and doctoral work funded by grants from IBM, Intel, etc.0

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?

[Branc0]

Here is an intersting article just about that.

It's not that we can't technically go on with Moore's Law, it's just that maybe it's not worth it.

Re:End of an era?

[Major_Small]

forget diamond wafers, why haven't they just switched over to quantum computing yet? must be politics...

</sarcasm>

Re:End of an era?

[aminorex]

Given that they are competing with AMD chips
labelled in P4-clock-equivalents, does this mean
that Pentium-M chips will start getting AMD-style
speed ratings? Now *that* would be ironic.

Re:End of an era?

[N1KO]

A few days ago they announced they'll be dropping Mhz for marketing and going with speed ratings instead.

Re:End of an era?

[blixel]

The 4 rules of karmawhoredom:
1. Write a joke and then write "It's Funny, Laugh!".
2. Post an unrelated comment high up in a discussion so everyone will see it and moderate it up.
3. Write "MOD UP" as AC after posting.
4. Ask a question that is explained in the article (guess why this is in bold).

5. Make a list of defamatory remarks.

Re:End of an era?

[VertigoAce]

Either that or a very extensive marketing campaign for the Pentium M. The Pentium M is easily confused with chips like the Pentium 4 M. I have a laptop with a 1.6 GHz Pentium M (not full Centrino since it has an IBM a/b wireless card). An IBM project manager for the T-series Thinkpads compared it to a 2.0-2.4 GHz Pentium 4. Given that it also gets 4+ hours of battery life (year old battery with 80% original capacity) while using wireless network, the system is definitely better than a P4 based system. It's also more expensive, which makes marketing it harder.

Re:End of an era?

[Anonymous Coward]

I put my 1.6 M up against my friends Desktop P4 2.6 in a kernel race and beat it by 30s. First Intel had to recant on the Mhz myth with the Pentium M. Now they are having to do it again because the lowly side project from Israel to beef up a P3 is beating the crap out of the US Engineering that created the Pentium 4 .

Re:End of an era?

[Gilk180]

actually, no. Moore's law has been misquoted and misused in so many places it is a shame.

The cost is not factored in to moore's law. The law pertains to chips with the lowest per-component cost. It has nothing to do with the ratio of cost to transister count.

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).

Re:End of an era?

[hesiod]

> The Pentium M architecture has a relatively high IPC,

Please excuse my ignorance, but what is IPC? Initial Product Cost? Wouldn't make much sense in context, but it's the only acronym of IPC I know. Well, International Pissing Contest too, but that's a politics thing, unrelated to tech.

Re:End of an era?

[DarthTaco]

I could be wrong, but I believe IPC in this context means Instructions Per Clock.

Re:End of an era?

[ball-lightning]

IPC = Instructions Per Cycle. Its the amount of work a CPU can do in a clock cycle. (The higher the IPC, the more efficent the processor, which is how AMD's processors can do the same amount of work with a lower clock speed)

Re:End of an era?

[Anonymous Coward]

I wish all EEs had talent in getting speed up that Intel has. Imagine how nice an Opteron would be at 3.4 GHz?

Re:End of an era?

[sparkywonderchicken]

Actually I prefer Necco wafers

Re:End of an era?

[trentblase]

Agreed... I can only assume that consumers caught on to the "frequency is everything" marketing tricks. However, ultimately I don't give a rat's ass about MHz or CPI. I want IPS (instructions per second).

Re:End of an era?

[Anonymous Coward]

You must not know what GOOGLE stands for either.

Re:End of an era?

[Citizen of Earth]

If this is true, Intel is abandoning the whole direction they took with the P4 = MHz at all cost.

I guess they were hoping to blow AMD away by labelling their chips with big numbers, but then AMD matched them by using virtual big numbers, eliminating Intel's marketing advantage. However, since AMD's big numbers are only used by the marketing department, they aren't caught up in the "at all costs" problem that the engineering department at Intel finds itself in.

Re:End of an era?

[hesiod]

> You must not know what GOOGLE stands for either.

Umm. Short for a Googolplex, inferring that they index a whole lot of pages? Or maybe it's Get On Our Great Life Engine.

Re:End of an era?

[Anonymous Coward]

I'm guessing you must be Jewish or something? What a nice contradictory post:

The Pentium-M is a modded Pentium-3.
Quite right. It's a Pentium 3, but with SSE2, the Pentium 4's bus, more cache and improved branch prediction. "Dothan" adds a couple of extra features on top of that.

Pentium-M was designed in Israel
Yes, but only insofar as Intel Israel was handed the P6 core and told "keep up to date, please". Moreover, the physical design (i.e. the main reason it runs cool and at a relatively high clockspeed for its complexity) was done in the US.

Re:End of an era?

[Anonymous Coward]

aboot the same as an intel P4 at 6GHz

Re:End of an era?

[wyohman]

Repeat until it sinks in, "There is no such thing as Moore's Law. There is no such thing as Moore's Law."

Re:End of an era?

[Anonymous Coward]

You're right, both IPC and MHZ matter- and the P4 design was successful if you look at it only in terms of IPC and MHz-

If you are an EE, then you would understand that increasing the pipeline depth isn't only a "trick" to get MHz up, but the only way to increase the clock speed for a given manufacturing process. Even if the IPC goes down, if the MHZ goes up by more than the IPC goes down, then overall performance will increase, and that is exactly what happened with the P4. (intel's 0.18 um process was stuck at 1.1 GHZ for the P3, and went up to 2ghz with the P4 using the same manufacturing technology-and IPC didn't decrease by as much as clock speed went up) When i buy a processor, what i care about is the actual performance, not the IPC or the clock speed of the processor.

It is true that the average consumer will look at MHz when buying, an added bonus for intel's archetecture decision, but the bottom line is that the P4 does deliver in terms of performance.

(and any moral high ground that AMD fan boys like to claim is negated by the fact that AMD doesn't even list the clock speed, and names their processors based on some arbritary performance rating)

What's wrong with the P4 archetecture is the fact that it can't scale up as high as originally believed- for the last 20 years the design paradigm was to shrink transistors and add MHz, with power consumption increasing from the milliwatt range to 100 watt in modern processors. At the time of the P4 design (probably starting more than 5 years ago), the engineers probably believed this trend could continue, and roadmaps predicted that the P4 archetecture would scale to beyond 10GHz.

Of course, procesor designers finally realized that power consumption can't continue increasing (someone claculated that in 5-10 years, the die temperature would reach the temperature of the sun if they continued to increase at the current rate) and that the 100watts of today's designs is about the most that can be practically achived in a single die. As a result, there has been a change in the direction of processor design, towards multiple cores instead of more complex cores (more surface area to remove heat) and greater efficency (performance per watt)

As you should know, power consumption is proportional to clock speed, so making a 10GHz, long pipeline processor not feasible from a power consumption standpoint, not because of IPC. The P4 design is still very competitive at this stage, (can be considered the "best", depending on what applicaitons you run) but the archetecture won't be able to compete 2 years down the road.

What's happening now is happening across the industry- Intel is planning on taking laptop processors and putting them in desktops, (exactly the opposite of today's trend) AMD is spending much more R&D dollars in their laptop processors (obviously probably planning the same thing), and virtually everyone is working on multiple core, low power designs.

What IS ironic is that the Pentium M is based on the pentium 3 core, which is itself based on the pentium pro, a processor widely critized for its low IPC when it was released 10 years ago.

Lemme guess...

[Predathar]

New socket meaning new motherboard right ? I miss the days when you were able to use the same board for more than 1.5 years....

Oh well..time to read the article.

Re:Lemme guess...

[Anonymous Coward]

And when was that?

Re:Lemme guess...

[tomstdenis]

Socket/Super 7 boards.... I recall having an MI, MII and a K6-2 350 in the same board.

Of course that would require both Intel and AMD to sit down and design some Socket1000 board or something. But that gets trickier cuz many of the pin [in Socket478 for instance] are grounds and power. IIRC there are 166 pins dedicated to power management. So the layout of the actual processor would be dictated somewhat by the location of power.

But it would be nice to be able to take out an Intel core and slap in an AMD core in the same motherboard...

Tom

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:Lemme guess...

[dasmegabyte]

I don't remember these hypothetical old days!

My first PC was a 386-40. Two years later replaced by a 486-66 (new motherboard needed). Two years later replaced by a 486-120 (new motherboard needed). Next up was a Celeron 166 (new motherboard again), then a K6-233 (new motherboard for EDO ram) and a K6-2 350 after that (new motherboard to support the higher multiplier). I went intel for a little while, dual celeron, and after that a p3. New motherboards for each. Finally, I settled down with a top of the line Athlon thunderbird 1 gig on a top of the line motherboard.

A motherboard which only supports chips up to 1.4 GHz. And whose top-of-the-line clock makes at 200 MHz.

*SIGH*. Each chip upgrade, usually performed at 2 year intervals, has required a new motherboard to take advantage of a higher clock speed or "better" memory or a new graphics system or USB. In fact, i'm beginning to thing that I'd be better served by a PC industry that integrated chips onto motherboards to save costs and improve cooling options. Over and under fans? You got it!

Re:Lemme guess...

[default luser]

Sounds like you need to learn to plan ahead, or buy a little more power at a time. I find I only need to upgrade every 3 years or so to stay powerful, and I plan my upgrades to be "upgradable" whenever possible.

Let's look at your list:

My first PC was a 386-40
You went cheap by buying the AMD 386 rather than a low-end 486. If you had invested in the 486, you could have done the DX2-66 upgrade without a motherboard change.

Next up was a Pentium 166 (new motherboard again), then a K6-233 (new motherboard for EDO ram)
Whose fault is that? You certainly didn't see monsterous performance improvements going from FPM to EDO, and you probably could have reused the P166 board. Don't blame your own thirst for performance on the industry, you had a choice.

dual celeron, and after that a p3. New motherboards for each.
You've got to be kidding me. Most BX boards could be upgraded to Coppermines (including the very popular BP6 dual Celeron board). If it couldn't support the low voltages, you could always pick up the Socket 370 PGA to FCPGA adapter from Powerleap. You just bought a new board because you were lazy, and now you're whining about it here. Hell, even my BH6 Revision 1.0 could run Coppermine processors (even though Abit officially said it could not), saved me a ton on upgrade costs.

a top of the line Athlon thunderbird 1 gig on a top of the line motherboard. A motherboard which only supports chips up to 1.4 GHz. And whose top-of-the-line clock makes at 200 MHz
Yes, but assuming you're not full of shit when you claim this system to have been "top of the line" at time of purchase (circa 2000), I'll bet you've liked not having to upgrade for over four years. Face it, you finally got your money's worth out of a system, and you're still whining.

My current system is an Athlon XP 1600+, and at the time it was nearly top-of-the-line, and for that investment I have had over three years of good performance. I expect to upgrade soon, and I do expect to have to buy a new motherboard...but I will expect this next upgrade to last me 3-4 years as well.

The point is this: I have used incremental upgrades (no motherboard change) and complete upgrades (new motherboard) to my advantage. Both paths have their benefits, but if you plan ahead and, research your options come upgrade time, you should be happy with either.

Re:Lemme guess...

[EvilTwinSkippy]

I was going to say something snide, until I thought of the bone-yard in my basement of discarded RAM chips, Mobos, and processors.

Fortunately they don't sit around. My computer's previous mobo, processor, and ram power my webserver in the basement, and the board, processor, and ram before that is running my Mom's home computer.

Trickle-down technology at it's finest. Or why I always buy white-box.

No more desktops

[fredistheking]

So they are only gonna support mobile-computing now?

-

Re:No more desktops

[Rick.C]

Time to RTFA.

There will be laptop and desktop versions. The big deal is that they're putting two or four lower power CPUs in one package with shared L2 cache.

Re:No more desktops

[hyc]

What makes you think that? Why wouldn't you want to put a Pentium-M into a desktop machine?

Re:No more desktops

[Anonymous Coward]

Nope. The P-M is significantly faster per-clock than the P4 Northwood or Prescott, and a hell of a lot cooler (thermally) as well. It's a win all around to put one in a desktop.

Finally...

[LoganTeamX]

Someone with half a brain in charge of development at Intel has stepped forward. Their mobile processors are all that's saving them for the non-enterprise arena. But oh no, they'll need to abandon the GHz race! Whatever will the fanboys do!?!?

Re:Finally...

[sjames]

But oh no, they'll need to abandon the GHz race! Whatever will the fanboys do!?!?

Perhaps pin count will be the new benchmarketing standard. I guess we'll need to have sockets with more pinholes in a lid. 'Ours is better because it has 4000 pins!'

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 (?)

Re:Good Idea

[Anonymous Coward]

Dont go an get the Ultra-Gigaherz-Processor but a descend,processor

The word you were aiming for is decent. I also have no idea why you put a comma behind it. Learn to spell, dude!

Re:Good Idea

[tedgyz]

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 (?)

I agree. I care more about building a silent, but capable PC than than a blazing overclocked hot rod.

Re:Good Idea

[Joel Carr]

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
I agree, and it will have the added advantage that no one will be able to run Longhorn if/when it is finally released.
Projected 'Average' Longhorn System Is A Whopper

---

It's good that they didn't call this pentium 5

[foidulus]

Because the name pentium came as the successor to the 80486,(ie 5, thus pentium) which was the successor to the 386, 286,186,8086. Thus the 5th generation of the 5th generation chip would have been kind of dumb. I think they should just abandon the pentium name all together, but by this point it has too much name recognition.

Re:It's good that they didn't call this pentium 5

[hyc]

Not just "name recognition", which implies a nebulous marketing benefit. "Pentium" is a registered trademark, and trademarks have to be used to stay valid. And there's no way they'd abandon any intellectual property, not even a name...

Re:It's good that they didn't call this pentium 5

[Ruie]

You have a good point.

Does this mean that marketing folks at Intel can only count using one hand ? :)

Re:It's good that they didn't call this pentium 5

[espo812]

Thus the 5th generation of the 5th generation chip would have been kind of dumb.

What's wrong with the fifth release of the fifth release? Sendmail version 5.5, PGP 5.5, Internet Explorer version 5.5, AOL Instant Messenger version 5.5.

Re:It's good that they didn't call this pentium 5

[rusty0101]

Well, when they come out with the Septium 7 chip, perhaps it will answer questions before they are asked.

Of course coming from Intel, one in 7 million responses will be wrong, there will be an instruction added to later dies that corrects for that and all binaries will have to be re-compiled to take advantage of that new instruction. The new instruction will cause two other errors to crop up in more common responses.

Re:It's good that they didn't call this pentium 5

[Anonymous Coward]

Seventh son of a Seventh son?

Re:It's good that they didn't call this pentium 5

[Alioth]

Bah. I was hoping they'd follow the 'increment the number with each generation' as they continued (80386, 80486, 80586, 80686) with the names - so the 586 gets called the Pentium, and the 686 gets called the SEXium!

