AMD and Intel Update CPU Roadmaps

vincecate writes "Recently AMD updated their processor roadmap. It shows their move to 90 nm and has a range of new processors over the next 1.5 years, including dual-core chips. An unofficial AMD roadmap shows speeds and performance increasing. Intel also recently updated their roadmap. Intel does not show anything faster than the current 3.6 Ghz in the next 11 months, including the recently delayed 4 Ghz chip, except to say '3.6 Ghz or greater.' Strangely, some of the recent SPEC benchmark results show the 3.6 Ghz chip to be slower than the 3.4 Ghz chip. One possible explanation for this is that the 3.6 Ghz chips will slow down due to 'thermal throttling' if you are not very careful to keep them cool. So it seems like heat may be the reason Intel's roadmap does now show much improvement."

Comments

Water cooling?

[Cyberax]

Well, why not just make water cooling mandatory for new CPUs, just like Apple did?

Re:Water cooling?

[Animekiksazz]

I've never used watercooling myself, but doesn't it require a bit more maintainance than say your fan cooled processor? I think that might be a problem for some people. Then again I doubt those people will be buying $1000 processors.

I guess though, at some point in the future water cooling will have to be implimented in some form.

Re:Water cooling?

[NoodleSlayer]

This still does not address the massive power consumption problem seen in the P4s, especially the "Prescott" version of the P4s. While heat is a problem, when you're already using +1 lb heatsinks more active cooling is just a temporary fix. Not to mention that Intel, while it can give recomendations it can't ensure that every OEM to produce machines to their specs. I've opened up a number of Dell boxes only to find below-spec power supplies and such.

Prescott in general has had more then its fair share of problems. Prescott is a massive CPU with a 31 stage pipeline, compared to the older P4's 20 and the Athlon XP's 12. I'm not sure off the top of my head how many stages the Athlon 64 has.

All this extra complexity is supposed to make it easier to clock up the processor, and was the same trick Intel used to gain clock speed from the PIII to the P4, so the marketing folks said "Do it again."

Of course the biggest reason why Intel doesn't show many (or any) speed increases is they've scrapped all their future P4/Prescott based designs, even projects that were closed to or already completed because of the problems they have had with Prescott. Intel's plan is to rework their Centrino/Pentium-M core into a desktop chip, but that will take several years.

Re:Water cooling?

[Wytter]

Actually, during a review session with the 3.6GHz LGA775, we experienced so high heat production that we had to use water cooling to ensure that the thermal throttling was not enabled. When using regular air cooling the processor would reach temperatures > 70 degrees during load, and from the results at this load we saw that at some times the processor had to use thermal throttling.

Another disadvantage with this high heat production is that other core components in the computer (such as the mainboard) will be exposed to more heat as well, hence the durability of these components will be lower.

If Intel and AMD continues to approach Itaniums heat production, water-cooling or similiar technologies will become mandatory for high end processors.

Re:Water cooling?

[Anonymous Coward]

For most uses conventional cooling with heatpipes should be quite enought.

Re:Water cooling?

[Anonymous Coward]

While heat is a problem, when you're already using +1 lb heatsinks more active cooling is just a temporary fix.

"Temporary" in this case means at least one, almost certainly three, and quite possibly five or more years. The extra time definitely gives Intel et al. a chance to figure out how to solve this problem without requiring hideously expensive upgrades.

Also, a technical point: the overwhelming majority of heat sinks do not weigh more than a pound, or even terribly close to that. The specification calls for a HSF combo of at most 450g. I know for a fact there are HSFs that weigh more than that, because I have one and it came with a warning about it; but even among aftermarket HSFs targeting overclockers it's very rare. I realize you're exaggerating to better illustrate your point - and I fully expect that within a year or two we'll have passed the 1lb figure - but I wanted to make sure everyone else realized that as well.

If you want to complain about weight, heat, and power consumption, set your sights on video cards. Mine is terrifically heavy and has its own goddamn power connector. The 6800 Ultras need two power connectors. My entire 7-drive backplane needs two power connectors!

Re:Water cooling?

[Jeremy Erwin]

At least Intel and AMD submit their results to SPEC...

Re:Water cooling?

[DJStealth]

Just FYI, 1 lb = 454 g

Re:Water cooling?

[Anonymous Coward]

"When using regular air cooling the processor would reach temperatures > 70 degrees "

So whats the ambient temperature? Do you have your air cond set a lot lower than 70 degrees or do you live in antarctica or somewhere.

Re:Water cooling?

[Anonymous Coward]

Prescott is a massive CPU with a 31 stage pipeline, compared to the older P4's 20 and the Athlon XP's 12. I'm not sure off the top of my head how many stages the Athlon 64 has.

I thought XP had 10 and A64 12..

Re:Water cooling?

[PingPongBoy]

Lordy, you can see a lot in those roadmaps I can't! I wish Intel was nearly so verbose.

Re:Water cooling?

[dfghjk]

"I've opened up a number of Dell boxes only to find below-spec power supplies and such."

Right. Identified by a "below-spec" sticker on them I assume. What's the "and such" specifically?

Using Pentium-M as a desktop processor won't take several years.

Re:Water cooling?

[Dominatus]

He means 70C

Re:Water cooling?

[Anonymous Coward]

No, but getting out a Pentium-M desktop CPU that performs better than the current P4 desktop CPUs might take some time...

Re:Water cooling?

[Rich0]

FYI - I have an Athlon 64 and heat hasn't been a problem at all. I just have the retail processor with the el-cheapo heatsink that it comes with (nothing fancy - just thermal compound and a reasonably-sized sink). I haven't seen it exceed 55C under heavy load. Granted, my case is fairly well-ventilated, but nothing excessive (well, the case was excessive, but I unplugged about half the fans). Oh, did I mention that I overclocked it by about 8% or so?

AMD used to have a high-heat reputation and used to be known for difficult-to-overclock processors. Honestly, I don't think that is nearly as much the case with their newer processors. The Athlon64 seems to run fairly cool, plus it supports frequency scaling when it isn't busy (note - the 55C figure I gave was under heavy load for considerable time - no scaling in effect). Right now, I'm typing on the machine and the CPU is reading 37C - only 1.5C higher than case temperature.

I think AMD is actually passing Intel in this respect. Intel had better watch out if they expect year-long delays - eventually AMD will be releasing 3-4GHz Athlon 64's and they'll be FAR faster than anything Intel currently has...

Re:Water cooling?

[Anonymous Coward]

Just to clarify:

The 6800 Ultra and Radeon X800 XT need two connectors, but they don't consume anywhere near twice the wattage of your card.

They have two connectors because those plastic (and metal, of course) connectors are a really weak spot in the juice line. Consider it a safety measure. They don't need two 12V power lines from the PSU, just two connectors. Actually, you'll do fine with just an Y-splitter from a single Molex cable.

I've seen pics of an overclocked Radeon 9700's power connectors turned from white to brown... scary. (While the card kept working well nonetheless.)

Re:Water cooling?

[Anonymous Coward]

Is it not incredible that you should even have to point this out? Maybe he was joking.

Re:Water cooling?

[Anonymous Coward]

Like a couple of days maybe. I know that 2 Ghz seems like a smaller number than 3.2 Ghz, but you may want to actually read some benchmarks before you make such a statement. The Pentium-M is only slightly slower than the newest Prescott chip. In some apps it's actually faster.

Re:Water cooling?

[multipart]

Prescott in general has had more then its fair share of problems. Prescott is a massive CPU with a 31 stage pipeline, compared to the older P4's 20 and the Athlon XP's 12. I'm not sure off the top of my head how many stages the Athlon 64 has. All this extra complexity is supposed to make it easier to clock up the processor, and was the same trick Intel used to gain clock speed from the PIII to the P4, so the marketing folks said "Do it again."

That's the problem Intel has right now, really. Marketing seems to say, "Make it sound faster", only looking for good warrior CPUs in the Mega Hertz Wars. IBM/Apple and AMD have not been trying to go for faster clock speeds but instead for faster CPUs.

Such long pipelines as the Prescott line may help achieving higher clock speeds, but 31 stages means that you'll see more pipeline stalls, so your CPU is happily running at higher clock rates, doing nothing. Of course, not all instructions actually have to go through all 31 stages, but still, it's impractical to have so many stages in an architecture when you know that every so-many-but-fewer-than-31 instructions you're going to hit a branch. Not to mention the additional complication for the on-die dependency tracking that you need in out-of-order cores like Prescott.