Re:It's good that they didn't call this pentium 5

[powerlord]

Well ... Haven't all those products contained expolitable bugs in them? Therefor all 5.5 products must be flawed, so the Pentium 5 would be flawed.

What? ...
ALL products contain bugs? ...
Oh well ... guess I'll write my own software from now on. That way I'm SURE its bug free.

=P
(removing tongue from cheek)

Re:It's good that they didn't call this pentium 5

[donnyspi]

Correct prefixes: Pent Hex Hept Oct Non Dec Undec Dodec

Re:It's good that they didn't call this pentium 5

[Louis Guerin]

Nah, Septium ain't gonna be worth nothing, it's the Sexium chips which will really take off...

L

Re:It's good that they didn't call this pentium 5

[Short Circuit]

What are you talking about? The i686 architecture has proven dead sexy over the years!

Re:It's good that they didn't call this pentium 5

[Dun Malg]

trademarks have to be used to stay valid.

Actually, they only have to be defended. I can name my own version of vinyl flooring Parpi-Klag Flooring, register "Parpi-Klag" as a trademark and, so long as I defend against others using the term Parpi-Klag to describe their product, the trademark remains mine. I don't have to sell a single square centimeter of vinyl flooring, nor even produce it. I can just dream about it and have just the trademark alone.

Re:It's good that they didn't call this pentium 5

[bhtooefr]

Sept(ic)ium: 786 (NetBurst)

Re:It's good that they didn't call this pentium 5

[hyc]

Have you ever registered a trademark? (Apparently not.) I have. You have to provide proof of actual interstate commerce before the mark will be granted. You also have to provide renewed proof after 7 years, otherwise the mark is abandoned.

Re:It's good that they didn't call this pentium 5

[LPetrazickis]

Uhm, since 5.0 is the first release, then 5.5 is the sixth release and 5.4 is the actual fifth release.;)

Oh, and the millennium changed over in 2001.^-^

Re:It's good that they didn't call this pentium 5

[Dun Malg]

Have you ever registered a trademark? (Apparently not.) I have. You have to provide proof of actual interstate commerce before the mark will be granted. You also have to provide renewed proof after 7 years, otherwise the mark is abandoned.

These conditions are outside the scope of the argument. Intel has, I'm sure, engaged in interstate commerce once or twice and could probably find some proof of this.

Re:It's good that they didn't call this pentium 5

[timmi]

Actually the whole reason they started on the whole "Pentium" moniker was because they simply couldn't x86, and it's derivitives.

Pentium mm

[Mad Man]

due to arrive in 2005

Shouldn't that be Pentium MMV?

Re:Pentium mm

[killmenow]

Shouldn't that be Pentium MMV?

I get the roman numerals thing, but I couldn't help thinking of a different meaning for MMV. So...

Shouldn't that be Pentium YMMV?

Re:Pentium mm

[mrogers]

Your megahertz may vary?

Re:Pentium mm

[slide-rule]

> Your megahertz may vary?

Almost... its "Your Math May Vary". ;-)

Re:Pentium mm

[NoMercy]

Pentium YMCA?

Re:Pentium mm

[SpryGuy]

Could be worse... they could come out with the Pentium DMV

Instructions have to queue up forever before they're executed. Talk about long pipelines!

Re:Pentium mm

[trukfixer]

Sounds good.. I think they considered calling it PMS at one time but.. Windoze already HAS that technology..

Re:Pentium mm

[Anonymous Coward]

Could be more than a little confusion when Intel releases its chips for 2010...

Re:Pentium mm

[smithmc]

due to arrive in 2005 - Shouldn't that be Pentium MMV?

If so, then the Pentium MMX was way before its time.

Re:Pentium mm

[JonLatane]

Or, if they want it to coincide with Longhorn's release, Pentium MMC.

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.

Re:Power consumption is important

[alienw]

the head produced could just vaporize the silicon

The Pentium may be a great processor, but I do not believe it can give head, despite its greatness.

Re:Power consumption is important

[drinkypoo]

Are they still using CRTs? While CRTs are still superior to flat panels in many respects, they certainly draw a lot more power.

Re:Power consumption is important

[arakon]

You mean all this time I've spent sweet-talking my computer has been wasted?

Re:Power consumption is important

[Anonymous Coward]

Go back to Physics 101. CMOS power is roughly proportional to C*V^2*f, until leakage starts becoming significant.

Re:Power consumption is important

[addaon]

Mod parent up. Frequency is NOT squared, voltage IS.

Re:Power consumption is important

[yngv]

Now be fair.. that's at least Physics 201. In 101, we learned things like Fudd's Law of Opposition:

If you push something hard enough, it will fall over.

Re:Power consumption is important

[Anonymous Coward]

get an mini itx from via for DTP, file and webserver. Would save some $$$

Re:Power consumption is important

[hkfczrqj]

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.

Altough not Pentium related, just imagine the efforts needed to power and cool down the BlueGene/L at L.Livermore. They have to cool down 131,072 CPUs! IIRC, BlueGene will use 1.5 MW, and should work at 0.5 Petaflops (as one guy from LLNL told me).

Re:Power consumption is important

[WuphonsReach]

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.

Which is why I think this is a good move for Intel. There's a lot more interest in the home market for small form factor and/or noiseless PCs then there was a few years ago. Probably mostly because computers have gotten 10x better at multimedia. Back when all your PC could do was crunch numbers, noise didn't really matter a lot (and older PCs were pretty quiet since they didn't have large power/heat issues). Now, it's a lot more common to see PCs being used in home theatre setups, or used as a simple stereo system playing MP3s or videos. IOW, uses where noise matters.

Looking back at my power bills from 2002, I used a measly 500 KWH on an average month. That creeped up to 750 KWH in 2003 and is now at 1000 KWH in 2004. Not to mention the amount of heat and noise that all that equipment generates in my office. So I'm quite interested in systems that are low-power / quiet yet fast enough to do what I want done. (I just built a mini-ITX gentoo server...)

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.

Re:Change of ideas

[Anonymous Coward]

P4 not PIV.

Re:Change of ideas

[4of12]

I think this may just help Intel get some more umph into their line, before 64-bit hits critical mass.

Well, the delays in native Windows for 64 bit to the end of 2004 have certainly helped slow the 64-bit bandwagon. Probably long enough for Intel to recover from the mistake it made in betting heavily on the Itanium for 64 bit.

But today's price cuts on Opterons will undoubtably increase AMD's sales at a time when Intel only has the 64 bit Itanium that lacks high-performance IA32 emulation.

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

[makapuf]

'Pentium IV' ? Ah, that sounds SO 2005 ...

Re:No Suprise

[rudeboy1]

I would tend to agree that Intel needs to concentrate on lower voltage chips. Lower voltage=lower heat. I have a P4 3GHz 800FSB chip running in my room, and if I leave it on all day with my door closed, I will come in at night and be welcomed by no less than a 5-10 degree difference in temperature from the rest of my apartment. Didn't have that problem with my P3 500Mhz. I am not going to be in the market for a new chip for a while, but someday I hope to have a computer that doesn't affect my home heating/cooling budget.

Re:No Suprise

[steveb964]

'Pentium IV' ? Ah, that sounds SO 2005 ...

ahh... people would pronounce it as 'Pentium i-vee', which would make me grit my teeth just as much when I hear people tell me that they have Windows 97 or Office 92 as an OS.

/rant

Re:No Suprise

[Jeff DeMaagd]

What's funny is that the next Pentium-M iteration will have 2MB l2 cache - same as that of the top of the Xeon MP line.

And if PM will get the 64 bit extentions, then it would be a terrific 64 bit platform, IMO, consuming only 70% of the power of the AMD Athlon 64 mobile chip with about as much CPU power.

I guess this means that there is an interesting CPU race again.

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.

Re:No Suprise

[drinkypoo]

You are assuming that the 64 bit extensions will not consume sufficient real estate to bring the power consumption of the Pentium M up considerably. Also AMD has just brought out their own low-power processors, which come with either 1MB or 2MB of L2 cache. The ~1.7GHz model comes with 1MB and consumes only 35W. Current Pentium M processors attain around the same clock rate and consume about 27W, though if these AMD processors do more per cycle than their intel-designed counterparts as we have come to expect from AMD, the additional power consumption should be worth it to nearly everyone. (Except people buying laptops with transmeta processors to keep their battery life up - they're obviously not interested in number crunching power.)

Re:No Suprise

[PeDRoRist]

With Pentium Ms (or Athlon 64s), you wouldn't have that problem, since they can run slower (ergo cooler) when idle, or performing low cpu time consuming tasks.

My A64 runs at 800Mhz when i'm surfing or reading my mail (although it used to stay at 800Mhz when I fired up Photoshop with the early version of the processor driver, which was quite annoying - fixed now)

Re:No Suprise

[servognome]

The Athlon 64 mobile 3000+ runs at 65W, the low power 35W A64-M is the 2700+ which has only 512kb of L2 cache, vs the 1MB in the 2800+ and higher. Until benchmarks come out on the 2700+ its apples and oranges comparison

Re:No Suprise

[servognome]

From this article The low power (35W) AMD64-M only has 512kb of L2 Cache, the 65W versions have 1MB.
The other models - 2800+, 3000+, and 3200+ are rated to operate at 65W while the 2700+ is rated to operate at 35W.
The 2700+ processor has 512KB cache while the rest of the models house 1MB cache.

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?

[the quick brown fox]

And VIA's EPIA processors are all named after Old Testament prophets: Samuel, Ezra, Nehemiah...

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.

Re:Religious Nomenclature?

[Anonymous Coward]

The darn chip was designed in Israel, so no duh! there are religious reiterations.

Re:Religious Nomenclature?

[Anonymous Coward]

Lets see... Dothan... A historical site in ancient Israel, Johan, A historical figure. Merom, a lake in Northern Palestine.
And this means they have religion. Yip yip yip yip yahoo. Apart from the fact that Intel is a private company and is entitled to do as it pleases, it should not surprise anyone that a chip designed in Israel would be coded after significant places and events in Israel and Jewish history.

Re:Religious Nomenclature?

[gooberguy]

Via does the same thing with their chips. The C3 has had cores codenamed after biblical figures such as Samuel, Eden, Esther, and Nehemiah. I think they use codenames from the bible because they can't be copyrighted.

Re:Religious Nomenclature?

[Anonymous Coward]

It could have been named after Dothan, alabama... there's a military base there... so its a militaristic, imperialist name, as well as zionist. 2 for 1 deal.

Re:Religious Nomenclature?

[Anonymous Coward]

you do realize the intel fab in israel was built on stolen palestinian land? they bulldozed the village that was in the way and drove off all the people in the area.

they made a ton of p4 cpus, mobile cpus, core logic (chipsets) at that fab.

read about the ghastly story at: http://www.inminds.co.uk/boycott-intel.html ....

Re:Religious Nomenclature?

[Anonymous Coward]

Palestinians are dogs. They have no right to property or even life. To suggest that highly advanced microprocessor fabrication is less important than the homes of some savages is clearly absurd. Remember: arabs are NOT human.

Re:Religious Nomenclature?

[wirelessbuzzers]

And VIA's EPIA processors are all named after Old Testament prophets: Samuel, Ezra, Nehemiah...

Properly, VIA's processor cores are named after Old Testament books which are named after people. That's why there was a I Samuel and a II Samuel core. (Dunno why no Joshua, though).

The next core to come out is Esther, who wasn't properly a prophet(ess), but rather a queen.

As a side-note, the preliminary specs on Esther are quite impressive: lower transistor count than Nehemiah (= cheap, small die), 90nm, 2GHz @ 5W (fewer IPC than a Pentium though), integrated RNG, AES, SHA-1... looking forward to using one of these in a quiet, low-power personal server and firewall.

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.

Re:Faster Pentium M?

[Groo Wanderer]

No, the ramp of Dothan has not been all that spectacular, it was set to around October of last year, and it is only coming out. There were several speed path issues that were resolved, and it is due out on the 10th.

That said, it is still not what was promised last year, power consumption is higher, no new FSB as promised, and other problems. When it hits what was promised, it will probably be a year late.

-Charlie

Re:Faster Pentium M?

[platos_beard]

I think the point is that ramping up speeds is no longer the priority it was. Performance gains will come from including muliple cores on one chip instead of ever higher clock speeds (and the power consumption, cooling requirements, and noise that come with them).

Since having many processes and threads is the rule in today's computers (I have, let me check, 46 processes and 466 threads going at the moment, and I ain't really doing anything), this seems like a pretty good plan to me.

Re:Faster Pentium M?

[10Ghz]

Well, if they used similar heatsinks/fans on Pentium-M as they do on regural P4, I bet they could increase the clock-speed to about 2.2-2.4GHz right now (and that that speed it would wipe the floor with even fastest P4's). Currently P-M runs at sub 2GHz-speeds because they are all in laptops, where you can't use massive cooling-solutions.

Re:Faster Pentium M?

[Anonymous Coward]

truthfully i dont think they need to ramp up the clock speed too much to get it to the current p4 performance... my laptop is a 1.3ghz pentium M.. 1.3 ghz most people would think "what? computers 3 years ago were faster.." but its actually about as fast and faster for somethings as my athlon xp 2200+ desktop here. the pentium M is a great design compared to the p4, has alot of L2 cache on chip and is paired with an almost perfectly tuned chipset. i like AMD but the pentium M was something i couldnt resist. the 2ghz pM has to be about as fast or a little slower than a 3ghz p4

Re:Faster Pentium M?

[CuriHP]

That may or may not be true. You are assuming the only problem with increasing the speed is heat dissapation. It's not. If the design is just barely making timing closure at it's current frequency, it may very well be impossible to increase the clock frequency without a redesign of the critical paths.

well...

[Hangin10]

We've been on 32bit chips for quite some time..
Is 32bits enough? Is that why 64bit chips don't
seem to be catching on? or does the fact that
AMD and Intel seem to have fairly different
workings to their interface (AMD's seems fairly
simple, I haven't looked at Intel's).

Slightly related,
It seems both Intel and AMD stopped shipping free
copies of their Architecture Manuals. :(

Re:well...

[sweet cunny muffin]

32 bits only allows 4 GB of RAM, which will be a real problem soon. 64 bits would allow 16777216 TBs.

Re:well...

[dead sun]

Meh, there are plenty of hacks (at the cost of a single extra cycle of latency) that extend memory access out to 48 bits. The problem being you can only have 4 GB pages, but for many, many applications it doesn't matter. For Joe Consumer who maybe, possibly, has 1 GB of RAM right now it's not a big deal. Heck, I haven't seen more than that preloaded into a computer at a retail outlet yet.

Sure, the people that need 64 bit memory access will welcome cheap 64 bit CPUs, but it isn't like 64 bit CPUs have been difficult to find in the past. The availability of cheap 64 bit CPUs will likely hasten their adoption so it isn't a big problem when we really need the extra memory, but that problem is still a good way off for most of us.

Re:well...

[alienw]

Don't seem to be catching on? Opterons are HOT in the server market, and the reason we don't see too many desktop 64-bit systems is because they are more expensive and Microsoft doesn't have 64-bit windows yet.

Re:well...

[skajake]

I think its because the cursed industry simply refuses to look at AMD as a serious contender. My CS professor got all giddy talking about Intel's forray into the 64bit realm with its new extensions. I couldn't beleive he did not even mention that it was spurred on by the Athlon64.

Sight, i guess until Intel starts pushing the boat just won't move.

AMD

[kpogoda]

Looks like the recent AMD press and popularity has forced Intel to rethink its business strategy. Here is my most recent ocrrespondence with Dell Sales Support. You should find it amusing: Problem Description: I am in the market for an AMD machine. I have been browsing your website but can't find an AMD processor-based machine. Do you sell any AMD machines? If not, I will shop elsewhere. Thank You. Dear Valued Customer, Thank you for choosing Dell Online Consumer Customer Care. I apologize for the inconvenience caused with regard to this issue. I have looked through your e-mail and show that currently Dell is not offering AMD machines. I have forwarded your message to management and I assure you they will look into this issue and will work on making improvements based on your feedback. Once again, I apologize and truly regret any frustration this matter may have caused. Thank you for your patience and understanding. They are both greatly appreciated. If you have any further questions or concerns, please visit the following website to contact us. www.DellCustomerCare.com Respectfully, Alexander ~DTC41593 7:00 AM to 3:30 PM CST Mon to Fri Dell's Online Consumer Customer Care

Re:AMD

[paitre]

Don't you just -love- canned responses?
Even though slightly modified to be a "direct" reply to your "problem", it's still a canned response.
'sides, Dell isn't going to offer AMD systems until they're forced to by -corporate- market pressures. It's happening, it's just slow (and they know they lost a decent sized purchase I just made because of it).

Re:AMD

[ryanvm]

Was the subject of your post "AMD (Score:5, Funny)" or did somebody really find that funny?

Re:AMD

[NemosomeN]

Apparently someone found it funny... (Obviously not me, or it would be back at +4 Interesting now)

Re:AMD

[Anonymous Coward]

Verdamnt! It won't let you put (Score:5, Funny) in the subject line, it automatically takes it out.

Re:AMD

[Spoing]

1. Do you sell any AMD machines? If not, I will shop elsewhere.

Is there a technical reason you want a specific brand of processor?

To me, the CPU brand is becoming much less important. The supporting chipset (features and quality), memory expansion, and system maker tend to sway me much more these days.

Re:AMD

[Anonymous Coward]

Hey, you're right. :O

Re:AMD

[TheLink]

Yeah. Maybe there have been one or two searches for opteron servers on Dell's search page :). Anyone looking for an Opteron version of the Poweredge 1600SC?

But their hands are probably tied. Rumour is that Dell has committed to buying USD5 billion of Intel stuff. I suppose that's how Dell gets real cheap Intel stuff? Now I'm wondering if Dell has a "get out" clause (they should if they are sane) somewhere, and if it does, what it is and whether it is close to applying... Watch Intel and Dell closely to see who is squirming the most, and perhaps you might figure more out.

It's worth supporting AMD just to watch the Intel and Dell show. Bwahahaha.

Still I'm sure Intel will manage to turn things around. Sure looks like they've the stomach to make the hard decisions based on technical stuff when it comes down to the crunch (plus plenty of reserve belly fat). Of course it took them a while (judging from the recent presentation by the ex-Intel chap) but there's plenty of inertia/momentum involved when making chips esp when you've been doing things well the past X years. Intel can afford to make a mistake or two every now and then, as long as it corrects them eventually.

Not sure about the Itanic though - my guess is it'll remain one of the fringe chips. If Intel doesn't make a good server class "Pentium M 64", then AMD is going to take that market (and Dell is in for a rough ride). If Intel does make a good AMD64 chip, I don't see that many people flocking to the Itanic. Heh.

If Intel screwed up/miscalculated[1] and can only launch a decent competitor in 2005+, Dell's competitors can take significant market share IF they play the Opteron card well. But which x86 server maker wants to piss Intel off by playing the Opteron card and which can actually pull it off? Sun? IBM? HP??

[1] Looks like Intel's 64bit extensions aren't 100% AMD64 compatible. That might be intentional, and not a problem in itself. The problem is if Microsoft insists on some things that Intel has left out (e.g. the NX stuff). Chips take some time to be fixed, tested etc ...

Re:AMD

[ChaoticLimbs]

Okay here's a better question- Do you have any computers with 64 bit architecture ala x86-64? In particular do you have any computer which can accept 6 GB of onboard DDR RAM?

Re:AMD

[Anonymous Coward]

Not sure about the Itanic though - my guess is it'll remain one of the fringe chips. If Intel doesn't make a good server class "Pentium M 64", then AMD is going to take that market (and Dell is in for a rough ride). If Intel does make a good AMD64 chip, I don't see that many people flocking to the Itanic. Heh.

Er... you mean Itanium? Or were you trying to draw a parallel between the Titantic and the Itanium?

The problem with the Itanium is that it's not x86 compatible. Which, on paper looks like a smart gutsy move because you're not limited by the limitations of the x86 platform. But out in the market, it doesn't sell real well because there's no "comfort" factor. (If I'm going to switch to something that's not x86 compatible... why not also look at Alpha / Motorola / SGI / Cray / IBM / Apple.)

AMD Opteron gets it right because it's a 64bit chip, that's still x86 compatible, *and* it's 32bit performance is on par with the existing 32bit chips. That means that there's little to no penalty for using the Opteron even if you're still using a 32bit O/S.

Which really takes advantage of the "comfort" factor when a company is trying to spec out what CPU to use on their next server.

braniac vs. speed demon

[the quick brown fox]

Interesting how P4 Northwood did so well against Athlon XP, mainly because of Intel's "speed demon" approach (trading IPC for higher clock speeds). Whereas Pentium-M brings them back in line with the kinds of IPC we see from AMD and IBM.

I won't pretend I know anything more about CPU design, maybe someone who does can comment on whether this means Intel made a bad bet with the long pipelines or whether there's some other reason why P-M holds more promise than P4...?

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: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).

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.

Intel is so far behind anyway

[Rosco P. Coltrane]

1979 - 86 (186 existed too but was a failure)
1982 - 286
1986 - 386
1989 - 486
1993 thru 2004 - Pentium (meaning 5-something), with a sub-version number

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

Re:Intel is so far behind anyway

[CrazyTalk]

Don't forget, when the Pentium first came out it was refered to unoffically as the 586, and the PII was refered to in the press as the 686, so even if intel had to change their naming scheme (they couldn't copyright a number like 80586, so they had to pick a name for it) people still thought of those chips under the old terms for quite some time. Now, of course, with the product line so divergent into different sub-classifications, it really doesnt make sense to map to the old numbering scheme.

Re:Intel is so far behind anyway

[vondo]

The Pentium Pro (pre-Pentium II) is the 686. The Pentium 4 is different enough from the P3 that it's really a new processor too. They stick with Pentium because it's a very well established brand name.

The Athlon in a K7, Opteron and Athlon64 are K8, but AMD isn't calling it the Octathalon either.

Re:Intel is so far behind anyway

[nelsonal]

The pentium was replaced by the pentium pro which was the die basis for the Pentium II and Pentium III. The pentium 4 was an entirely new architecture that was designed around the philosiphy of attacking big problems with lots of rapid teaspoons. The pentium M is another new architecture that was designed around the goal of power constraints. By your rendering these chips would be in the Octium family (P2 & P3 hexium P4 Heptium) but why screw up a good brand name.

Re:Intel is so far behind anyway

[glwtta]

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

Somewhere an Intel marketing exec just spastically grabbed his desk and ejaculated violently.

Re:Intel is so far behind anyway

[elmartinos]

And where are AMD's Biathlon and Triathlon?

Re:Intel is so far behind anyway

[DarthTaco]

"Don't forget, when the Pentium first came out it was refered to unoffically as the 586"

I always heard that they called it a pentium instead of 586 because when they powered up the new processor and added 100 to 486, they got 585.999871.

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:Intel is so far behind anyway

[JDevers]

Actually, the Pentium M is NOT a new architecture it is a Pentium III all dressed up for war. More or less it is a P3 with the P4's branch prediction unit, an ever so slightly longer pipeline, and a few other niceties from more modern processor designs.

See: http://arstechnica.com/cpu/004/pentium-m/pentium-m -1.html for a somewhat technical discussion...

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.

Re:Intel is so far behind anyway

[Anonymous Coward]

Actually, the Pentium Pro (remember that?) was referred to as the 686.

Re:Intel is so far behind anyway

[yarisbandit]

But the classily-named sextium processors have been in the works since april '99... What's the delay? ;)

Re:Intel is so far behind anyway

[3770]

With the 86, you mean the 8086, and that was actually a failure too I think. Back then the IBM PC's and XT's used the 8088 if memory serves me. I think the difference was the width of the memory bus, being 8 instead of 16 bits.

Re:Intel is so far behind anyway

[drinkypoo]

I'm trying to envision what you would do in an octathalon... Run, bike, swim, ski, shoot, ride, fence, and screw? (The last one would really spice up the olympic games, summer or winter.)

Re:Intel is so far behind anyway

[bstone]

80186 was not a failure. I just was not used in PC's.

Don't tell me that. My 1st PC was a TRS80 with a 186 processor.

Re:Intel is so far behind anyway

[Anonymous Coward]

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

That is an oversimplications that I don't think you understand. Most likely you have read Slashdot and a few other PC/enthusiast sites but not having real knowledge of the fact.

Re:Intel is so far behind anyway

[maw]

They definitely couldn't copyright a number like 80586. It's even possible that they couldn't trademark it.

Re:Intel is so far behind anyway

[jkovach]

The 80186 still isn't dead - you can still buy them. I have a 186 single board computer sitting in a box on my desk that I assembled and programmed last year for a microprocessors course. The chips are still sold by DigiKey and the other usual suspects. I have a copy of the 186 datasheet here that actually says "updated June 2002" on it. Old hardware never dies...

Re:Intel is so far behind anyway

[Anonymous Coward]

yeah, i always confuse those two. (Fortunately, IANAL!)

Re:Intel is so far behind anyway

[Anonymous Coward]

186 wasnt a failure it was popular is washing machiens ect

Re:Intel is so far behind anyway

[Cyno01]

The PIII is simply a better chip than the P4, servers still ship with PIIIs in em. Running at the same MHz a PIII will outperform a P4 by quite a bit IIRC.

Re:Intel is so far behind anyway

[green pizza]

>> 80186 was not a failure. I just was not used in PC's.
>Don't tell me that. My 1st PC was a TRS80 with a 186 processor.

Umm, the "80" in "TRS-80" refers to its Zilog Z80 processor.

Re:Intel is so far behind anyway

[troon]

The RM Nimbus educational PC used a '186 clone as its main CPU. You could argue it was borderline PC-compatible, but it did run a modified version of Windows.

Re:Intel is so far behind anyway

[socode]

It was in an RM Nimbus I used to use...

Re:Intel is so far behind anyway

[Anonymous Coward]

Intel stopped focusing on the Pentium because the engineers there are to busy smoking Opium.

Re:Intel is so far behind anyway

[Detritus]

See here, it was one of the few PCs that used the 80186.

Re:Intel is so far behind anyway

[RhettLivingston]

Have to second this one. If you look at the 80186 in terms of the total # of different designs that used it, it may very well have been their most successful architecture. They tried to kill it once with an embedded systems version of the 80386, but didn't come close. They ended up bringing the '186 line back after that and actually expanding on it. I was involved with at least 25 different custom Multibus II board designs through the 90's that used a '186 based core.

Re:Intel is so far behind anyway

[Anonymous Coward]

Come on, they skipped the ability to name a processor Sexium!

Well, more correctly Sextium, but Sexium is better.

Re:Intel is so far behind anyway

[Anonymous Coward]

True, but you you can patent numbers. I have applied for a patent for 69. Man, I am go ing to be rich...

Re:Intel is so far behind anyway

[bhtooefr]

Here's some confusion clearing info:

8086: 16 bit internal/16 bit bus CPU, first x86
8087: Math Co for 808x x86 CPUs
8088: 16 bit internal/8 bit bus CPU, first popular x86
80186: Enhanced 8086
80187: 8087 modified for 8018x
80188: Enhanced 8088
80286: 80186 with memory protection
80287: Math Co for 80286
80386 DX (technically, 80386): 32 bit internal/32 bit bus CPU
80387: Math Co for 80386 series
80386 SX (technically, 80388): 32 bit internal/16 bit bus CPU
80486 DX: 386DX with integrated cache controller, math co
80486 SX: 486DX without math co
80487: 486DX for a second socket on 486SX board, disables the SX
Pentium (technically, 80586): 32 bit internal/64 bit bus CPU, adds pipelining
P6 (consists of Pentium Pro, Pentium II, Pentium III, P2 Xeon, P3 Xeon, Slot 1 and Socket 370 Celerons, and Pentium/Celeron M, technically, 80686): 32/64, adds RISC86
NetBurst (consists of Pentium 4, Xeon, Pentium 4-M, Socket 423/478 Celerons, technically, 80786): 32/64 or 128, lengthens the pipelines drastically

Re:Intel is so far behind anyway

[Anonymous Coward]

That is an oversimplications that I don't think you understand. Most likely you have read Slashdot and a few other PC/enthusiast sites but not having real knowledge of the fact.

Thank you for your content-empty post!

Re:Intel is so far behind anyway

[skeptikos]

He probably had a Tandy 2000, AFAIK the only PC using an 80186. I have seen it described at "TRS 80 model 2000" in some places.

hrm

[millahtime]

I want to tear down a few things.

Longer batterylife for laptops

I use a desktop, I don't care about my battery life there.

Silent PC's

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

Longer lifetime of the processor

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

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.

Re:hrm

[Junks Jerzey]

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

That was back when Intel x86 chip ran fairly cool. The real question is "how long does a Prescott-level P4 chip last?"

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

[pohl]

It's the fan in my pc that's loud not the processor.

Ever wonder about the reason why the fan is sitting there right on top of the processor? Could there be some sort of relationship between the two? I'd wager that if the processor ran cooler, you might be able to dispense with the processor-fan altogether and just use a heat-sink which, being just a solid hunk of metal, makes no noise whatsoever. This could be the next giant leap in Quiet Computing! Imagine a beo...