Of course in-order architectures with full predication ISAs would solve some of the problems with longer pipelines, but I guess we can't say that this other Intel architecture, ia64, is such a great success ;-)

Re:Water cooling?

[NoodleSlayer]

A PIII with a power supply rated under 200w.

As for Desktop chips based off of the Pentium-M if you take a look at Intel's short-range road maps you'll notice that they don't have them. Pentium-M right now is geared heavily towards low energy and will take some serious retooling before they can get it up to performance levels that Intel is comfortable with, after all there was a noticable gap inbetween the PIII being replace by the P4 and Intel releasing the Centrino/Pentium-M, a gap of several (being two) years. It will be about two years until you see Pentium-M based chips as serious performance-based options in the Desktop market, that Intel truly is comfortable with, and would be in a postion to replace the current P4s, however flawed they might be. And considering Intel's recent track record on launches and yeilds, probably longer.

Is the processor clock rate trend coming to an end

[Ckwop]

The clock rate of the CPU went up madly through the 90s but the wind appears to have gone out the sails a little. Is the actual speed of the CPU still climbing but they're doing this without adjusting the clock rate?

Don't really keep up on the hardware these days.. :P

Cheers,

Simon.

Whew

[Anonymous Coward]

An unofficial AMD roadmap shows speeds and performance increasing.

And here I was, afraid that they had decided to not increase speeds and performance. That was close.

Re:Whew

[Ari_Haviv]

unlike Intel...

Re:Whew

[Aragorn992]

I think they meant that not only are speeds increasing BUT performance along with it. Intel (for example) increased speeds a while back and didn't really increase performance, i.e. P4 Willamite compared to P3 Coppermine (and the 0.13u version of the Coppermine, can't remember its name). Intel have also increased performance without increasing speeds, i.e. P4 Northwood compared to P4 Willamite.

Green slashdot

[Anonymous Coward]

Color harmony

Re:Green slashdot

[Anonymous Coward]

mod parent up, get rid of that ugly ass color that looks like newborn baby-shit

So I'm screwed?

[reub2000]

I just got an MSI K8N Neo Platnium, which is a socket 754 motherboard. Looks like socket 754 is going no where.

Re:So I'm screwed?

[mondoterrifico]

It has been known for some time that amd was going to go with socket 939, until 2006. The socket 754 will simply move into the budget category, replacing the xp line.
Recent price drops of athlon 64 3000 make socket 754 solutions attractive price wise.

Re:So I'm screwed?

[beakburke]

Actually, even the semprons won't be much faster than your athlon XPs. The 754 is a dead end, even for the budget stuff. The Semprons are going to be on the 939 as well, where they will likely get better performance.

Re:So I'm screwed?

[Mesaeus]

And no worries about heat production either. I just assembled a quiet pc for a customer, containing the Athlon64 3000+. Boy what a cool chip. I used a 80mm 1800 rpm fan instead of the default and still it gets really cool. In fact on the motherboard I used, Asus K8V, there's a Q-Fan technology that lowers the speed of the cpufan according to the temp, and the fan regularly stops completely. And yes, this is by design. When not under load, the cpu temp actually goes UNDER the case temp if you disable Q-Fan and thus let the fan run at its full 1800 RPM (very quiet BTW). On top of that, it's real easy to undervolt the cpu, so you can run it at 2.0 gigahertz (3000+) and lower the voltage from 1.55V to 1.30V, this gives another big bonus in temps. The largest temp I saw while running under full load was 42 degrees celsius, the lowest 32. At the same time Intel has serious problems with heat, AMD seems to have made their coolest chip in years. Anyone else have positive experiences with this ?

Re:So I'm screwed?

[vandan]

I wouldn't worry too much. Although some people talk up the ability to drop in a faster CPU down the track, this is largely either not possible or not worthwhile.

For example, I've got an EPOX motherboard ( can't remember which one ) and an Athlon 2100XP. It's got a 266MHz FSB ( from memory - I may be wrong ). I'm pretty sure if I wanted to, I could put a 3200XP chip into it. But at the moment, I really don't see the point. This runs all games quite well indeed ( partly due to my video card of course ). I don't see much point in upgrading until I can get a 4000XP equivalent ( ie twice as fast as what I've got ). And even then I may wait, depending on my needs at the time ( eg games ). At that point, I really wouldn't consider putting a brand new CPU into this motherboard, whether it would work or not. It would be time to upgrade for a number of reasons, mainly new technology.

So don't look at a motherboard being any more permanent than the CPU that's plugged into it. If you do, you're looking at things the wrong way.

Re:So I'm screwed?

[EpsCylonB]

For example, I've got an EPOX motherboard ( can't remember which one ) and an Athlon 2100XP. It's got a 266MHz FSB ( from memory - I may be wrong ).

Weird I have a 2100 and epox MB, wouldn't be the 8rda+ would it ?

Anyway you are absolutely right, most upgrades are short term temporary solutions, eventually you have the bite the bullet and upgrade teh whole MB, CPU and RAM combo.

Plus if you are a gamer, you really should wait till next year before spending any money, with PCI Express just around the corner anything you buy now will be obsolete in 3 months.

Re:So I'm screwed?

[Fweeky]

Cool 'n' Quiet keeps the CPU ticking over at 800MHz when idle, where it puts out about 35W; hence the fan stopping when idle.

Re:So I'm screwed?

[Mesaeus]

Actually, I've managed to stop the fan even under full load (when Cool 'n Quiet doesn't work). The important part is the temperature, and undervolting the cpu can drop the temp quite a bit. The important thing is that the Athlon64 can be made to run REALLY cool, while it is already a cool processor to begin with. So I don't have to use loud fans to cool it, that's why it's so fantastic for "quiet" pc's. I've lost count of the number of people who have asked me for recommendations to lower the noise of their jet-engine-screaming prescott towers.

Re:So I'm screwed?

[nusuth]

When not under load, the cpu temp actually goes UNDER the case temp if you disable Q-Fan and thus let the fan run at its full 1800 RPM

Are you aware that means either your case temperature reading or the cpu temperature reading is wrong? Fans are not heat pumps.

Re:So I'm screwed?

[Anonymous Coward]

Don't forget that DDR2 is around the corner. When you consider the bus speed difference in a CPU you get now vs. later, different RAM, and PCI Express - worrying about the socket type is small potatos. Pretty much everything is about to change =P

Re:So I'm screwed?

[Anonymous Coward]

I guess the parent ment that the CPU gets cooler than the motherboard. There is no termal readings for the case air temp, AFAIK.

Re:So I'm screwed?

[Mesaeus]

Yep, that was what I meant. Doh !

Re:So I'm screwed?

[ThisIsFred]

The socket 754 will simply move into the budget category, replacing the xp line.

One problem with this, the s754 line stops at the 3500+, the majority of the s754 Athlons aren't available in the s939 package. Right now the best price/performance combination that also has an upgrade future is the s940 Opteron 100 series. Which is really confusing to me. I wish AMD would re-release some of the sub-$200 Athlon 64s in s939, so I can start with something that is worth the price but also has upgrade potential. I'm also looking forward to the day when s939 K8T800 mainboards are sold for less than $150.

Re:So I'm screwed?

[ameoba]

You do realize that CPU temps -can't- be under case temps (when using air cooling) and your sensors are just poorly calibrated, right?

Re:So I'm screwed?

[Anonymous Coward]

I have the AMD Athlon 64 3000+. As I understand it, isn't the Q-Fan and Cool n' Quiet different? I can use Q-Fan on my Linux box, set from bios, but don't believe I can use Cool n' Quiet, which is non-functional in my bios version, anyway.

Not sure of the different, though.

My CPU fan regularly stops, too. Q-Fan has moved the fan from a default of around 5900rpm to around 3200rpm most of the time. I also have a dual-fan Thermalake 420w power supply, which seems to do a good job.

So, undervolting to 1.33v is safe? When would this NOT be a good idea? How can monitor the consequences of this, once I do the change? Any other suggestions for keeping everything cool.

Re:So I'm screwed?

[Anonymous Coward]

re: above post
I'm using the K8V Deluxe mobo.

Re:So I'm screwed?