Re:hrm

[millahtime]

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.

hehe, my daily used desktop for almost all my personal stuff is a G4 running OS X 10.3. I am also an electrical engineer who has done chip design. I really do understand the inner workings of a chip.

On the note of the fan. Make then quieter. I believe I read a /. article about fans coming to market that run almost silent.

We also live in a disposable society. DIsposable cars, chips, and everything else. They make things to not last long so you have to buy new ones. That's business. Going for longer lasting items is literally a foreign concept.

Re:hrm

[Shinobi]

Smaller fan sounding less than a big fan is all relative. To move the same amount of air in a given time, the small fan needs a far higher rpm, which increases noise.

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: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:hrm

[bubkus_jones]

Is that all?

Re:hrm

[bhtooefr]

Let me tear down a few things on YOUR post:

Longer battery life on laptops:
It does mean lower power consumption on desktops too - lower power bill.

Silent PCs:
While the fan is loud, the lower heat of the P-M makes it easier to get rid of the fan.

Longer lifetime of the processor:
Hmm, the PMMX only puts out ~10W, right (my P233MMX has a heatsink w/fan, but could easily be passively cooled)? Well, the P4 puts out between 80 and 103W. The P-M consumes ~23W - which puts the power consumption MUCH lower than the NetBurst CPUs.

Re:hrm

[jkabbe]

This should have been visible from a mile away. A month or so ago there was an article (on The Register I think) about Intel abandoning it's Mhz marketing in favor of a simple number system (for example, 5xx series for one processor line, 7xx series for another processor line). It was obvious to me at the time this was because they would be dumping the current P-4 and going with the P-M. Why else would you remove Mhz from the equation?

Re:hrm

[LoneStarGeek]

Bock...Bock...Bock... Peck.... Peck.... Peck.... [Crowing] So what did your calculations prove.

Re:hrm

[BlowChunx]

I think the latest kernel comes with a microcode patch that fixes the erroneous calculation of your non-existence. But you still get nice framerates on the proto-fowl...;-|

Re:hrm

[mattdm]

Smaller fan sounding less than a big fan is all relative. To move the same amount of air in a given time, the small fan needs a far higher rpm, which increases noise.

That's not the point -- it's not about necessarily literally physically smaller, but smaller-in-power. Which really *is* less noisy.

Re:hrm

[Anonymous Coward]

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 --.

But apparently not enough time to use a spell checker.

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

[perrinkog]

You missed the point.

The grandparent is saying that a smaller fan would be fine due to lower processor heat. Thus the fan would not have to move as much air.

Re:hrm

[Planesdragon]

Some rebuttals.

I use a desktop, I don't care about my battery life there.

No, you care about your power bill and your circuit breaker. (And, FWIW, LCDs or other large-watt power savers are more important than processors at the moment...)

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

You only need a fan in your PC because of your too-hot processor. My first two computers didn't have any fan outside of the power supply.

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

About that long. CPUs didn't start to get Too Hot until after the Pentium I.

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.

Re:hrm

[coldmist]

Smaller fans = less noise.

Riiiggghhhtt.

Small fans, because of their size, push very little air. In order to push any appreciable amount of air volume, they usually run 5-7,000 RPM. A fan, no matter it's size, makes a lot of noise at anything higher than 2-3,000 RPM. (Think of a car that can do 120mph. Driving it at 50mph will be much quieter than driving it at 125mph.)

Bigger fans can push much more air volume, and can run anywhere from 500 to 5,000 rpm. Once again, anything over 2-3,000rpm make a lot of noise, and the lower you go, the less.

A 120mm fan spinning at 500 RPM will push a lot more air than a 40mm fan running at 7,000 RPM, while at the same time being a lot quieter!

Re:hrm

[Leomania]

for example, 5xx series for one processor line, 7xx series for another processor line

Well, they'd better hope that BMW doesn't decide to sue 'em for using their numbering scheme...

;-)

Re:hrm

[2WheelCowboy]

AMD will be too busy reworking all of its model numbers.

Re:hrm

[tius]

Problem? What problem? Marketing and modern media have ensured that the average person's attention span is no more than 5 seconds.

Besides, imagine the spin: "We continue to Innovate into the 21st century."

You don't even need to mention MHz, pipeline, or any of yesterday's lingo...get with it baby!

Re:hrm

[N1KO]

So that's why all those light fetishists take the time to fill their fans with LEDs.

Re:hrm

[Anonymous Coward]

When the original Pentium I came out, it was ripped on by the nattering classes as a "waffle iron", "spaceheater", "frying pan", etc. Funny how people's perspectives change over time.

Re:hrm

[WingNut7]

Actually smaller fans make more noise. Smaller fans need to run at higher rpm's to produce the same amount of airflow. This means smaller fans will produce a louder, higher pitch noise then a larger fan.

Re:hrm

[Knuckles]

Esp. since BMW has also used the M1, M3, M5 names for their top sports models

Re:hrm

[timeOday]

No problem, Intel will now adopt the "PR" rating system, all part numbers will be estimated comparable speed to P4 chips at speeds that never existed :)

Re:hrm

[MikeCapone]

Smaller fans = less noise.

Actually, that's not true. Bigger fans move more air per rotation and so spin slower, thus making less noise.

Also, when big fans spin fast, they tend to make a less annoying noise than small fans with their high-pitched whine.

Re:hrm

[Anonymous Coward]

5 seconds? That seems a bit harsh doesn't it? Wait...what was the question again?

Re:hrm

[pipingguy]

small vs. large fans

Some aftermarket vendor should start making replacement side doors that are equipped with big (150mm?), slow intake fans for cases. With an easily-replaceable/washable filter. Or is that a dumb idea?

Re:hrm

[sharkey]

Nope, it makes the Internet faster too!

Re:hrm

[Cyno01]

Its not a bad idea, but most people who own dells/compaqs/hps would never buy one and would proably ever need more than stock cooling (i say most, i do have a compaq with a aftermarket video card and a second HD, but i threw in a casefan...). Those of us that build our own PCs know enough (ussually) to properly cool our rig.

Re:hrm

[Cyno01]

Just had some more thoughts, even if PCs shipped with this people wouldn't clean the filter ever. I've opened peoples cases and found dust bunnies larger than my fist inside. Also, just FYI, the biggest of the big PC fans are 120mm, 80mm is "standard" and 92mm and 40mm are the only other sizes i've ever seen for sale anywhere. Hmm, guess it is a dumb idea, sorry...

Re:hrm

[pipingguy]

Yeah, it would have to be for non-mainstream users, but then there seems to be a market for other add-ons and gadgets. The most useful features would be the easy-to-clean filter (assuming the case would be overpressurized by the large fan) and relative ease of installation.

Re:hrm

[pipingguy]

We have post overlap.

I suggest that the reason that peoples' case interiors are that dustbunnyish is *because* it is so troublesome for them to get inside and clean it out (scared of breaking something, sharp edges, etc.). Do most case fans come equipped with filters? Are most case fans even easily accessible? Most I've seen are located down low behind the front facia.

Re:hrm

[Cyno01]

Most dell/compaq/hp cases do not come with filters, and most just come with exhaust fans and vents in the bezel or sidepanel.

Re:hrm

[Cyno01]

There is a market for add-ons. Although most people who install fans in adition to where there are brackets/screw holes in their case can install one anywhere. Its really not difficult to cut a fanhole in the side of your case or a blowhole in the top.

Re:hrm

[pi42]

Actually, I believe I read that Intel cleared it with BMW, believe it or not.

Source on this

Quoted here:

Future incarnations of Intel's entry-level Celeron microprocessor will be dubbed the 300 series. The more powerful and pricey top-of-the-line Pentium chips will be designated the 700 series. Intel executives acknowledged their debt to the German automaker. Don MacDonald, vice president of Intel's sales and marketing group, acknowledged the new names' nod to BMW. "Yes, we talked to them first," he said.

Re:hrm

[pipingguy]

Thanks for your insight. I do wonder about your "...not difficult to cut a fanhole..." comment.

The market I'm thinking about is somewhere between tinkering handyman, Dremel-owning computer geeks and your average home computer user.

Then again, nobody cares that much to spend an extra $50 for easy (physical) maintenance when buying a new PC.

Re:hrm

[Anonymous Coward]

Also remember that, when AMD was trying to get rolling, the speed "increases" Intel would come out with would be something like the "premium" Pentium-100, and the bulk chips were P-90 or P-75's (or celerons gack! but they were all the same chips, however. The P-90's and P-75s were P-100s that didn't pass all the acceptance tests at 100Mhz).

Then AMD comes out with a 120 Mhz Pentium-compatible chip for the price of a Celeron. Voila!

It did force Intel's hand and made them work harder, also. And AMD had to keep pumping things up as much as they could, as well.

Then AMD realized, "we keep saying, and benchmarks justify, that a 1.4GHz Athlon is equivalent to a 1.8GHz P4, so why not just call it an Athlon 1800+?" (while keeping the cost down as well compared to the 1.8Ghz P4)

Re:hrm

[utexaspunk]

>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...

yeah, it's like a beanie with a propeller- since the processor is the "brains" of the computer, it's kind of the nerd of the bunch...

Re:hrm

[Crazy Eight]

You can also find 170mm fans designed to run on 12 Volts.

Re:hrm

[Crazy Eight]

Well, regardless of your desire to nitpick I'm enthusiastic about this change in Intel's roadmap. I like what I've read about the Pentium-M's design so much that I'd consider a desktop built around it right now. It's possible to build a motherboard for it with the E7205 chipset. If someone could hack smp support (perhaps with socket adaptor?) I've got no doubt the results would seriously rock.

Re:hrm

[drunkenbatman]

Woah, you forgot marketing 101: never, ever make people feel stupid or as though they made a stupid decision. :) Not to say your other points aren't valid, but that's the big one.

Say its new and revolutionary, and that this is why it is this way now, and sure some people will be able to read between the lines... but that's fine. Let them save face. But when joe average has made a decision based on GHz, and you essentially call him stupid and bring his ignorance to light, you are not going to get the reaction you hoped for. Even if it makes you feel good.

Just an additional scheme for reducing heat

[elwinc]

After all, you can't go in increasing CPU wattage indefinitely. I can recall the days far past when 30 watts was considered power hungry for a CPU. Sure, you can win a little with more and more rococo CPU cooler designs, but at some point you have to look for still more ways to limit CPU power. The mobile chips do it by varying their clock rates and turning parts of themselves off part of the time. Just think of it as an additional scheme for reducing CPU heat output.

Re:Just an additional scheme for reducing heat

[Rosco P. Coltrane]

After all, you can't go in increasing CPU wattage indefinitely.

For the n-th time, "wattage" isn't a word (perhaps it is, but then only in the M-W). It's called "power".

I can recall the days far past when 30 watts was considered power hungry for a CPU.

I remember a day where it took between 1 and 3 days to complete a raytrace in Povray, where it takes 2 hours tops today. You want to go back to that?

Sure, you can win a little with more and more rococo CPU cooler designs, but at some point you have to look for still more ways to limit CPU power.

I don't think they're waiting for you to tell them to go look for solutions. You'll notice that CPU speeds increase much faster than the power they require, that it's not a linear relationship, and I suspect this doesn't happen magically.

The mobile chips do it by varying their clock rates and turning parts of themselves off part of the time. Just think of it as an additional scheme for reducing CPU heat output.

Run distributed.net or setiathome, or do raytracing on a laptop, and I guarantee you your battery won't last long. The point of power-saving by throttling the CPU and turning parts of it off is to save power while the user does nothing (typically 99% of the time when doing word processing).

Re:Just an additional scheme for reducing heat

[nomadic]

I remember a day where it took between 1 and 3 days to complete a raytrace in Povray, where it takes 2 hours tops today. You want to go back to that?

Well it did make you a LOT more careful when writing the POV files. Kind of how like some writers prefer to write books via typewriter or pen because it forces them to think about sentences carefully before they write them.

Re:Just an additional scheme for reducing heat

[teh*fink]

For the n-th time, "wattage" isn't a word...

Playing devil's advocate.

Re:Just an additional scheme for reducing heat

[ForceOfWill]

I remember a day where it took between 1 and 3 days to complete a raytrace in Povray, where it takes 2 hours tops today. You want to go back to that?

You're just not making complicated enough scenes ;-)

Re:Just an additional scheme for reducing heat

[Anonymous Coward]

For the n-th time, "wattage" isn't a word (perhaps it is, but then only in the M-W).

watt*age (wotij)
n.

An amount of power, especially electric power, expressed in watts or kilowatts.

The electric power required by an appliance or device.

Source: The American Heritage® Dictionary of the English Language, Fourth Edition

Re:Just an additional scheme for reducing heat

[Dun Malg]

For the n-th time, "wattage" isn't a word (perhaps it is, but then only in the M-W). It's called "power".

What the heck are you talking about? Wattage is a word, and it means "power". I'm an electrician. We use the term "wattage" rather than "power" because the latter is too ambiguous. A watt is a unit of measurement, so asking "what's the wattage on that bulb" will get you an answer in watts. Asking "what's the power on that bulb" will get you either puzzled stares or an answer in watts. It is always preferential to speak about quatities using their units of measure. That means amps, volts, and watts. "Power" is not a unit in ANY system.

Re:Just an additional scheme for reducing heat

[ChrisMaple]

Very well expressed. Ask for voltage, not "electro-motive force".

Re:Just an additional scheme for reducing heat

[raygundan]

Wattage and voltage are both perfectly cromulent words.

Excellent

[Abit667]

It was obvious right from the beginning that the Pentium M was far better then the p4, I'm glad Intel is actually making this move.

Sounds like a boxing Match

[ZHaDoom]

We have speculated that this could only happen if Intel knocks Tejas on the head. Sounds like a boxing Match

More info

[Groo Wanderer]

There is more on The Inq here:
http://www.theinquirer.net/?article=15749
http://www.theinquirer.net/?article=15760
http:// www.theinquirer.net/?article=15768

And more coming soon, this story is far from over.

-Charlie

Disclaimer: I write for The Inq, but I did not do these stories.

Model numbers vs frequency

[Mad Quacker]

Now it all makes sense why Intel wants to use model numbers, their newer (faster) cpu's will run at a lower clock rate. Looks like they let marketing run engineering when the produced the P4, and now it's come to resolution.

Re:Model numbers vs frequency

[BCW2]

Gee and I thought they were just copying AMD.

And still losing sales.

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.

Wow, pentium 1000

[lone_marauder]

I've been out of the processor loop WAY too long!

Re:Wow, pentium 1000

[Anonymous Coward]

Is it "Pentium M" or "Pentium m". 'M' is for 1000 but 'm' is 1/1000 like in milli... A Pentium must be from B.C.

Assuming facts not in evidence

[spotteddog]

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

Not eveyone comes from a Judeo-Christian background

Looks like Intel got some religion....

I highly doubt it. Well, maybe they worship the $.

Re: Assuming facts not in evidence

[CastrTroy]

No, they don't, but it's important to know some stuff about other peoples' religions and histories to be a properly rounded person. I'm not buddist, but when the Dalai Lama came to Canada, I knew who he was, and the significance of this event.

Re: Assuming facts not in evidence

[Anonymous Coward]

I highly doubt it. Well, maybe they worship the $.

Yeah.. Wouldn't it be better if they worshipped some entity that someone invented because they didn't know what causes lightning?
And killed others because they had invented something else to worship instead.

Religion sucks f00!

Re: Assuming facts not in evidence

[Anonymous Coward]

Dalai Lama aka Little Debbie

Re: Assuming facts not in evidence

[meringuoid]

Not eveyone comes from a Judeo-Christian background

But you read the Beano, right? So you should know that Jonah was a sailor who, every week, would get a job on a ship and contrive by misadventure to sink it by the end of the strip.

I mean, come on - this level of ignorance of our cultural heritage is astounding, whether you're a believer or not.

Re: Assuming facts not in evidence

[spotteddog]

You are giving people too much credit. My wife is a kindergarten teacher. Most of here career has been spent in inner-city poverty stricken schools. I assure you, after meeting many of the parents, uncles, aunts, grandparents, and siblings of her students that this level of ignorance of cultural heritage is not astounding in the least.

You used the phrase our cultural heritage. My point is that one cultural heritage is not everyone's cultural heritage. I would say you know nothing of my cultural heritage as you do not know me (at least I don't recognize your userid as one I know).

Re: Assuming facts not in evidence

[The-Dalai-LLama]

I've never been to Canada.

...sorry, couldn't resist...

Re: Assuming facts not in evidence

[Anonymous Coward]

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

Not eveyone comes from a Judeo-Christian background

...or has read Moby Dick...

pentium m descended from pentium III

[Anonymous Coward]

so they are dumping netburst for updated four year old technology 'bout time they got a brain. think of the chaos that is about to ensue with code being optimized to P IV instructions. this is exactly why I use amd they certainly don't dump a whole new architecture on us now do they?

Dear Slashdot:

[Paulrothrock]

Is that pronounced Pentium "emm" or Pentium "One Thousand?"
Signed,
Confused Mac User

Re:Dear Slashdot:

[nelsonal]

In the notebook market it has been pronounced Pentium M (or Centrino, if you listen to Intel's marketing department).

Re:Dear Slashdot:

[It's all Krista's Fa]

Do you call your computer the "One Thousand ac"?

Re:Dear Slashdot:

[Paulrothrock]

No, I call it Swordfish. Or Sexy. Depending on my mood (I kid! I kid because I love!)

Multicore cores - OS/software

[CodeMonkey4Hire]

In the article, they mentioned 2- and 4-core chips coming out in 2005 and 2006. How long will it take the unix and windows operating systems to take full advantage of all the cores? Are they already there? Does software running on these systems need to be specially written and/or compiled in order to take advantage of the multiple cores or will they mostly serve to handle multitasking?

Re:Multicore cores - OS/software

[avdp]

I believe all that this means is that it's essentially multiple CPUs on one chip (with apparently some cache sharing). Should be handled out of the box by any SMP kernels.

Re:Multicore cores - OS/software

[Sique]

There are four different approaches to handle several cores.

a) Tread them like different processors. This requires you to use either an SMP capable operating system (virtually all Unices, Windows NT Series Server edition) to fully leverage the advantages, or...
(Operating system level)

b) ...have your software being aware of multiple cores and use a multithreaded approach. So calculations can be split into different treads and those dispatched to different cores.
(Application level)

c) Have your compiler optimize the software for parallel execution of instruction during the compile process. That means for instance to try to fit operations together in a way that two consecutive operations don't depend on each other.
(Compiler level)

d) In the prefetch queue of the processor check for interdependencies between operations. If two operations are independent of each other you can put them into different cores and execute them in parallel.
(super scalar processor)

Those four approaches are more or less coupled together.

a) is a special case of b), if you call the operation system a special application which governs all other applications. If your computer is supposed to run only a singlethreaded application at a time, then having an SMP operating system doesn't really speed up your work (ok... printing in background works better ;) ), if the application doesn't use multiple threads.

On the other hand having no SMP aware operating systems means that one core gets most of the work because it runs all processes. Even multicore aware programs are started on the first processor and only dispatch certain threads to the other cores.

c) and d) are also tightly coupled. If the compiler is able to optimize the code in a way that successive operations are independent of each other, then a multi core processor is better able to keep all cores working in parallel.

b) and c) are somewhat intertwined, because redesigning software to take advantage of a multi threaded model means also that you have to loose the dependance of different operations in your program and to design your algorithms in a way that most of it can be executed in parallel.

So as a conclusion: You will probably notice a speed advantage by going multicore from the beginning, because some of the conditions are already meet. A processor can dispatch operations to different cores even though the software may not be optimized for that. So a slight increase of speed is to expect even from very old software. As you add newer software to your computer and update your operation system, other speed ups will be possible.

Re:Multicore cores - OS/software

[bhtooefr]

Actually, I think you're not quite right on only NT series Servers supporting SMP.

Why else would I see stickers saying stuff like "Windows 2000 Professional 1-2 CPU" and "Windows XP Professional 1-2 CPU" on computers that ship with those OSes?

Re:Multicore cores - OS/software

[Sique]

I don't know where you got the "only" from :) I just mentioned the WinNT Server series as an example.

Mircosoft and CPU licensing.

[achilstone]

If Intel is moving to low power multicores chips what will this mean in terms of os licenses that limit the number of cpus which can be used at any one time such as MS? Will the likes of MS relax the licensing terms so that this type of technology will take off or will they use it as an excuse to make more money and kill off demand for multicore cpus? I wonder if Intel has discussed this with MS?

Re:Mircosoft and CPU licensing.

[BCW2]

No one will pay M$ extra for this and they will look even worse if they try to enforce it. I think you could win the argument with the one socket = one CPU, no matter the CPU design. It will be fun to watch.

Re:Mircosoft and CPU licensing.

[achilstone]

Damn I wish /. had a spellchecker.

Re:Mircosoft and CPU licensing.

[jo_ham]

You have no choice - I thought the current version of Windows can refuse to run on MP architectures unless an MP-safe licence code is entered when setting up, and to get one of those you have to call MS.

Re:Mircosoft and CPU licensing.

[bhtooefr]

I thought it just ran one CPU. BTW, WinXP and WS2K3 think of HT in the kernel as a second CPU, but in the licensing system as part of the first. However, Win2K was made before HyperThreading even came out, and two HT CPUs can't be run in HT mode on 2K Pro - it'll detect 4 CPUs, and only CPU0 and CPU1 (the HT part of CPU0) will be used.

Re:Mircosoft and CPU licensing.

[bhtooefr]

Just make the multicore chips lie to Windows, and say that they're HyperThreading chips. Windows will run on it like it's two CPUs, though it's on a one-CPU license (note - this stunt doesn't work in 2K, as it doesn't recognize HT for what it is, but the editions of Win2K/XP I think you're talking about support two CPUs anyway), and there's nothing to worry about. Just don't let Intel HT the cores...

Hook 'em

[mgs1000]

Well, there goes my plan for having a Tejas Longhorn computer...

Good Move Intel

[two_stripe]

This is great news from intel. The Pentium-M absolutely kicks the Pentium 4's butt! Itll be interesting to see how they market this seeing as they've been the ones pushing the "More Mhz means a faster CPU" Idea. Also, i'tll be interesting how they market the celeron line as well.

Re:Good Move Intel

[bhtooefr]

They'll market the Celeron M line.

Prescott chips...

[Anonymous Coward]

I always thought naming a processor chip after a deputy minister who is responsible for transport but prefers to drive high-performance cars (Jaguars) for short distances, was the equivalent of putting on metal armour, climbing to a mountain during a thunderstorm, and waving a sword in the air while making disparaging comments about the gods.

Re:Prescott chips...

[Patrik_AKA_RedX]

Common, you really think the Gods would waste a perfectly good lightning bolt on an unpolite idiot in armor?
Besides it's much more funny to strike the person standing next to the idiot, who's trying to explain the dangers involved.

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:Return of the son of the revenge of the P6

[barawn]

Quad-pumped FSB, longer pipeline

A quad-pumped FSB might make sense, although I doubt that the PM is actually all that memory-hungry, as the old P6s weren't, and neither were the Athlons.

A longer pipeline is virtually the definition of the P4 - it has one of the (if not the) longest pipelines in desktop processors anywhere. A long pipeline is what causes low IPC.

I really doubt that they'll lengthen the PM's pipeline much. Look at the Athlon XP -> Athlon 64 evolution - the pipeline was only stretched by a couple of clock cycles.

This is a curious point for Intel, as processors can't continue to get faster in a simple way - the heat issues are just too large right now. The PM will probably start getting the standard tricks that others are playing - hyperthreading, like the P4s, integrated memory controller, maybe even an L3 cache. But definitely not a long pipeline - that was the P4's mistake.

Re:Return of the son of the revenge of the P6

[MyHair]

The P6 (PPro, PII, PIII) architecture is coming back to the desktop. This does make pretty good sense.

I agree. Intel had to brutally kill PIII because it was beating out the higher clockspeed PIVs. As a consumer that never made sense to me, and I've never wanted a PIV. I'd rather have had a PIII 1.3GHz or an AMD.

One of my users just got a new laptop with a Pentium M. I didn't really know what it was until yesterday, and just now I realize the clockspeed is 1.6GHz where his old PIV laptop was 2.4GHz. I haven't found any GHz claim or speed rating on the laptop or box yet, just a "Pentium M" logo prominently displayed below the keyboard. Heh, that 2.4GHz laptop is a joke to me; it has a rather loud fan that kicks on frequently when using it. I have yet to hear the Pentium M laptop make any unfriendly noises. (Besides the time I had that unbalanced CD in the drive, but that's not the laptop's fault.)

-Typing this from a PII 266MHz laptop. The techs in my company get the bottom-end equipment. Go figure. Not that I'm complaining; I think it's funny.

Re:Return of the son of the revenge of the P6

[Anonymous Coward]

(Score:5, Informative) ???

Pentium 4 tricks ... longer pipeline ???

Re:Return of the son of the revenge of the P6

[zsazsa]

I really doubt that they'll lengthen the PM's pipeline much. Look at the Athlon XP -> Athlon 64 evolution - the pipeline was only stretched by a couple of clock cycles.

That's true. According to articles I've seen about the Pentium-M, its pipeline is only a little bit longer than the PIII's. I guess I misspoke - mod parent up!

Re:Return of the son of the revenge of the P6

[bhtooefr]

Most estimates place Banias at 14 stages. AFAIK, the other P6 CPUs have a 10 stage pipeline, Williamette/Northwood/Gallatin have a 20 stage pipeline, and Prescott has a 30 stage pipeline.

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:You are groping for....

[Anonymous Coward]

No he was after Opium.

As long as Intel tries to follow Moore's law,

[Anonymous Coward]

who cares about the details. Or, has Intel been stricken by save-the-environment virus.

Dual-core P4 4GHz for Longhorn?

[ghostdancer]

Does this mean... We have to wait longer, in order for us to run Longhorn on its recommended hardware?

Lot of speculation makes my head spin...

[tomstdenis]

From what I've read the Pentium M is nothing more than a souped up P6 core [slightly longer pipeline, 1MB power optimized cache and SSE/SSE2 iirc].

So really it's the PIII v2.0.

Last I checked though the Athlon XP [and Barton] still whoop the PIII's ass and probably the Pentium M at equal clockrates [let's not forget that 2.2Ghz AMD Bartons exist...].

So my guess is if they do ditch the P4 it won't be directly to the Pentium M but instead a design based off it [but with a higher IPC than the P-M] cuz otherwise they'll get screwed in the processor speed wars and that would look awful stupid.

Tom

Re:Lot of speculation makes my head spin...

[barawn]

Read the Ars Technica article on the PM - its main benefits are instruction grouping and a massively improved branch predictor. Intel claims that just the branch predictor alone improves the PM's performance by 6%, and a 6% jump in real-world performance is very impressive. All in all, with all of the PM's improvements, it's about 15-20% clock-for-clock faster than the P3 was.

Don't judge the PM by its roots - it's a much newer design than the old P3, and I can't imagine that clock-for-clock, the Athlon and the PM wouldn't be very competitive.

The PM is one of the highest IPC x86 processors out there. You can't just improve on it easily.

Re:Lot of speculation makes my head spin...

[tomstdenis]

I'm not saying the PM is slow I'm just saying compared to an Athlon it's not faster. Well at least at ALU work. I'll bet the PM has a stronger SIMD unit which is where the AMD K8 comes in ;-)

Point is a desktop cpu won't be a verbatim port of the PM for long... ;-)

Tom

where does that leave performance freaks like me?

[JustNiz]

Its finally time to build myself a new PC. I just want the most powerful Intel-driven box I can get my hands on, for heavy-duty emulation etc. I don't care about power/heat/noise issues as I am building a desktop and will sacrifice all of that for outright performance.

It sounds like the fastest new pentium M will be 2Ghz, will that REALLY compete performance-wise against the forthcoming 3.6Ghz Pentium 4?

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

Incredibly smart move!

[Junks Jerzey]

Desktops have been in trouble for a while. A new CPU comes out with an 8% higher clockspeed, and then it uses 15% more power. Obviously there was a limit to how long that could continue, especially as those diddly performance increases weren't providing tangible benefits (compare an 8% clockspeed increase with switching to a dual core processor, for example). And at the same time the desktop market has been being heavily outpaced by laptops and mobile devices.

Re:Incredibly smart move!

[steveha]

It's not so much an incredibly smart move, as it is walking away from an incredibly dumb move (pinning the future on the Pentium 4 architecture). The Pentium 4 is so broken in so many ways.

steveha

Re: the solution is ...

[Anonymous Coward]

The solution is SOIULVLDOBDWMC: Silicon On Insulator Ultra Low Voltage and Low Density On Big Die With Much Copper.

open4free (c)

hmm

[Anonymous Coward]

ya know, I think there's a limerick in there somewhere...

Re:hmm

[Anonymous Coward]

There once was a man from Nagasaki
who one night drank way to much sake
Later searching the Net
much to his regret
he found he had lead role in a bukake
(video that is.)

Re:hmm

[Anonymous Coward]

Except that it's pronounced Boo-kah-kay.

Cheap laptops

[SpinyManiac]

This means the end of desktop CPUs in laptops.

Decent battery life in a cheap laptop? Nah, they'll cut down on the batteries instead.