[nusuth]

Well, there is no way to measure average (or minimum, maximum whatever) temperature of a motherboard either. The thermal diode of the CPU is embedded and literally in the CPU. Thermally speaking, CPU can be approximated as a single homogenous temperature device (at least in the plane perpendicular to CPU to heatsink direction) with its small dimensions and the main mode of heat transfer being conduction.

OTOH for motherboard and case there is just one sensor, located somewhere on the motherboard, poking outward, measuring the temperature around it. The whole structure is too big to have a homogenous temperature distribution. I don't know the location of the sensor for his particular board, it may be located just over some heat generating device like a harddisk or a specific hot chip soldered on the board and his uncalibrated readings can be indeeed correct and overshooting the average case temp. However, in all likelihood, the sensor is strategicly placed to give a more or less good reading of the average case temperature.

The P4 heat problems

[Jugalator]

Strangely, some of the recent SPEC benchmark results show the 3.6 Ghz chip to be slower than the 3.4 Ghz chip. One possible explanation for this is that the 3.6 Ghz chips will slow down due to 'thermal throttling' if you are not very careful to keep them cool. So it seems like heat may be the reason Intel's roadmap does now show much improvement."

Sure, but this is just another form of performance problem. That it takes the form of heat is just a sign of a too inefficient design for the speed. I'm of the opinion that when you get problems to cool the processor good enough (I assume those testers used proper fans for the CPU like they should always do) at 100% usage, then it's not your problem anymore.

Is it just me or are people stupid these days?

[Anonymous Coward]

The reason the 3.6GHz processor runs slower than the 3.4GHz processor is because they're different processors, not the same processor running at different clockspeeds. Just look at the die photos (www.chiparchitect.com) and you'll see what I mean. The idea is that the new processor will scale to higher clockspeeds which it, uh, already has. (Just look at the "OC records": nobody got an old Pentium 4 beyond 4GHz with standard HSF cooling - nobody. On the other hand, this is more or less straightforward with the new Pentium 4s.

What I don't understand is why more people aren't building Pentium M desktops.

Re:Is it just me or are people stupid these days?

[wtarreau]

it's not you, people really are stupid. They didn't even notice that eventhough integer perf dropped between 3.4 and 3.6 GHz, flotting point perf did increase, so it proves that the CPU really has a higher frequency and is not throttling due to heat. Moreover, if it was throttling, it would certainly be by a few percent !

Willy

Re:Is it just me or are people stupid these days?

[FFCecil]

This just goes to show that the old adage "don't judge a processor by its clockspeed" is becoming more and more true. I think AMD has got the right idea by moving away from clockspeed measurements. That's not to say that their system is perfect, but it's definately a large step in the right direction.

Re:Is it just me or are people stupid these days?

[kinema]

"What I don't understand is why more people aren't building Pentium M desktops." It's simple. Intel's megaherts myth has finaly come around and bit them in the ass.

Re:Is it just me or are people stupid these days?

[Technician]

What I don't understand is why more people aren't building Pentium M desktops.

Um, It's the Mobile processor. As such, it's designed to use less power by running at lower voltage and lower speeds, especialy when running on batteries. Why pay extra for a chip that runs slow, and by design slows down on battery power? The Pentium M is a speedy chip, but not at the speeds of the desktop chip. Take a regular Pentium and underclock it if you want to save heat and power.

Re:Is it just me or are people stupid these days?

[hayds]

Being designed for laptops, arent Pentium M chips much more expensive than P4s? Also, can you get ATX mobos for them?

Since its already a cutthroat pricing market, I guess PC makers dont see the need to up the cost of a PC by putting in a Pentium M chip when they can use a P4. Even if the P4 does use more power.

Re:Is it just me or are people stupid these days?

[Ianoo]

Being designed for laptops, arent Pentium M chips much more expensive than P4s?

I can't speak for American prices, but over here we can get a 1.7GHz Pentium M for about 190GBP, the same price as a Prescott 3.2GHz. So yes, there is a definite price premium, but no, the difference is not huge, especially since the Pentium M's are very overclockable, or so I've heard.

Also, can you get ATX mobos for them?

Yes

Re:Is it just me or are people stupid these days?

[Anonymous Coward]

I modded u flaimbait cause that isn't true. The Pentium M is comparable performance to the P4.

Re:Is it just me or are people stupid these days?

[gunnarE]

The reason the 3.6GHz processor runs slower than the 3.4GHz processor is because they're different processors

most importantly the 3.4GHz one on the quoted SPEC page has 2MB of L3 cache, but the 3.6GHz one has none.

Re:Is it just me or are people stupid these days?

[Anonymous Coward]

nice work.

pentium m kicks buttisimo.

Re:Is it just me or are people stupid these days?

[Anonymous Coward]

Also, can you get ATX mobos for them?

Yes [hardwareanalysis.com]

That link points to a micro-ATX board. It has only 3 PCI slots and no AGP. A quick Google search didn't reveal any ATX boards for the Pentium M.

In the case of Mobile Athlons, they appear to be compatible with normal Athlon motherboards. See this link for example.

Re:Is it just me or are people stupid these days?

[Anonymous Coward]

There are a bunch of ATX 479 Socket (Pentium M) Mainboards. They're just not too common.

http://www.radisys.com/oem_products/ds-page.cfm? pr oductdatasheetsid=1158

http://www.nexcom.com/0330/nexweb/weben/ObjView. as px?ObjID=Prod*10000231

Re:Is it just me or are people stupid these days?

[Technician]

OK, I'll bite...

I modded u flaimbait cause that isn't true. The Pentium M is comparable performance to the P4.

I thought the Pentium is out in speeds up to 3.4 Gigahertz. I thought the desktops 3.4GHZ and mobile chips running at up to 2.0GHZ would have been informative, not flaimbait. I guess I unintentionaly started a flame war and so my first comment is flamebait.

A clip from the Intel site regarding the Pentium M..

Processor, Intel® Pentium® M Processor. ... Architecture, 90 nm, 130 nm process technology, 130 nm process technology. L2 Cache, 2MB, 1MB, 1MB. Clock Speed, 1.50 to 2 GHz, ...

URL is here;

http://www.intel.com/products/notebook/processor s/ pentiumm/

I'll try not to flamebait again. It wasn't my objective.

Re:Is it just me or are people stupid these days?

[Chad Page]

It dosen't have to be more expensive... both the .13u and .09u Pentium-M's are actually quite a bit smaller than either Northwood or Prescott even with the -M's larger L2 caches, and therefore actually cheaper to make. Intel's positioning simply allows them higher margins on the -M.

Re:Is the processor clock rate trend coming to an

[Nazmun]

Kind of but not really... There was a time in the nineties where if you waited two years you could get a system at least 2x as fast.

I built my system about two years ago (actually it's a few months short of two years). AMD would have to release the equivalent of 5600+ within a few months to match the speed of the 2800+ they released almost two years ago.

If they were a few months late that would be normal but it looks like it will take far longer.

*sigh*

[SinaSa]

As I wait for the skin to grow back on my eyes from this horrible colour scheme, I can consider the information in the story summary.

We're obviously starting to see a convergence between the industrial processor market and the end-user one. I mean three years ago you would get a dual 3.2GHz (1.6 * 2) system to host a medium sized website, and that kind of horsepower is probably still adequate today. So what kind of apps (I mean, apart from Doom 3) do end users need this kind of grunt for? 3GHz? 3.6GHz? 4Ghz?! If Architects could use AutoCAD 2000 on a 950MHz cpu, without complaint, what has changed? Obviously a speed increase is nice, but three or four times that?

Are we going to see a point where the convergence turns to over taking, and end-user CPU's need to be faster than a lot of corporate stuff?

p.s: I'm aware of shit.slashdot.org, no karma whores please.

Re:*sigh*

[SinaSa]

I just read my own post and realised I forgot to make a point. The point was, maybe the reason Intel and AMD have slowed down on their roadmaps is because we have a over abundance of CPU horsepower at the moment?

Re:*sigh*

[Anonymous Coward]

yep yep.. the games are limited by graphic cards and the pros with their photoshop and maya dont like a box on their desk that imitiates the noise of a starting plane ..

Re:*sigh*

[XennM]

Or there is an under abundance of applications that take advantage of the current CPU horsepower.