Re:Cheap laptops

[ashot]

I have had the pentium-M Thinkpad T-40 which has an larger battery option for about a year now. Runs for 8 hours with (5 with non-stop music/games), not cheap though.

Pentium4 oven?

[elmartinos]

A lot of people have electrical heaters. Why not replace the heating elements with processors which are designed to be able to run at a very hot temperature? Imagine the processing power if every heater actually was a cluster of 20 pentium4 processors! And all these machines are networked. This could be expanded to every heat producing machine. Just imagine you toaster, boiler, hair-dryer are a beowulf cluster calculating seti@home packages - a geek's heaven. It could also be useful for companies who need that processing power to lend this for little money.

Re:Pentium4 oven?

[Trumpetgod2k1]

$$$ to build P4 driven space heater >> all money you and I will ever see

Re:where does that leave performance freaks like m

[paitre]

Then you obviously haven't used anything from them recently. Seriously dude, AMDs current crop (Barton XP and Athlon64) are kick ass chips, and the boards are stable.
Don't blame AMD for the chipset companies provided half-assed chipsets that predominantly sucked. That's pretty much been fixed, and what do you know, AMD's king of the performance hill.

But to answer your direct question:
No, a 2GHz Dothan will -NOT- be able to compete with a 3.6GHz Prescott. But then, you might be able to aircool dothan, or use a lightweight fan if you can't, and you're borderline needing a water-cooler for the highest end prescotts.

P4 can't be speed up.

[Anonymous Coward]

110w is the power consumption/disipation of the lastest P4 3.4 EE (With 1MB of cache).

So what's the downside to this move?

[Tyfud]

Seriously, if it's as good as everyone says, and the architechture is better, and it runs smoother, and faster, what would be the downturn of this move?

Re:So what's the downside to this move?

[Anonymous Coward]

Intel has to backpedal on their carefully crafted "More MHz means more better!" propaganda. Bad for Intel, but good for consumers.

Re:where does that leave performance freaks like m

[FSWKU]

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

Then you must be doing something wrong. I've used both Intel and AMD CPU's and have never had a problem with stability on the hardware side of things. In fact, my current AMD system has been up and running since I switched it back on after spring break. According to my command line: "\\SHODAN has been up for: 19 day(s), 11 hour(s), 12 minute(s), 8 second(s)" -- WinXP Pro on an AMD processor. Completely stable. To be fair, I've never had an issue with Intel chips either. I have a 733mhz P3 that's still going strong after 4 years (not total uptime, but its still stable nonetheless).

Intel's nordic PR representative is denying this

[Jugalator]

From an article at IDG.se. A quick translation:

- Intel has never used to comment unreleased products, or so called "road maps". And there is no press conferences or presentations planned for today where this kind of announcement can be made.

But can't the news about the Tejas and Jayhawk cancellations still be true?

- I doubt that. In that case, it is the first time for seven years that I have not had the information available in advance.

So either it's a miscommunication on Intel's side, or the news are false...

Pentium M vs AMD xxxx+

[swillden]

Will this move make the AMD "marketing ratings" irrelevant? They were invented to help make it clear that AMD processors performed as well as their P4 competition, even though the P4s were clocked much higher. The Pentium M, however, is based on the P3 architecture, which has always had performance that is comparable to the AMD chips at a given clock rate.

Will this now force AMD to find a graceful way to drop the marketing ratings, lest they appear to be artificially inflating their processors' performance?

Re:Pentium M vs AMD xxxx+

[JDevers]

More than likely they will actually follow Intel's and their own Opteron strategy. In other words, simple model numbers indicating performance.

Re:Pentium M vs AMD xxxx+

[Anonymous Coward]

Nope. Intel is the ones who have insisted clock speed is the king of performance. Now they're the ones that have to change their marketting to compete with AMD's performance ratings.

Re:where does that leave performance freaks like m

[MoronGames]

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

You're probably coupling them with some other sub-par parts then. The processors themselves are fine. Are you using cheap RAM? A cheap PSU? A cheap motherboard? All of those things make a huge difference in stability.

Re:where does that leave performance freaks like m

[sydres]

yeah or at least very close considering I have seen benchmarks that put a 1.6 ghz pentium M in aboout the same performance arena as a 2.4 ghz pentium 4 so maybe a little lower. but the new one comes with or so they say 64bit extensions. as well as an increased ipc or so I would imagine then the fact of the matter is their will be dual and quad core chips available so yeah it probably will compete plus burn less juice

Why 64-bit x86 chips aren't catching on

[MarcQuadra]

Well I think the reason AMD64 chips aren't catching on as quickly as they could is that the AMD roadmap clearly shows how they're moving to a new socket soon. Why would you buy or invest R&D into a machine that is essentially dead-ended in terms of CPU availability?

I think that once the socket-939 chips come out and the platform 'congeals' into a long-term solution you'll see more of these things selling.

It also doesn't help that there's no version of Windows that takes advantage of 64-bitness yet, or that a HUGE portion of the IT people making purchasing decisions still doesn't trust AMD CPUs.

Re:Why 64-bit x86 chips aren't catching on

[paitre]

It also doesn't help that there's no version of Windows that takes advantage of 64-bitness yet, or that a HUGE portion of the IT people making purchasing decisions still doesn't trust AMD CPUs.

I don't resemble that remark! And neither do a lot of the HPC geeks. We need bang for buck, primarily, and we're getting with with Opty right now :)

Re:Why 64-bit x86 chips aren't catching on

[MarcQuadra]

Hey, I'm with you, but every time I price out an AMD desktop I get shot down because "it might not be 100% compatible with our intel software". It's amazing how far the PHBs can push their heads into their asses.

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.

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.)

Re:Increasing number of cores...

[Anonymous Coward]

Whats wrong with 'The Register'?

Re:Increasing number of cores...

[robnauta]

Theregister is notoriously unreliable because they usually post unverified rumors. Sometimes they claim X, then a week later the opposite of X, then when the company releases any news, whether it's X or not X, they will boast about predicting it right by pointing to the right guess, while deleting the other one.

Re:Increasing number of cores...

[Anonymous Coward]

Register sucks. If you want juicy rumors, try The Inqurier -- they correctly reported that the 64-bit Pentium and the fact that the P-M would replace the P4 for most desktops months before anyone else.

Re:Increasing number of cores...

[Anonymous Coward]

8. A dual-core processor is "like putting two cylinders in a car instead of having one big cylinder."

Re:where does that leave performance freaks like m

[IWorkForMorons]

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

Bud...then you're limiting your speed and power right there. AMD systems are quite stable from my experience. Usually, it's when people buy bad equipment to surround their chip that causes them to think the chip is bad. Such as my own recent stablitily problems. After replacing the CPU fan and heatsink, replacing the power supply, and very nearly sinking $300 into a water-cooling system because I thought the chip was over heating, I spent $25 on a memory heatsink. Haven't had a single problem with unstablity since. And believe me...this system holds it's own against the hot and heavy Intel systems that are rated at several hundred Megahertz faster...

Scaling dead? Have we hit a clock rate wall?

[mc6809e]

The new 90nm Pentium 4's really didn't get much of a clock rate boost, which was a surprise. Reducing feature sizes has usually given us a good bump in clock rate. Remember the original Pentium 4 when it came out? There was a big jump in clock rate. This lastest shrink hasn't provided much. Now we hear that Intel is going to the Pentium-M: a chip with a lower clock-rate. That doesn't mean the chip is a poor performer. In fact, it runs very well. Like the Athlon, it gets much more work done per cycle than the Pentium 4.

Still, process shrinks in the past have yielded easy speed increases, but not this time around. Intel's move seems to confirm that there might be trouble ahead.

It looks like the folks at IBM also have concerns:

"Somewhere between 130-nm and 90-nm the whole system fell apart. Things stopped working and nobody seemed to notice."

Intel is floundering

[Mr. Neutron]

It seems to me like Intel just doesn't know what it's doing these days. While AMD does new and innovative things, like the first consumer desktop 64-bit x86 archetecture chip, what's Intel doing? Die shrinks and more absurdly drawn-out pipelines, it seems. If I were in charge of the shop at Intel, I'd set the following priorities:

1. Make a 64-bit challenger to Athlon64. If it means butchering the Itanium die and adding a 32-bit co-processor, so be it.

2. Enable SMP on something faster than Tualatin.

3. Wake up to the fact that Intel can no longer dominate the CPU market on name recognition and MHz rating alone.

All I can say is, at least Intel is opening up the way for more competition. It won't be long before the market share is split 60-40.

Re:Intel is floundering

[aksansai]

1. Make a 64-bit challenger to Athlon64. If it means butchering the Itanium die and adding a 32-bit co-processor, so be it.

Intel has already announced that it will also be releasing a variant of the x86-64 technology that was developed by AMD. You can see their announcement here. While technology analysts see that there are indeed differences, it is approached fundamentally in the same manner that AMD used - making compiler development for the "extended" 64-bit Intel processors easier.

This does not mean Intel is simply give up on the Itanium. They have more than a decade worth of R&D dollars into the processor. I don't believe they will actively pursue integrated the two 64-bit processors under one flag, either, because it would be easier to keep one facility churning out Itaniums and all their other facilities to make modified P4/PM chips with 64-bit extensions.

2. Enable SMP on something faster than Tualatin.

This is a matter of market preference. The market prefers a single processor (right now). I remember seeing supporting statistics but I can't find those at the moment. It was better than 90%. R&D is currently focusing on making single processors more efficient (Intel's hyperthreading is a good example). All these improvements will eventually trickle down to the small SMP sector Intel supports.

Also, Intel wishes to keep its Xeon and Xeon MP line strong. To do so would be to limit the offerings of SMP capable chips and chipsets to focus the multiprocessor market for higher profitability.

3. Wake up to the fact that Intel can no longer dominate the CPU market on name recognition and MHz rating alone.

Intel is beginning to realize this, but this does not change the fact that there are many people that will still choose Intel over AMD just based on name alone. I run into these types on a regular basis.

Re:Intel is floundering

[RESPAWN]

Intel is beginning to realize this, but this does not change the fact that there are many people that will still choose Intel over AMD just based on name alone. I run into these types on a regular basis.

It's like the old saying goes. "Nobody ever got fired for buying Intel." Personally, I feel the same way. I'd rather cover my ass and buy Intel all the time instead of AMD. Also, I haven't yet seen much data regarding AMD's performance/reliability in the business desktop arena. I haven't seen anything like that for Intel either, but that really doesn't matter. Intel is the standard and before a move to AMD can be considered, it has to be shown to be better than the standard.

Gates: 64-bit chips the norm by 2006

[Anonymous Coward]

http://www.taipeitimes.com/News/worldbiz/archives/ 2004/05/06/2003154365

Gates: 64-bit chips the norm by 2006

BLOOMBERG
Thursday, May 06, 2004,Page 12

Microsoft Corp chairman Bill Gates said nearly all personal computer chips on the market will be able to process data in 64-bit chunks instead of the current 32 bits by the end of next year.

"Between now and the end of 2005, we'll go from having very few 64-bit chips out there to virtually 100 percent of what AMD [Advanced Micro Devices Inc] ships and the majority of what Intel [Corp] ships," Gates said at a conference in Seattle.

Software companies and chipmakers introduce new technology to convince consumers and businesses to trade in old machines for new products that do more. By processing data in 64-bit pieces, currently only done by high-end server machines, chips can manipulate data more quickly.

Advanced Micro, which last year began selling processors that let PCs run the software usually used by servers, said it is ahead of its larger rival. Intel, which hasn't disclosed when it will release a 64-bit-capable desktop chip, said it will enter the market when demand warrants it.

"We said that we would add 64 capabilities when the customer demand is there," said Robert Manetta, a spokesman for the Santa Clara, California-based Intel. "We haven't given a firm timeline for when we would do it for PCs."
This story has been viewed 340 times.

DELL support FAQ - AMD

[Anonymous Coward]

Here is Dell's response in a question about an AMD Dell. Click the link under knowledgebase. (A little old, but sadly their most current response.)

How to sell this

[HarveyBirdman]

Call it the Pentiummmmmmmmm and get Homer Simpson as a spokesperson.

Yeah, that should be the another nail in Intel's coffin.

Oops, I didn't mean to type that last bit out loud!

right, but you need to move less air...

[caveat]

...with a cooler processor (a given volume of air can remove a given quantity of heat, so if you have less heat, you need to push less air to remove it). your smaller fan can spin at a lower RPM, hence it's quieter.

Re:right, but you need to move less air...

[Shinobi]

But the large fan needs to spin even less quickly, so will sound even less too.

My current CPU fan is far larger than my old one, moves more air, but is still more quiet than the old one.

One P4 sucks more power than a whole dual G4 box!

[caveat]

it's a bout freakin' time; i was curious what the power needs of a G4 proc were as compared to a P4, so i went hunting - my entire dual 1.25ghz G4 box sucks down 113 watts, while a single 3.2ghz P4 wants ~115W IIRC...that's just re-freakin-diculous!

I wish I could tell..

[DamienMcKenna]

my ex-co-worker who was a major Intel biggot and thought that the Tejas was going to be the best thing ever. Yeah, best thing ever for frying eggs on your PC case.

Damien

Never going to happen

[msgmonkey]

Power pins are the least of the problems, bus protocols is the main reason for incompatibility.

AMD have gone with hypertransport and integrated membory controller in the Opteron/Athlon 64.

AFAIK Pentium-M uses a "hub" architecture with a 400MHZ link with the hub also providing the DDR memory controller amongst other things.

The reason Socket7 boards worked with different chips is because they all used the orginal Pentium bus protocol. With Pentium Pro, Intel went with a new GTL bus which was n't licensed to AMD, so AMD went with the Alpha bus with the K7(as there was ex Alpha people working on K7).

Re:Never going to happen

[Anonymous Coward]

(as there was ex Alpha people working on K7).

Oh, was there?

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:Another version 4 failure

[TeknoHog]

Interesting how Windows skipped version 4...

Windows 95 and 98 are actually version 4.something, which you can see with the 'ver' command in a DOS shell. Bloated and disappointing, you said?

Re:Another version 4 failure

[/dev/trash]

It did? It was called Windows 95, but internally it was version 4.xxx.

Re:Another version 4 failure

[phoenix_rizzen]

Windows didn't skip version 4. Windows 95 has the internal version number 4, as does Windows 98 and ME. There's also Windows NT 4.

The DOS 4 that was released was horrible. The DOS 4 that didn't was simply amazing (multi-tasking and everything). Only a few European customers got to see that one. :(

And the Pentium4 isn't the fourth iteration of anything. It's nothing more than a bizarre marketing moniker. Especially considering the Pentium-II was actually the third CPU to have the Pentium name (fourth if you include the Pentium/MMX).