Re:*sigh*

[no longer myself]

Actually, there is an over abundance of bloat in a number of applications that more than take advantage of the current CPU horsepower.

Seriously... Even the fastest CPUs can't begin to overpower poorly written programs.

Re:*sigh*

[Anonymous Coward]

If you want to give your eyes a rest, click here, but if you really want to see good taste, click here

Re:*sigh*

[sholden]

As I wait for the skin to grow back on my eyes from this horrible colour scheme

Didn't that skin make it hard to see the colour scheme?

But back to the actual topic...

CPU speeds are just stupid for most people. I write code for a living (at the moment anyway) and the 400Mhz "Mobile PII" I'm using at the moment is adequate for that.

I understand that if your job involved compiling Mozilla or X11 a lot then more CPU (and more importantly more and faster RAM) would make you more productive.

I understand that if you are doing computer generated animation or physics modelling or whatever then more grunt would make you more productive.

But for most people, computers were fast enough a long time ago. This 400Mhz laptop is my fastest machine - both home computers are 300Mhz or so. One of them runs Windows XP Pro just fine, runs Office just fine, runs firefox and IE just fine, runs gimp just fine and even does all of those at the same time just fine.

Every time someone asks me for advice on buying a computer, I ask "do you want to play games?", and if the answer is no then my answer is "buy the slowest CPU model they sell and spend any extra money on RAM".

Re:*sigh*

[Okonomiyaki]

Are we going to see a point where the convergence turns to over taking, and end-user CPU's need to be faster than a lot of corporate stuff?

I think we will actually. If I understand your meaning correctly when you say "corporate stuff" I'm thinking web, file, email servers and so on. Like you said, 3 year old machines are fine for most of that stuff now and will continue to be for some time. On the other hand, the end user is going to be requiring more and more power and not just for games or pretty interface animations. Apple and Microsoft have both been talking about the idea of the PC as a digital hub (well, I don't think MS uses that term exactly because it may be a Steve-ism) for a while. As it becomes a hub for more and more devices it's going to need more power. Loading an iPod with songs is trivial. Manipulating digital photos is a bit tougher. Beyond that you get into editing video and burning DVDs. Encoding and Decoding video. Music creation software. Maybe it won't be long before we see easy to use, prosumer quality 3D animation software...

We've seen a lot of things that used to require very expensive, specialized equipment make their way into the consumer space in the past few years. It's not too hard to guess where that trend may go next. One thing is for sure, it will continue to require more and more powerful processors. Not everyone will need all that power every day but when you get back from that European vacation and you want to do something cool with all the video you shot, you'll be glad it's there.

Re:*sigh*

[Anonymous Coward]

The Mhz Myth has been around for a while, and even some everyday-joe consumers are catching on. In the mean time, I'm typing this on my ThinkPad T21 which has something like an 800Mhz P3 in it. I only use it for Web, SSH connections to my server, p2p and a MS Office, along with tools such as my digital camera and scanner. Oh, and then I also rip loads and loads of CDs on it. Oh, wait, I also watch DVDs and DiVX files and... and... well, I do quite a bit with it. And I'm satisfied. No reason to buy a new laptop anytime in the near future.

I also have a PowerBook (Titanium) with an 800Mhz G4. This sucker gets a much harder load. I do video editing on it. I have loads of external peripherals plugged into it (5 FireWire devices, 4 USB devices) and it really gets a beating. I could use a little more power, sure. But it's not that I'm feeling stressed out over the TiBook's slowness. It isn't slow. I might buy a new machine in a couple years.

So really, I think most users (~90%) are satisfied with the machines they have. The other 10% are either slashdotters, or people that have very specific needs for very specific applications. (One I can think of is the Genetic Algorithm based optimzer system we use at the office, which could use a nice sized cluster if we could afford one. We're in the process of porting it to a 4 processor Itanium2 running HP-UX at the moment, but this has more to do with memory issues with a 32bit OS.) So in most cases, the people that really need the bleeding edge can afford the really expensive (and power hogging, and coolant requiring) systems.

So where does it leave us? I think we're doing pretty damn well right now. Computers are actually affordable now (I could't afford an 80386 in college, when I really needed one to compile my lab assignments) and the standard tools, including power hoggers like video editing, and this is really nice for a change. I actually just trashed my AKAI sampler and Yamaha MIDI synth yesterday because I can do everything I need to via GarageBand and a MIDI keyboard. (Yes yes, I know a lot of people will give me a reason why the AKAI sampler is better, but for what I do it really isn't, and I suspect this is a case for the majority of users.)

On the other hand, I really, really like Apple's new 30" LCD, and than alone is a tempting reason to try and buy a dual 2Ghz G5 system. (And sell a kidney to do so!)

Re:*sigh*

[Anonymous Coward]

> If Architects could use AutoCAD 2000 on a 950MHz cpu, without complaint, what has changed?

Well, AutoCAD 2002, 2004 and ... drum roll ... 2005 have been released. Try running those bloated whales just for the kicks. They are god awfully slow. I'm not saying they are better in any regard, but people will upgrade to them, eventually.

(To be fair, most of the bloat comes from the sky rocketed amount of dll's and arx's they load, so memory and disk speed "help" more than cpu speed here.)

Re:*sigh*

[KingOfBLASH]

. I mean three years ago you would get a dual 3.2GHz (1.6 * 2) system to host a medium sized website, and that kind of horsepower is probably still adequate today

A dual 3.2 GHz P4 would host a web site quite a bit larger than "medium sized." Consider this: for a web server, the vast majority of the computational resources are spent in bandwidth. If you had a number of static pages, you could probably serve up web pages from a single 1Ghz to millions of visitors a day[1]. Anything more than that, you'd need for running scripted pages. But even then, we're not talking about an awful lot of horsepower for web development (unless an incredibly bad programmer doesn't realize you need to max out the CPU to post a form)

[1] Watch when a site gets /.ed. With the exceptions of the articles where someone is running apache on their wristwatch, you usually can still get a small trickle of bandwidth from the web site. Depending on how the web server handles things, it may drop your connection randomly to prevent DOS attacks (OpenBSD). But you'll see the server is still up, it's the bandwidth that is the problem.

Re:*sigh*

[Scarblac]

Every time someone asks me for advice on buying a computer, I ask "do you want to play games?", and if the answer is no then my answer is "buy the slowest CPU model they sell and spend any extra money on RAM".

Another good thing to spend the extra money on is monitor, keyboard, mouse - the things you actually interact with. I've seen people pay through the nose for the absolute latest Pentium, then buy a shitty mouse with it...

These are things you look at and touch all the time when using the computer, improving them improves your whole experience a lot.

Re:*sigh*

[sholden]

A very good point.

In fact I'm using my laptop right now, because the significantly better desktop machine at the in-laws place (which is where I am) has an especially bad mouse (and to make it worse not enough "mousing space" on the tiny desk) - so bad that my laptop touchpad is nicer to use...

Re:*sigh*

[Wetware]

Well... CPU speed is very important to me. One app I tend to spend a lot of time on is iMovie (run on an Apple machine of course, but IBM seems to be running into the same sort of troubles in its transition to 90nm processor manufacturing that Intel and AMD are experiencing) and I am always frustrated in waiting for each title, effect, transition, etc. to render. I don't think that iMovie is a particularly inefficient application. And home video editing is certainly no longer an obscure application. For those doing such work, every extra bit of speed will help.

Re:*sigh*

[Archibald Buttle]

It all depends on what you want to use your computer for.

If all you want from your computer is word processing, web browsing and email access then you don't need much processing power at all.

If, in contrast, you want to do video editing, applying plenty of real-time effects, and decoding/encoding to a compressed format then a high-end dual processor machine is handy.

People are getting used to their computers being able to do things in real-time. Consumer-level applications make a lot of use of real-time processing, since a regular person who doesn't understand about computers is less tolerant of the need that a computer has to process things and expect things to happen instantly.

For example, I just wrote an application which is a modern version of an app I helped write over ten years ago. The app makes graphics files. The original app took its input settings and created the output file when you saved it. When using the app it was common to create several files before you got the output you really wanted. The modern version produces the graphics in real-time and displays them, so the user knows exactly what they're getting before they save the file. Users tend to prefer the new app over the old one, even though right now there's a couple of features missing.

As for seeing an overtaking, I think that depends on the application being used for corporate things. Plain web serving of HTML or file and print serving doesn't take that much power, but video-on-demand is much more demanding.

Re:*sigh*

[sholden]

Clearly that comes under the "grunt work" I mentioned. In fact I'd classify it under "computer generated animation" which I specifically mentioned.

[Aside: I've done video editing (with Premiere, which I suspect was a particularly inefficient application) on a Power Mac 7100 (I think, may have been a 6100 or an 8100...) - we went home for the weekend while it rendered and hoped it didn't crash]

Re:*sigh*

[Anonymous Coward]

Well... CPU speed is very important to me. One app I tend to spend a lot of time on is iMovie (run on an Apple machine of course, but IBM seems to be running into the same sort of troubles in its transition to 90nm processor manufacturing that Intel and AMD are experiencing) and I am always frustrated in waiting for each title, effect, transition, etc. to render. I don't think that iMovie is a particularly inefficient application. And home video editing is certainly no longer an obscure application. For those doing such work, every extra bit of speed will help.

The point is, there's only *maybe* a 10% performance improvement going from 3.2Ghz to say 3.6Ghz. Given today's memory prices, that's usually enough of a price difference that you can put 1GB RAM in instead of 512MB.

For users who open a lot of applications at the same time, the performance increase from not having to swap will far outweigh what they would get from a CPU upgrade. For probably close to 10 years now, I've always instructed folks to go with a slightly slower (and usually *much* cheaper) CPU and use the cash saved to bump the memory up as much as possible. Note that since I used the phrase "cash saved" indicates that these folks have a spending limit.

Buying the absolute fastest processor is generally a waste of good cash, but I'm glad there are suckers like you who are willing to foot the bill.

Re:*sigh*

[Anonymous Coward]

For those doing such work, every extra bit of speed will help.

Which means that you'd have been better served by either a dual-CPU system or even an inexpensive second system to offload work to.

I don't use iMovie (being an x86 shop), but there are currently two big CPU bottlenecks in my flow. Filtering a video stream to prep it for encoding, and the actual MPEG2 encoding. I tend to do my filtering on my primary box because it requires a bit of interaction to get the batches setup. But once they're prepped and ready, they get run overnight.

I use a second box to do the MPEG2 encoding. That system is a slower, older machine that pretty much runs 24x7 doing encoding work as I keep filling the job list with additional work.

Re:*sigh*

[Anonymous Coward]

p.s: I'm aware of shit.slashdot.org, no karma whores please.

Bah, I've been going with I.slashdot. It makes me feel witty.

Re:*sigh*

[Wetware]

Yah, but you implied that most people don't do this sort of thing. I think a lot of people do want or need the speed. Games (for better or worse) are a primary reason for a many a computer purchase. They need speedy processors. Video editing is now a mainstream application, only because consumer computers are powerful enough to support it. Who knows what else will make it into the consumer space as CPU speeds rise?

Re:*sigh*

[kf6auf]

So what kind of apps (I mean, apart from Doom 3) do end users need this kind of grunt for? 3GHz? 3.6GHz? 4Ghz?! If Architects could use AutoCAD 2000 on a 950MHz cpu, without complaint, what has changed? Obviously a speed increase is nice, but three or four times that?

IDL, IRAF, Mathematica, Matlab, etc. In other words, physicsists and astrophysicists can always use faster computers for their everyday work. Even more so, (astro)physicists running fluid dynamical systems of galaxies need every bit of speed possible.

Granted going from 8GB to 16 GB of RAM is the largest benefit but increasing a CPU from 2 GHz to 4 GHz would also help out a lot (though 3.2->3.4 would not be worth it).

I also know some molecular biologists that run detailed simulations about, umm, molecular biology (excuse me but I am not a biologist) and are always looking to get all the speed they can as well.

While scientists certainly do not represent the average computer user, there is demand for fast computers in research.

I hope someone appreciates this as I am giving up my right to moderate but felt that I should mention it because I didn't see any mention of it.

Re:*sigh*

[UnknownSoldier]

I prefer the apple look of this article.

Re:*sigh*

[sholden]

A lot of people might, but I still think *most* people don't.

Of course I would love for the wife to get into video editing, since then I could buy her a computer that actually had enough grunt to run a game more recent than Baldur's Gate...

Re:*sigh*

[Anonymous Coward]

ah, but in 5-10 years that dual g5 will be a worthless doorstop, but your kidney will be very valuable (especially if your remaining kidney goes bad)

Sell someone ELSE'S kidney :)

Re:*sigh*

[roger_and_out]

Not everyone will need all that power every day but when you get back from that European vacation and you want to do something cool with all the video you shot, you'll be glad it's there.

Americans taking a European vacation? I don't think so!

Re:Is the processor clock rate trend coming to an

[opk]

I've noticed this. I got a 1GHz Athlon a few years back and it doesn't seem to be much behind the latest Athlons (especially when I count my athlon's overclocked speed). My previous machine was a 100MHz pentium and that seemed to go out of date really quickly.

Are the new processors really much faster?

Re:Is the processor clock rate trend coming to an

[no longer myself]

It's more along the lines that the clock speed was getting faster, but the performance wasn't proportional. Right now I've got a 2GHz Celeron laptop that is much slower at compile than my Athlon 1.47GHz. (Both have a modest 256MB RAM.) So obviously, clock speed isn't everything.

Don't get me wrong... I still love my Laptop! :-)

The important concept to keep in mind is that all these computers are powerful enough to do what I need them to do, so merely making CPU clocks tick at a higher rate isn't going to persuade me to run out and upgrade.

Re:Is the processor clock rate trend coming to an

[Nazmun]

Well, it's a matter of opinion but, imo... yeah they are.

But then and again your upgrade cycle seems to be longer then mine. I think it all depends on whether or not you want to run some apps that require a more powerful system or actually runs them a good deal faster.

Forget CPU, enter the GPU

[llZENll]

If there was a way to tap into the raw power of our GPUs for normal processing we could see a destkop with 3-10GHz performance right now in some applications and benchmarks. First to come out with an Open Source GPU offload driver wins!

Re:Forget CPU, enter the GPU

[rokzy]

GPUs are not powerful, they are just very efficient at matrix manipulations and calculations related to graphics operations. You can use this to increase performance in some situations, but 99.99% of the time you'd just bottleneck yourself.

Re:Forget CPU, enter the GPU

[pmjordan]

In addition to this, until we start seeing widespread use of PCIe, the downstream AGP bus is still a serious bottleneck as well. Uploading data to the GPU is really fast, downloading maxes out at ~133MB/s.

I haven't had the chance to play with a Pixel Shader 3.0 card yet, so I don't know how useful for generic computation they are. It usually helps if you're trying to process many sets of the same kind of data, rather than evolving one calculation through a long or iterative algorithm.

~phil

Re:Forget CPU, enter the GPU

[Anonymous Coward]

Sick, sad, and ultimately moronic

A GPU is still only a CPU that has been optimized for multicore vector operations
eg: a GeForce 6800 is approx 10 programmable pipelines, with some entangling fixed function pipes, such as triangle setup, cache, memory and IO, etc.
6 of those cores are independant vertex processors
4 of them include 4 "pixel pipes" each, which, unfortunately, aren't actually independant, they just include writemasks to prevent unused ones from affecting what they shouldn't. if one pixel pipe slows down, the rest slow with it. Each of these pixel pipes has a 2 stage pipeline, much like the early, early CPUs(im thinking pre-pentium), and 2 execution units, which each consist of a 4 floating point vector processor with inline integer math for addressing, filtering, etc.

so basically it is 10 processors that can handle 20 instructions per clock with an average of 176 data outputs per cycle(22 "pixel/vertex pipes" * 2 execution units * 4 components per vector)
Or 17.6 data outputs per processor per cycle

compare to a prescott
1 processor, 31 stages, 1 pipeline, 4 instructions per cycle, up to 4 components per vector(SSE) = 16 data outputs per cycle, 8x as many cycles per second, 1/2 as many transistors = 128 data outputs requiring half the silicon.
therefore, a dual core prescott could easily do (256/176=) 45% more work at the same price and:
1) have a decent level of precision(64 bit)
2) have integer math
3) be much more flexible and not rely on the AGP bus
4) be a hell of a lot easier to code for

what are you whining about?
consider the concept processor, Niagra, an 8 core, multi-ghz, 3 instruction per clock, 4 way hyperthreading processor, undoubtedly with vector extensions...who needs a GPU for some shitty 32 bit, AGP-bottlenecked results when the CPU is obviously superior?