Re:Another version 4 failure

[pfraser]

Interesting how Windows skipped version 4...

Never heard of Windows NT 4.0?

Re:Another version 4 failure

[jak163]

You forgot Rocky IV.

Re:Another version 4 failure

[Ben Hutchings]

Some companies that use generation numbers for their products just skip number 4 because it is considered unlucky in Japan in the same way that number 13 is considered unlucky by some people in the west (the Japanese words for "four" and "death" are similar). This is why there is no Palm IV for example. Maybe there's something in it. ;-) However IE 4 was very successful - in my opinion it was the first usable version of IE and it would probably have killed Netscape even if it wasn't bundled.

I am seeing a lot of posts...

[anethema]

that say this is a great idea. I'm personally not so sure. They havent got the core beyond 1.7 ghz. Why? This is a P6 core cpu! Thats right, pentium pro, pentium 2, pentium 3, etc. They added some features, new stepping and thats it.

This has also been mentioned, but what I'm saying is, I have NO idea how they will get any kind of clockspeed out of this thing. It's been tooth and nail to try to get it to run where it is even. They can drop the process size to get more speed out of it, but that only takes you so far. You will have to lenghten the pipeline eventually which carries with it all the problems of the P4.

I just honestly cant see them taking the P6 core any further. I'm suprised they can even do what they do with it.

I also doubt they will drop the Pentium 4 core while they still have a lead on the athlons performance wise. P4 is a core they spent a LOT of time and money developing. I doubt they will just drop it.

Also, in the article, they say that intel will be tearing up their roadmaps and they SPECULATE Pentium-M will be the replacement. Might be a modified Pentium 4, or maybe a Pentium4-PentiumM mix.

Just my 2c.

Re:I am seeing a lot of posts...

[aksansai]

This has also been mentioned, but what I'm saying is, I have NO idea how they will get any kind of clockspeed out of this thing. It's been tooth and nail to try to get it to run where it is even. They can drop the process size to get more speed out of it, but that only takes you so far. You will have to lenghten the pipeline eventually which carries with it all the problems of the P4.

I just honestly cant see them taking the P6 core any further. I'm suprised they can even do what they do with it.

I see a plethora of possibilities for the P6 core without having to adopt all the performance penalities introduced in the latest P4 chips. Again, Intel's direction is exploring the use of multiple cores (like SMP on a chip), adding or enhancing existing the ALU to make processing threads more efficient, etc.

The one thing Intel has been able to achieve since the popularity of the 8088 (God has it been that long??) is being able to develop innovative ways to get a little better performance out of a processor.

AMD has achieved this quite nicely (even getting the speeds above 2GHz) on a much simpler design. Although they can't obtain the latest P4 speeds, performance translates to near equal of the faster (GHz) Intel Pentium 4s.

Re:I am seeing a lot of posts...

[pkesel]

Mod this up. He used 'plethora' in his post. That's got to rank something for insightful (maybe even inciteful).

Re:I am seeing a lot of posts...

[phoenix_rizzen]

You have to look beyond MHz and to overall CPU performance. Sure, the Pentium-M is only around 1.7 GHz right now. But that translates into a *lot* better performance than a Pentium4 at 2.0 GHz.

Intel has finally realised they can't compete on pure MHz anymore. Look at what that got them: a super-fast Pentium4 that requires more electricity and generates more heat than the electrical heating system in my apartment.

The Pentium-M architecture won't scale into the 10's of GHz. But the performance will be much better than a Pentium4 that does scale into the 10's of GHz.

Which would you rather have: a super-high GHz CPU that maybe does 3 or 4 instructions per cycle, or a slower, less power-hungry CPU that can do 6 or 8 instructions per cycle? Do the math, and the low-speed, high IPC system will (always) beat the high-speed, low IPC system.

That's the prob with tech-support

[sindarin2001]

That's the problem with being in tech-support. We can always breath new life into old hardware, there-by not giving us the excuse to ask for new hardware.

Damn the M....full speed ahead

[Prince Vegeta SSJ4]

I have a P4 machine on every floor in my house Just when I was about to cancel my Natural Gas service, the go and try to Lower the heat output of chips.

Huh?

[aksansai]

Interesting how Windows skipped version 4...

Umm... where have you been?

Windows NT 4.0 was released in 1996. Eight years later, it is still supported by Microsoft for security fixes, bug fixes, and moderate enhancements. NT 4.0 is still a powerful, light-weight operating system that can run even some of the newest technologies Microsoft is churning out, like the .NET Framework.

Re:Huh?

[yngv]

I admit I was thinking in terms of home PC OS, but was wondering if Microsoft's decision to not name the successor to Windows 3.11 as Windows 4, but instead to go with dated versions (Windows 95), was in any way related to the apparent version 4 curse.

And where have I been? More like where haven't I been!.. After graduation, I quickly became involved in an international ring of sexual criminologists, culminating in...

Read the rest of this comment

Re:Huh?

[Slack3r78]

Given most of the software you named wasn't really around prior to Windows 95, I seriously doubt it. You're grasping at straws here.

Go back to a variant of iCOMP...

[aksansai]

Intel used to stamp the iCOMP index on their processors so that people could compare the Intel benchmark of performance of the 80486 with the Pentium. Intel touted the slower Pentium 60 as being faster than the 80486-66 which was true. However, the stamp was nice and small - and was not a major advertising point. Intel dropped it, I believe, during the latter portion of the regular Pentium era.

AMD's rating gives me a good idea of X processor is slower than Y processor and the difference in cost is N. If the companies switch to something, hopefully they will KISS to maximize what it is intended to do - show the basic improvement of one processor over another.

Wow - my computers didn't just have a CPU!

[chriso11]

It seems that there are a few here who don't think of things like the MB! It uses power too! And so does the Video Card, HDs, and memory.

In fact, I'd bet that the CPU is only about 20-30% of the power consumed in a system. Generally, systems with higher speed CPUs (hence more power), generally have the higher power graphics and HDs.

So while you Macaddicts may be absolutely correct about the CPU powers, it isn't the whole story. That's why I buy AMD!

Is the English translation "Moore's Law is dead"?

[Anonymous Coward]

Enquiring minds etc. etc. yadda yadda....

Re:where does that leave performance freaks like m

[Anonymous Coward]

It sounds like the fastest new pentium M will be 2Ghz, will that REALLY compete performance-wise against the forthcoming 3.6Ghz Pentium 4?

I wouldn't be surprised. Clock speed has become the least useful measure of processor performance there is. You can't make any useful comparions between processors by clock speed. So, when that 2GHz Son-of-686 is compared to a Pentium 4, don't be surprised when the Pentium M wins. Perhaps this will finally put an end to the MHz myth, and companies will start marketing processors that don't suck to consumers.

You Forgot One!

[TheWizardOfCheese]

But that's excusable. The 386s[u]x deserves to be forgotten. However, as others have mentioned, the 186 was actually quite successful in embedded applications.

...and I'm an idiot

[getch();]

Um yeah. It's C*V^2*f not C*V*f^2. My point remains though. Without getting into the specifics, power per unit area is increasing currently.

I was trying to avoid getting into things like subthreshold leakage current (to explain why Vdd can't be reduced arbitrarily).

That's what I get for posting ten minutes after I wake up I guess.

Re:...and I'm an idiot

[default luser]

But there's not much to get into. There's a particular threshold voltage required to activate a CMOS gate (after you take leakage and impedence into account), and this is the limiting factor in your voltage.

Incidentally, this is one of the reasons power usage in processors has skyrocketed in recent years. Only 10 years ago were were throwing 5v Vcc into chips. Each successive die shrink brought us down (3.3v, 2.8v, 2.0v, 1.6v), reducing power required by the square. This meant we could increase processing capability (adding both parallelism and clock speed) without a huge power usage increase.

Now, when was the last time you saw Vcc go down with a new die shrink? It's been a long time. We can only push the threshold voltage so low and still have a product that works and is profitable.

Diamond Age

[dpilot]

The execs are afraid that they'll be lured mindlessly away into secret underwater tunnels and communicate through hot sex.

Re:Diamond Age

[AviLazar]

You would think they would be happy to get some?

Re:Diamond Age

[dpilot]

Look like you haven't read the book. When I said 'mindless' I meant 'mindless', and when I said 'hot', I meant 'HOT'. Wouldn't be around mentally to appreciate it, or physically afterward.

Re:Diamond Age

[rebel47]

And the problem with communicating through hot sex is??

Re:Diamond Age

[SquadBoy]

He means hot as in spontaneous combustion.

jesus what is the world coming to when a Diamond Age joke has to be expained on /.

Itanium: Linux box of the now and future?

[wandazulu]

This is total tin-foil-hat speculation, but hear me out:

Considering the enormous amount of resources Microsoft is going to have to keep pouring into Windows development, I can't imagine they are going to want to keep spending the $$$ on maintaining multiple versions for various CPU types, now down to basically two: x86 and Itanium. They saw that NT on Mips, PowerPC and that NEC chip (forget what it's called) were lousy sellers, and you've got to imagine that though NT on the Alpha was probably going to be a big seller, the implosion of DEC pretty much put the lid on its future and Microsoft decided quickly it wasn't going to be spending the time and resources to keep up an Alpha port, no matter how good it probably was.

Even though Intel says otherwise, you have to think that they see the Itanium as not the future for Windows; people don't really trust Windows to operate well on the x86, why should they think it'll run better on the Itanium? I predict that the server version of Longhorn will be an x86-only product (of course, probably *requiring* 2 or more of those as-yet produced multi-core cpus).

So let's presume that Windows won't run on the Itanium in the future. What's left? Solaris, AIX, or Irix on the Itanum? Nope (3rd paragraph) That leaves only one OS that already runs on the chip, our friend the penguin.

So what will Intel do with it? Seems like the future points to a processor that runs Linux (probably NetBSD too) and little else. If Intel doesn't can the thing altogether (which is unlikely given the huge resources spent, and the serious black eye it would give them), then presumably we would have a cpu to call our own: a machine built with nothing but Linux in mind. Combine this hardware with all the software that's a little more than a recompile away, and you'd have a pretty sweet setup to counter the Windows juggernaut.

Is Israel a bad location?

[Nimey]

I have to wonder if Israel is a good location to have Intel's design shop, given the continuing violence between the Israelis and Palestinians. It'd suck a P-M designer or two was killed in an attack. :-(

Re:Is Israel a bad location?

[Anonymous Coward]

you do realize the intel fab in israel was built on stolen palestinian land? they bulldozed the village that was in the way and drove off all the people in the area.

they made a ton of p4 cpus, mobile cpus, core logic (chipsets) at that fab.

read about the ghastly story at: http://www.inminds.co.uk/boycott-intel.html !

this is bound to happen

[john_uy]

intel started to introduce model numbers to their processors instead of the actual mhz count. this may be the reason to it.

Intel isn't stupid

[Sebastopol]

Despite what all you teenaged know-it-alls post, Intel isn't stupid. AMD is a very good competitor, and the shuffle of smart people between these two companies has been going on for decades. Intel is huge, it can afford to explore all possible paths. AMD cannot. In fact, no other semiconductor company on earth has the resources to push the limits of design as rapidly as Intel does. Yes, an IBM lab can demonstrate a technology on a test chip, but IBM won't collapse if it can't bring it to market, Intel will.

By having multiple, very different designs cooking at the same time, Intel has wisely used its resources to compete in a rapidly changing market.

I doubt anyone on this board (save a few) could make intelligent decisions with a $40 billion dollar company without 20/20 hindsight.

like the Moto 68010

[Anonymous Coward]

The Inel 80186 and the Moto 68010 both were largely attempts to make a processor easier to integrate into a system. So that you needed fewer support chips. Hence the use in embededded solutions where chip count mattered more.

Both had some real improvements in them, but were very minor. I think the largest was the 68010 had 24 external address lines (allowing up to 256M of address space) instead of the 20 of the 68000 (16M of address space).

Now the 286, it was a disaster. Intel's segmentation (instead of paging) idea of virtual memory just wasn't the way the industry went, so the new capabilities of the 286 were a dead end.

Pentium M "Not Invented Here" Held It Back

[CritterNYC]

Actually, the reason it took THIS friggin long to come to this decision was the resistance on the part of Intel in the US to fully accept the design of this chip by Intel in Israel. Apparently, there was quite a bit of "not invented here" mindset on the part of the US Intel folks, even though it was still Intel that created the Pentium M... just in Israel.

I can't find any info regarding this online at the moment, but I did get this information from a reliable source. Anyone else read this?

Hyperthreading?

[nakedsource]

Does the Pentium M support Hyperthreading, or something new and better?

I've wanted a desktop Pentium M system...

[ChaoticCoyote]

...ever since I saw the first specs for the chip. As compared to the Pentium 4, the Pentium M runs cooler, and is well-suited to multicore chips. I'm looking forward to comparing multicore Opterons against multicore Pentium-Ms.

Parallelism could be important

[ChaoticCoyote]

This is a matter of market preference. The market prefers a single processor (right now). I remember seeing supporting statistics but I can't find those at the moment. It was better than 90%. R&D is currently focusing on making single processors more efficient (Intel's hyperthreading is a good example). All these improvements will eventually trickle down to the small SMP sector Intel supports.

Multicore chips avoid many of the problems involved in trying to increase raw, single-processor MHz. Simply put, a good dual-core system will whip the pants of a single-processor system, even if the MHz is the same between the two systems. However...

...the dual proc system only wins if software is written to take advanatage of parallelism. And therein lies the rub; parallel programming is still something of a dark art, and few programmers really grasp the issue involved. FOSS compilers lack support for simplified parallel programming (OpenMP), posing another barrier to full utilization of multiple cores.

The potential of multicore and SMP systems is largely predicated on training programmers in parallelism, the availability of tools, and the willingness of software companies to invest in parallel software development.

Re:Parallelism could be important

[aksansai]

...the dual proc system only wins if software is written to take advanatage of parallelism. And therein lies the rub; parallel programming is still something of a dark art, and few programmers really grasp the issue involved. FOSS compilers lack support for simplified parallel programming (OpenMP), posing another barrier to full utilization of multiple cores.

I believe this was largely an accurate statement before the development of very well written kernels (Win2K+, Linux, *BSD). Modern kernels take advantage of the availability of more than one processor to off-load the primary processor for user-duties. Also, one cannot neglect the advancement in threading models in the aforementioned kernels. Even though applications may not be specifically taylored for multiple processors directly, the use of threads allows the kernel to distribute the tasks across multiple processors.

Software development using threads will also have to take the next leap in properly constructing method calls to be thread-safe. This provision allows many plain ol' applications to take advantage (albeit maybe not at a peak level of efficiency) of multiple processors with little effort.

The potential of multicore and SMP systems is largely predicated on training programmers in parallelism, the availability of tools, and the willingness of software companies to invest in parallel software development.