Re:Forget CPU, enter the GPU

[BlowChunx]

In addition to this, until we start seeing widespread use of PCIe, the downstream AGP bus is still a serious bottleneck as well.

I thought this was an AGP driver issue, not a hardware limitation? Who knows, maybe the same guy will write the PCIe driver and history will repeat itself...

Re:Forget CPU, enter the GPU

[EuropeUnited]

The AGP bus is only limited to 133Mb/s when running in 1x-mode (32bit, 33mhz). The AGP found in computers from this millenia are quite a bit faster, with both DDR and much higher clock frequencies.

In more recent computers, the AGP is running at 8x, which means theoretical maximum speeds over 1000Mb/s.

Besides, no matter what speed you're running the bus which your graphics are sitting on, going beyond local card memory will always be slower due to latency and such.

Re:Forget CPU, enter the GPU

[idiot900]

consider the concept processor, Niagra, an 8 core, multi-ghz, 3 instruction per clock, 4 way hyperthreading processor, undoubtedly with vector extensions...who needs a GPU for some shitty 32 bit, AGP-bottlenecked results when the CPU is obviously superior?

So you can use the CPU for things other than rasterizing scenes?

Re:Forget CPU, enter the GPU

[pmjordan]

Nope, you're mistaken, the >1GB/s transfer is only from Northbridge/System RAM to Graphics card, the other direction is as slow as a normal 33MHz, 32bit PCI slot. This is one of the reasons why PCIe is catching on so fast.

~phil

Re:Forget CPU, enter the GPU

[pmjordan]

It's a hardware issue, as the AGP bus is actually a tagged-on second PCI bus. Graphics Card->Northbridge runs at PCI speeds, Northbridge->Graphics Card runs at whatever AGP revision you have. (1x, 2x, 4x, 8x)

~phil

Re:Forget CPU, enter the GPU

[Anonymous Coward]

> compare to a prescott 1 processor, 31 stages,
>1 pipeline, 4 instructions per cycle, up to 4
> components per vector(SSE) = 16 data outputs
> per cycle, 8x as many cycles per second, 1/2
> as many transistors = 128 data outputs
> requiring half the silicon.

You've got a number of facts wrong. First of all, there is only 1 SSE unit on Prescott, and it only does half a 128-bit vector each cycle (common misconception!) Therefore, it outputs two single precision float results (NOT multiply-add like GPUs/PowerPC/IA64, just one or the other each clock). At peak bandwidth, this means 3.6 GHz * 2 floats/cycle = 7.2 Gflops.

By your math, assuming 4-wide vectors x 2 units x 2 ops (mul+add) x 10 pipes x ~500 MHz = 80 Gflops for the GPU (assuming full memory bandwidth). How is the P4 better again?

Yawn...

[Anonymous Coward]

Wake me when AMD or Intel realizes that with the same amount of silicon, they could have dozens of Pentium 3s/Athlons at close to the same clock speed, providing much better performance at a much lower R&D and silicon cost

Re:Yawn...

[Xilman]

Wake me when AMD or Intel realizes that with the same amount of silicon, they could have dozens of Pentium 3s/Athlons at close to the same clock speed, providing much better performance at a much lower R&D and silicon cost

And much higher software development cost though, to be fair, most of that would not be done by AMD or Intel.

I've written code to run on medium scale clusters and have made very good use of them. On the right sort of application, a 32-cpu cluster of PII-300 machines beats a single 3GHz P4 hands-down and a 32-cpu cluster of PIII-1000 blows it completely out of the water. I've extensive experience with both clusters. Unfortunately such applications are very much a minority interest. My interest is in computational number theory. There are a few others, but most people just can't make cost-effective use of more than a few processors.

Also unfortunately, writing code for medium sized clusters is not easy and it can be very difficult to find any parallelism in a task let alone exploit it efficiently. If it were easy, we would have seen clusters in widespread use for several decades now.

Paul

Re:Yawn...

[dragonp12]

How easy it is to say stuff like that...

But of course, if they had loads of Pentium 3s etc running at the same clock speed, there would come a point when there was nothing more to be got out of that processor. More R&D is always good.

Re:Yawn...

[Anonymous Coward]

You've hit on way I am afraid. Multi-core processors are useless for existing apps that are single-threaded. Not to mention algorithms that cannot be calculated in parallel will never benefit(is there a term for algorithms of this sort?). I think a major theoretical breakthrough is needed real soon.

I am hoping that someday compilers will be smart enough to infer parallel computability but I think this is at least a decade off. A new language, not based on C/Java, that addressed this would be nice.

Re:Yawn...

[Xilman]

Not to mention algorithms that cannot be calculated in parallel will never benefit(is there a term for algorithms of this sort?).

"Intrinsically serial" is the phrase I've seen used for such things.

Paul

I'm starting to worry

[rokzy]

I have always thought AMD is better than Intel (price/performance, no annoying jingle, no annoying "... inside", no "MHz myth"), but now it seems Intel is getting its arse kicked so much I worry AMD might get too complacent.

Re:I'm starting to worry

[lakeland]

How about we wait until they're getting close to 50% market share _before_ we start to worry about AMD turning evil? AMD has a better product, and they have had a better product for _years_. Thanks to their better product, and our support, they have very slowly being gaining market share. However, they are still very much the underdog. If we withdraw support now and go with something else (transmeta?) then I would guess both AMD and transmeta would end up as minority players.

Re:I'm starting to worry

[Anonymous Coward]

AMD hit +50% desktop market share for the first time a few months ago. I haven't heard if they maintained it, and they only have 10-20% in the huge notebook and server spaces. Server marketshare is growing quickly because of the Opteron though.

Re:I'm starting to worry

[Anonymous Coward]

Thats only for those who build computers (I forget the exact term) not for desktops in general. The market where AMD finally pulled ahead is the smallest peice of pie by FAR when looking at the overall market. Intel is still selling more by a pretty large margin.

Re:I'm starting to worry

[Anonymous Coward]

AMD hit 50% for one week. The Q1'04 market share numbers are 15% desktop, 8% mobile, 5% x86 server Just because we think AMD should be kicking ass doesn't mean it is so. There is alot more that factors into the equation than performace numbers or design.
I doubt AMD could even reach 50% in the near future, since they would need to add large amounts of manufacturing capacity.

Re:I'm starting to worry

[Ziviyr]

Well, since intel's faster chips are running slower than their slower chips, AMD actually has a 64 bit product out, and theres still plenty competition towards speed boosting features. Oh, and AMD isn't charging limbs for their CPUs yet.

I'll worry about complacency when the threat of stale tech pops up.

Clock speeds seem to have stalled.

[arcade]

More people than me seem to have noticed that clock speeds seems to have stalled. I don't necessarily see this as a bad thing - as computers has grown fast enough for me lately. I'm still content with my 1.3Ghz Duron.

What I personally really, really want to see is cooler CPU's. CPU's that doesn't require a huge fucking fan. CPU's that are content with a heatsink would be nice.

Furthermore, I would love it if Dual configuration became more widespread (and thus cheaper). Personally I would love a multi-CPU machine far more than single-CPU ones.

My personal wishlist:
- 64bit CPUs to become the norm (seems to be happening).
- Cooler CPUs, not requiring fans (seems to be happening, look at the VIA EDEN CPU's)
- Dual/Quad/Multi -CPU configurations becoming the norm in home computers.

I don't care much whether single CPU's grow much faster at the moment, as there doesn't seem to be applications requiering it for regular use. There are of course specialist tasks that require more horsepower, but those are .. specialist tasks.

Re:Clock speeds seem to have stalled.

[Anonymous Coward]

- Get yourself first cheap Athlon XP you can see
- downclock and undervolt
- result: cool CPU

Re:Clock speeds seem to have stalled.

[10Ghz]

My personal wishlist:
- 64bit CPUs to become the norm (seems to be happening).
- Cooler CPUs, not requiring fans (seems to be happening, look at the VIA EDEN CPU's)
- Dual/Quad/Multi -CPU configurations becoming the norm in home computers.

You can have those, just not at the same time. Via Eden runs fanless. But it's still 32bit! And it doesn't run in SMP-configurations (yet. there has been some info about SMP-solutions).

I think you could buy an Opteron 2xx-machine, underclock it to around 1GHz so it might run fanless. Then you would have your fanless 64bit SMP-machine,

Re:Clock speeds seem to have stalled.

[dj245]

What I personally really, really want to see is cooler CPU's. CPU's that doesn't require a huge fucking fan. CPU's that are content with a heatsink would be nice.

Well I don't know about fanless, but if you're really after a quiet computer, and don't care at all about speed, there's always underclocking. I bought an AMD 2600+ for my dorm room PC and underclocked it (gasp) to 2100+ speeds. The Shuttle power supply has a small 40mm fan, but the only other fan in the case is a Zalman 80mm fan with a big fat resistor running in series with it. With that fan turning at barely 1600 RPM, and the processor at 114F under full load, you can barely tell its on.

Almost all computers need some ventilation (even most of the supposed "fanless" PC's) but if you can run them near silently, then the only thing you have to worry about is a fan giving out.

Re:Clock speeds seem to have stalled.

[arcade]

Some people here doesn't seem to realize that what I'm stating as my personal wishlist isn't what I want _today_, it's where I really wish companies put their research and development.

Re:Clock speeds seem to have stalled.

[Tim C]

There are of course specialist tasks that require more horsepower, but those are .. specialist tasks.

Once upon a time, viewing images was a specialist task. Then, viewing them in full colour, then editnig them. Sound used to be a specialist task, then editing sound.

Now, things like video encoding/editing is a specialist task, requiring (relatively) serious amounts of horsepower.

Well, once that sort of horsepower becomes commonplace, the task stops being specialist, as more and more people do it simply because they can.

True, there will always be truly specialist tasks that never become mainstream (animation and rendering work, phsically simulation and similar number crunching), but there is stuff now that could most certainly benefit from more CPU power (whether it be from single- or multi-cored machines) that would become more mainstream when that power became affordable.

Re:Clock speeds seem to have stalled.

[denobug]

In my line of work one of our main servers has dual processors. It is having a real hard time publishing the data coming from remote locations fast enough. Processors are on a slightly lower spec. we're constantly having one processor running at 100% and locks up the I/Os of the computer.

It is simply not true anymore that "any" coding work and database processing work can be done on just "any" computer. That is a myth that peole has been spreading. They are not claiming without some truth to it: Many tasks' minimum processing requirement has been met long time ago. I have an old Athlon 800 MHZ desktop sitting at home next to a P4 system and it runs just fine if all people does is using Word and surfing the net. What I don't agree with is the misconception that the processor world can rest on their laural and not improve the processor's processing power (not necessarily the clock speed, enought heat already!)

on an off-topic notem, anyone remembers the day when a passive heat sink does the job without the fan on your CPU?

Re:Clock speeds seem to have stalled.

[dfghjk]

Curious that you see no requirement for a single CPU to get faster yet desire to see multi-CPU configurations become the norm. That doesn't make any sense. Two half speed CPU's are a liablity compare to a single full speed one though a system designed in that manner may be much easier to build.

The future may be MP via multicore processors because that's the easiest path to superior performance. It won't be because there's no demand for a faster single processor though.

Processor manufacturers have to compete with something and so far it is raw speed. VIA does it on cost, small dies, and low power but their popularity is limited. That's a real market proof that fan-less has limited appeal unless the performance is truly competitive.

Re:Clock speeds seem to have stalled.

[PingPongBoy]

Why can't CPUs run so slow during idle periods their fans are off? When I want to watch a movie or play some music the CPU should not be used - there are lots of tasks that don't really need CPU although the CPU option should be there for fine control. Common tasks like moving a window, etc. should be learned by special circuitry and not require CPU work. Then CPUs can run at several speed levels, which can be controlled by users who want cool laptops or satisfied deadlines.

Re:Clock speeds seem to have stalled.

[Anonymous Coward]

That's exactly what AMD's "Cool 'n Quiet" in Athlon 64 models does. When the CPU notices it's not under full load, it downclocks to 800 MHz where it outputs about 35W. Most Athlon 64 motherboards have temperature controlled fan connectors, which slow the heatsink fan down accordingly, making your late night movie watching very quiet :)

They got the clue already :)

Re:Clock speeds seem to have stalled.

[ameoba]

Where have you been, and where do people find these crazy screaming HSFs for stock-speed CPUs?

The first thing to do is stop buying cheap cases/power-supplies. It costs money to make things quiet; extra sensors to throttle fan speeds & quality bearings add a few $$$s to the cost of a unit.

With that said, I just threw together an Athon64 3200+ with the stock AMD heatsink into one of these and I couldn't hear it over the ventilation system. This is under full load, not sitting there idle with some sort of CPU throttling in place. It's not quite the fanless but it definately runs more quietly & cooly than your Duron 1.3.

If only specialists need faster CPUs, what's the justification for pushing SMP machines into home machines? I can't see the average home user needing multi-processor systems any more than they need 3MHz CPUs...

Re:Clock speeds seem to have stalled.

[0111 1110]

In a recent article, anandtech reported that Shuttle is planning to release a SFF desktop platform for the Pentium M by the end of the year. A 1.1 Ghz ULV Pentium-M is supposed to dissipate a scaldingly hot 5 watts. Even at 2 Ghz it is only supposed to dissipate 21 watts.

Considering my 1.13 Ghz PIII-S dissipates 29.9 watts, I'm pretty impressed. I could keep my current clock speed and only use 5 watts. That Shuttle is going to be my next upgrade. Hopefully the Pentium M chips will have dropped somewhat in price by then as well.

Re:Clock speeds seem to have stalled.

[kylemonger]

True, there will always be truly specialist tasks that never become mainstream (animation and rendering work, phsically simulation and similar number crunching), but there is stuff now that could most certainly benefit from more CPU power (whether it be from single- or multi-cored machines) that would become more mainstream when that power became affordable.

Animation isn't going to be a specialist task forever either. The porn applications of it are obvious. One example: remember how Celebrity Deathmatch was popular on TV? Imagine Celebrity Sexmatch, rendered right on your home PC. You choose the celebs, locale, camera angles and sex acts performed and it is rendered with ILM quality in a few hours.

In general, rendering realistic 3D locales has interior design uses. Imagine if your designer could show you exactly what the inside of your house would look like instead of the dinky models they use now. The designer could put you inside the model with good VR so you could see if the lighting made you look bad or if the sun blazing though the windows at certain times of the year would diminish or enhance the enjoyment of your favorite room.

Re:Clock speeds seem to have stalled.

[MarkCollette]

I think we need to realise that sometimes a task requires specialists because the human skillset is specialised, and sometimes because the required tools are beyond the reach of most people.

So, if a task is doable by most people, then increasing computational power will help bring that ability to the masses, and will drive computing innovation. These tasks should not be used as an argument against the need for faster/better computers.

Re:Clock speeds seem to have stalled.

[arcade]

on an off-topic notem, anyone remembers the day when a passive heat sink does the job without the fan on your CPU?

I, then guy you're answering too, remember. I also remember when I didn't even need a heat sink.

What I really want is for computing to return to the days of the heat sink. No movable parts for cooling is a good thing. Less maintainance.

I don't see any reason to return to the pre-heatsink days though, as a heatsink isn't a moveable part. :)

Re:Clock speeds seem to have stalled.

[arcade]

Curious that you see no requirement for a single CPU to get faster yet desire to see multi-CPU configurations become the norm. That doesn't make any sense. Two half speed CPU's are a liablity compare to a single full speed one though a system designed in that manner may be much easier to build.

It makes lots of sense.

With one CPU, one task typically occupies the entire CPU for an an entire timeslice, however that is defined by the kernel at hand.

With two CPU's, two different tasks may occupie the different CPU's, running in parallell.

If only one of the two tasks is CPU intensive, the entire system is still extremely responsive, as they've got an idle cpu there ready for use.

Re:Clock speeds seem to have stalled.

[Anonymous Coward]

Common tasks like moving a window, etc. should be learned by special circuitry and not require CPU work

Is this a joke?

Re:Clock speeds seem to have stalled.

[arcade]

If only specialists need faster CPUs, what's the justification for pushing SMP machines into home machines? I can't see the average home user needing multi-processor systems any more than they need 3MHz CPUs...