We are in complete agreement here. If many developers were to focus on the possibilities of a multi-core chip (or an SMP system), it could benefit even the uniprocessor environment by creating better constructs and honoring memory thresholds to a greater degree.

Re:Parallelism could be important

[Doctor Faustus]

the dual proc system only wins if software is written to take advanatage of parallelism. And therein lies the rub; parallel programming is still something of a dark art, and few programmers really grasp the issue involved.

This is an issue with higher levels of parallelism, and for computers that perform a single task. For desktop use, though, you'll get the benefits of multiple processors just by running several programs.

Nope, just ride on Apple/AMD's hard work

[morton2002]

Actually now that Apple and AMD have successfully?/vigorously argued against the MHz myth, Intel can simply ride on customers' nascent acceptance of performance numbers.

Compared to the incongruous numbering schemes between most consumer electronics, and especially car models, customers will likely accept a marketed performance scheme and forget about any meaning behind the raw numbers.

Intel's new math

[paranoic]

1 processor with 2 cores at 2MHz each = 4MHz chip.

Re:Intel's new math

[bhtooefr]

That's an FSB of... 111.1_KHz...

Re:Intel's new math

[bhtooefr]

Gahhh... bad math... 100KHz...

I was basing it on a 18 multiplier, derived from a 1.8GHz CPU, but a 2.0GHz CPU would have a 20 multiplier...

Almost right

[willy_me]

Size also makes a difference. It's the speed of the tips of the blades that makes the difference. If you have two fans running at the same RPM then the tips of the bigger fan are going to be moving faster then the smaller fan - hence, making more noise.

But the general tone of your statement is correct. It is generally better to go with a slower, bigger fan then a smaller, faster fan.

Intel also innovating

[youknowmewell]

What about that light speed CPU they're developing? Once it hits the market it could change computing forever!

Dothan?

[Angst Badger]

Wow, Dothan is a strange choice for the name of a processor. No offense intended to those who are condemned to live there, but Dothan, Alabama is by a wide margin the goddamned ugliest, most depressing industrial pit of hell in the entire South. It might be the ugliest town in the entire country, but for the existence of Cheyenne, Wyoming, which is appallingly unpleasant to look at even on a moonless night during a power outage.

Re:Dothan?

[ciroknight]

Dothan in the Intel sense is taken from the biblical meaning... something to do with two wells. Don't ask me, ask those crazy Isralies that invented the thing. Either way it's a damned fast chip.

Re:where does that leave performance freaks like m

[jmorris42]

Amen to that. My previous machine was a t-bird 900 on an above average MB. Slightly unstable until I finally threw money at the problem and bought a big-ass, high quality power supply to replace what Enlight tossed in for free with the case. Zero problems after doing that.

This time I started out with a nice heavy, well ventilated case with a beefy p/s filled with another high quality MB, name brand RAM and an Athlon64. I have been running a lot of test builds of whiteboxlinux on this puppy to stress it before building a final x86_64 release and the temp rises a bit after 20+ hours of straight compiling, but it has run rock solid.

Moral of the story, don't buy AMD to save money and expect a system as stable as a quality Intel box. Buy AMD as part of a quality system and it won't be the weak link.

DMV

[steveha]

DMV == Department of Motor Vehicles, where you go to get a license to drive or a registration for your car. This is not a universal term.

I used to live in California, and I've spent a few hours standing in queues at the DMV.

In Washington state, where I live now, we don't have the DMV, we have the DOL (Department of Licensing).

I remember seeing a TV movie, made in California of course but supposedly taking place in Seattle, where someone said "I need to go to the DMV" and the other characters understood. It made me wonder how many places use the "DMV" acronym; if you live in e.g. England, do you say "DMV"? How many states in the US use this term? And does the fact that Hollywood uses the term in movies and TV shows mean that everyone understands it now?

steveha

Re:DMV

[bhtooefr]

Ohio: B(ureau)MV

Re:DMV

[EvilTwinSkippy]

In Pennsultucky, er, Pennsylvania we have PENNDOT. Or simply DOT (Department of Transportation.) They also handle road maintenance and highway construction, so we get to blame them for more lines and delays than the one to get your license.

Hyperthreading on both cores?

[Sgs-Cruz]

With multiple cores, and Intel's hyperthreading, does that mean that my PC will soon 'see' four processors instead of the two I now see with hyperthreading?

That would be pretty cool, and it would have the benefit of making multi-core things more mainstream. This creates more of an incentive for software developers to write things that can take advantage of multiple threads. And that can only be a good thing!

Re:Hyperthreading on both cores?

[DrVomact]

Leave it to me to see the leaden lining of every silver cloud. If you have four processors, then won't you have to buy a server-class Windows license? IIRC, AFAIK, IMHO, etc. etc. the limit of regular Windows OS licenses is two processors.

Re:Hyperthreading on both cores?

[Anonymous Coward]

The Windows XP Pro license supports two physical processors, not logical processors. (A single hyperthreaded chip provides two logical processors.) So a dual-processor hyperthreaded system would indeed be supported by the XP Pro license, and it would appear as a 4 logical processor system.

I've seen XP Pro work happily on a dual Xeon workstation and Task Manager reported 4 processors.

speculation can be taken way too far

[Anonymous Coward]

Has anyone considered that the current heat and power issues with Prescott might be due to some speed path issues causing Intel to hike up core voltage more than origionaly planned to make the 3.2 GHz bin split? There might be a new stepping in a few months with significantly less power consumption. This might have been a stop gap to avoid the embarresment of a PIII/PIV situation when the newer more scalable archetecture being significantly slower to start. There is a reflexive hate the P-4/love the P-M attitude here, which seems to overlook that Dothem was 6 months too late.

Coppermine had a record 13 core steppings (some not released) to eek out tiny speed gains. With Banias not ready yet, the P4 was the right thing at the right time, and it worked . . . well.

It takes years many years to make a chip and even more to design the process. Intel has been planning more efficiency and multi-core for quite some time. PAT Gelsinger was an early evangilist on the heat density issue.

I'm saying a _lot_ less than I know . . . stuff this is out in the public already.

Re:where does that leave performance freaks like m

[steveha]

Concur. I've been using AMD for years, and my computers are stable. I buy good motherboards, good RAM, and good cooling solutions.

The worst thing about the Athlon XP is how easy it is to kill the chip by putting on a heat sink incorrectly. AMD has learned its lesson, and the newest chips (Athlon 64, Opteron) have a metal heat spreader protecting the actual chip. (That's one thing Intel did right with the Pentium 4 and it's about time AMD did it too.) The newest AMD chips are better than the Intel counterpart chips in every way, including price.

steveha

AMD is showing how it is better...

[aksansai]

Intel is an excellent offering for their continued development on their processor families. AMD is showing that even though they don't have the cycles-per-second title, they can produce extremely fast processors at an affordable price. It's another example of capitalism at work. AMD is a force to be reckoned with the industry, and AMD will do what it needs to continue their R&D and profitability.

I agree with you that Intel is the standard because, frankly, they've earned it. While many companies have come and gone to create an x86 compatible chip - Intel has constantly provided satisfactory desktop performance not only through their CPUs but also their motherboard chipsets. It is up to their major competitor, AMD, to aggressively market their product with major computer manufacturers to get their product into the hands of consumers through support contracts, incentives (price reduction in quantity, etc.), and by provided a superior product than Intel's family of processors.

Re:AMD is showing how it is better...

[RESPAWN]

Actually, your mention of major computer manufacturers brings up another good point. Could we see more AMD acceptance in the business world if Dell decided to start shipping comptuers with AMD processors? Dell was the number 1 PC manufacturer for a long time, and it was quite a while after the HP/Compaq (number 2 and number 3)merger before they finally eclipsed Dell in sales. Which brings me to my point that these days I tend to see mostly Dell in the business. Whenever I come across Compaqs they are generally a good bit older. And I almost never see HPs in the business sector. Which makes me think that HP/Compaqs inroads to becoming number one had a lot to do with the fact that they sell a lot to home users through BestBuy. My point finally being, that if Dell, the second largest PC manufacturer, and one of the largest in the business sector, were to start shipping PCs with AMD processors, then maybe AMD would finally be able to make some inroads to business PCs.

Voltages are still going down

[phsdv]

I work for a big semiconductor companie. For our 120nm process we have core voltages of 1.2V. For the 90nm you have a choice, either 1.0V or 1.2V depending how much static leakage you are willing to take.

The max voltage depends on your oxide thickness in your transistor. So you want high speed take the thinner gate oxide, but you will have more static leakage. You want to be even faster, lower the treshold voltage, but you will get more leakage.

Far the fast process option the gate, in a 90nm process, will be more like 65nm. Even more leakage.

All this together will already dissipate a few Watts and the chip is not doing anything yet.

When you want to go real low power, for example for your cell phone, make the gate 90nm, take the thicker gate oxide, increase the threshold voltages, and most important: do not run at high clock frequencies when not needed and you battery will last many hours.

It's used in the HP 200LX

[Solandri]

Probably the best/most rugged palmtop computer for close to a decade. It's basically a DOS machine with 2-4 MB of memory, a ramdisk, option for a flash PCMCIA card, and a bunch of embedded apps like Lotus 1-2-3. Mine still runs and I'd probably prefer it to my Palm if there were a way to replace the display with something more modern and backlit.

HP 200LX technical details

Great...

[davidle]

We might actually get a processor that doesn't take down the National Grid.

It's mismanagement, caused by social breakdown.

[Futurepower(R)]

The AC post above this comment (now at -1) is so typical. Something terrible happens in technology, and, instead of talking about the technology, technically knowledgeable being attacking each other!

The parent post is exactly right: "I'm an EE in Microelectronics, and I had been very disappointed in Intel's tricks to get MHz up."

Intel agrees! The whole point of abandoning the Pentium 4 is that Intel is dissapointed, also. Certainly the engineers at Intel did not know it would happen this way.

The Inquirer article says the same thing: "The heat dissipation/power consumption problems the company has had with the first 90nm desktop Pentium 4, 'Prescott', may well have convinced it that if it's to bring multi-core CPUs to market, its needs to completely rethink the architecture of those cores."

Not very long ago, AMD had 8 to 10 percent of the desktop market, now the share is closer to 50 percent.

Intel has not been doing well. I've had extensive conversations with people who work for Intel, including people who help design Intel microprocessors. In my opinion, it seems that Intel is suffering from years of bad management. Intel has been a company that treated its employees badly, and now the entire company is suffering.

The self-destruction of Intel seems to me to be related to the self-destruction of other U.S. companies, like Enron and WorldCom and the Tyco . It is my understanding that the self-destruction of these companies is related to a general social breakdown that is happening in the United States. I've written a draft of an article about the problems: Social Breakdown in the United States. (This is only the May 7, 2004, 12:46, draft version.)

Re:It's mismanagement, caused by social breakdown.

[Anonymous Coward]

Not very long ago, AMD had 8 to 10 percent of the desktop market, now the share is closer to 50 percent.

For values of 50 == 12, maybe. What crack are you smoking? There are more APPLE MACS in businesses than AMD machines.

Intel vs. Sun

[Anonymous Coward]

OK, I'm confused about something. We read that Intel is ditching several next-generation designs in progress (ones they've already dumped money into) in favor of a current-generation design that's more compatible with multi-core chips. And everyone says that's a good thing and a smart move on Intel's part.

But a few weeks ago, Sun did EXACTLY the same thing, canceling some next-generation SPARC designs in favor of some current-generation stuff that they are the process of making into a multi-core chip. When Intel does it, it's a smart move and good engineering. But when Sun does it, it's obviously just a sign that they are falling to pieces and utterly doomed as a company. (The Slashdot story on this is here.)

Maybe we should reconsider whether Sun's move is really an act of desperation. It's starting to sound like it might be just a reflection of a strong industry trend toward multi-core processors.

Is this the beginning of the end of desktops?

[Anonymous Coward]

It very well may be. Laptops can do everything, and are portable. Most people don't need Geforce 6000s, and even laptop video cards are becoming more capable of running new games. PCI cards are big and clunky, and most PC users still don't want to touch them with a ten foot pole. When kids go to college, what do their parents get them? Laptops, not desktops. Soon desktops will be for hobbyists only.

Now with wireless internet everywhere, who wouldn't want a laptop? When they make laptops that you can upgrade (including sound, video, hard drive, etc), I think desktops will be a thing of the past altogether.

Re:Is this the beginning of the end of desktops?

[obeythefist]

Let's assume that you don't 100% require mobility out of your system (which currently is the only reason people buy laptops, aside from PHB's that believe a higher price tag means a better system).

Desktops are more powerful than laptops. Aside from some lan gamers who again desire portability over power, this makes desktops the computer to use. Go look at a big LAN party. There will be more people bringing desktops than laptops. Gamers, therefore, are definately people who'll stick with the desktop (or erode away to the console). Gamers shouldn't be confused with hobbyists, either. Hobbyists are the ones who cool CPU's with fluid nitrogen gas and move thier systems only on pain of death. Gamers are a much larger proportion of the marketplace.

Desktops are cheaper than laptops. They will be for a long time to come, too. Laptops carry additional costs over PC's, namely they contain all the components a PC does, but they're smaller and miniturisation costs extra. Laptops also contain components that desktops don't need, like PCMCIA cards and batteries and LCD monitors.

I find it difficult to believe that any time in the near future people will decide to pay more money for less powerful computers.

With wireless internet everywhere, there are more and more reasons most businesses, (you know, all the people who buy computers that aren't gamers or mac users), aren't touching wireless technologies. The security risks are insane, and the costs of securing the networks to mitigate those risks are even more insane. Companies will not dump 100MBit CAT5 networks for 54MBit wireless networks (at least none of the companies I've had contact with, and that's quite a few).

Uninformed, not insightful

[Wesley Felter]

Make a 64-bit challenger to Athlon64.

You mean like EMT64 that was already announced?

Enable SMP on something faster than Tualatin.

You mean like the 3.2 GHz Xeon that's already shipping?

Occupation Inside

[Anonymous Coward]

you do realize the intel fab in israel was built on stolen palestinian land? they bulldozed the village that was in the way and drove off all the people in the area.

they made a ton of p4 cpus, mobile cpus, core logic (chipsets) at that fab.

read about the ghastly story at: http://www.inminds.co.uk/boycott-intel.html

The pentium is dead

[DrDNA]

Long live the pentium.

Sounds like BS

[Anonymous Coward]

Tejas is the Pentium 5, not a p4 revision as stated. Unlike Prescott which was massively behind schedule and broke ass, Tejas is fast and has already taped out.

Thats a damn strange time to change plans.

Yay!

[rofthorax]

30% less power and if it has centrino technology, we can clutter the airwaves with microcode at the same time..