Doing many tasks at the same time. One music player, one movie player shoveling a movie to the TV in another room, some CPU intensive work for the guy sitting at the computer, some this and some that, while the system still being responsive.

With a single CPU, this takes a lot of timing, prioritizing for interactive applications and a lot of _crap_. With several CPUs the CPU intensive tasks gets their own CPU, while the less intensive tasks just sits on a mostly idle CPU.

Re:Clock speeds seem to have stalled.

[TheLink]

That's just a scheduler problem.

If you don't mind HALF your processing being wasted, just get a scheduler that _never_ allows a single process/thread to use more than 50% of CPU. Next reduce your time slices to something higher than 100Hz. Given multi GHz CPUs, even switching every 2KHz may not be a problem.

Then it'll behave very like a dual SMP machine with 2 CPUs at half speed.

That said, I wonder if SMP machines still lock up and wait for the CDROM drive for Windows machines?

Re:Clock speeds seem to have stalled.

[denobug]

So true.

I have seen some small cases with the modified heat sink transfering the heat from the top of the processor to the back of the case. That seems to be a pretty good idea. I think at one point Intel's Xeon processor was (or plan on) using this kind of technology. Since I never work with one (besides what Intel's manager's presentation) I don't know if it is actually used by Intel.

That plus some re-engineerin of the processor should get us to a no fan design. Hopefully someone other thank Apple thought of this.

Apple's premiums can handle the extra cost...

[michaeldot]

...but maybe the cheaper PCs cannot?

Also, a liquid cooler is probably inherent harder for Intel to package with an OEM processor. Affixing a liquid cooler to a processor requires more case aware design than simply clipping a fan to a mainboard socket.

Opteron, Linux 2.6 and Java 5 benchmark

[gregluck]

Last week I benchmarked the 2.2Ghz Opteron on 64 bit Linux 2.6 and Java. I got almost three times the performance of a 3Ghz Xeon. For details see http://gregluck.com/blog/space/start/2004-07-29/1# AMD64,_JDK1.5.0_and_Linux_2.6_rock!/

Re:Opteron, Linux 2.6 and Java 5 benchmark

[Anonymous Coward]

They didn't use the same JDK for both systems. This makes it an uneven test.

Re:Opteron, Linux 2.6 and Java 5 benchmark

[khuber]

True, he should go back and test 1.5 on the Xeon, but it's probably not a dramatic difference. The specweb results do verify that a dual Opteron 248 handles about twice the workload as a dual Xeon 3.06 though. The Xeon CPUs are a bit cheaper.

Re:Opteron, Linux 2.6 and Java 5 benchmark

[vincecate]

Can you run it again with JDK1.5.0 32bit on Xeon to see how different it is with same JDK version? Also, I assume you were running 32-bit Linux on the Xeon of the same FC2 release used on AMD?

I tried to post on your blog but could not see how to get a username/password. You should have a link to "make new user" on your login page.

Re:Is the processor clock rate trend coming to an

[Anonymous Coward]

I've got a 2GHz Celeron laptop that is much slower at compile than my Athlon 1.47GHz.

The Celeron is a severly crippled chip, unlike the Duron, which is a respectably performing budget processor. It only has 128KB cache, which is CPU sucide on a P4 core. The P4 needs large amounts of cache to keep its long pipeline filled. People who buy high clock speed Celeron, thinking they're getting a fast CPU are getting massively screwed by Intel. So much so it borders on being an unethical and immoral business practice. The chips are not near as fast as their clockspeed would indicate. One would be much better of with an Athlon XP, Duron, or a slow P4 as a budget processor.

Where is the roadmap I want?

[MancDiceman]

Where is the roadmap for low-power consumption chips that can operate either fanless, or with low less cooling gear?

I survived just fine on a PII for several years until recently biting the bullet and getting myself a P4 box in a Shutttle Zen XPC case (relatively quiet). I seriously considered getting myself an EPIA box as my main machine, simply because it would be lower power (therefore cheaper to run), silent and enough umph to use mutt, firefox and ssh into the server kit where the real work is done. The only reason I ended up with a P4 is because a friend had a 3GHz one going very cheap.

I want less power, not more. The idea I should overclock, buy liquid cooling systems and should pay a ridiculous amount so I can play some games? I'm sorry, what planet are you all on?

Re:Where is the roadmap I want?

[eRacer1]

Where is the roadmap I want?

Here is one. A 25W mobile Athlon 64 should be a good candidate for a low power desktop system with very modest air cooling.

Re:Where is the roadmap I want?

[PingPongBoy]

Many low power computers available now as handhelds. Can they be docked to a regular size keyboard and screen? It isn't all the time that one needs a CD, speakers, etc. Wireless networking will eliminate an outside port (though I don't like holding all that wireless in hand as there is speculation of a cell phone cancer link).

Re:Where is the roadmap I want?

[Ritontor]

I know it seems crazy, but it's not just the current batch of games that want you to have big, power-hungry cpu's. As the days move on, EVERYTHING will get bigger and smarter, until we reach a point where AI is king - human level intelligence on (presumably) silicon. Reckon your PII is going to keep up then? The drive for greater computing power far exceeds the limited requirements of the current moment in time. While you, personally, might not want to play new games right now, i'm pretty sure you'll be interested in running basic AI expert systems in 10 years time...

Re:Where is the roadmap I want?

[taj]

I recon Intel wont have the 4 Ghz CPU's out by then.

Your thinking is very vector like with high language teaching from schools being the target. In reality there are other directions CPU's are going too. If the original poster is building a gateway with no moving parts including CPU fans, hard drives, ... The machine may well be the gateway for the trendy AI expert systems in 10 years and still be working fine.

Z80, 80C52, ... are more than enough for many applications. But in these applications you do often concern yourself with power.

Hand held gps units do not need every feature possible. They need long battery lifes.

A frig manufacturer does not want to change its electric use rating because someone said AI is cool just so they can have it report failures over the internet. Computers with monitors represent a minute fraction of all computers out there.

Re:Where is the roadmap I want?

[Chad Page]

I don't have one (yet) but the desktop Athlon 64's have various low power modes. The mobile ones are much nicer still, but they don't work on that many motherboards aparently, alas.

Re:Where is the roadmap I want?

[0111 1110]

True AI is always 10 years away. The hard truth is that it will more than likely be 100 years or more. Even the youngest slashdotters will not live to see it.

AI will likely only move forward along with advances in robotics and bio-connectionist systems. Traditional programming doesn't lend itself well to the task.

"Expert systems" is an unfortunate term. I prefer "knowledge based", but even those programs have proven impractical. The most ambitious attempt has been Doug Lenat's CYC project.

Faster processors certainly won't hurt AI, but the answers do not lie there. Of that much we can be sure. As with space travel, we are all waiting for some kind of fundamental breakthrough.

Maybe it will be some combination of robotics, human and experiential trained (maybe wetware) connectionism, along with some CYC-like rules.

Re:Where is the roadmap I want?

[DirePickle]

I suspect that we are on the planet where some people are gaming enthusiasts, and are willing to toss that much money into a computer. Some people are into cars, some computers, other folks do other things. No one's saying you have to.

Re:Where is the roadmap I want?

[arcade]

I know it seems crazy, but it's not just the current batch of games that want you to have big, power-hungry cpu's

Except for a very few specialist areas - yes, those are the only applications - and not even the current batch of games seem to rely mostly on CPU power, but on GPU power.

As the days move on, EVERYTHING will get bigger and smarter, until we reach a point where AI is king

We'll need a breakthrough in AI first. I doubt it will come very soon.

The drive for greater computing power far exceeds the limited requirements of the current moment in time.

I hear some people (like you) claim that. I don't agree. My computer (Duron 1300) is currently fast enough for me to have xmms in the background, several downloads going on, a mail client with tens of thousands of emails accessible, a web browser, a compilation of KDE 3.2beta2 in the background, and a gazillion of xterms lying around.

Oh, and I've got a good old Matrox G450, so I actually run a dual-monitor configuration.

The point is that most of the time, the AMD Duron 1300Mhz is more than fast enough for me. There is days where I could really use a little bit power, say, 20% more or so .. but considering that this i a pretty old system, and that the top of the line models these days are at least 3-5 times as fast as my current machine.. heh. :-)

There are of course CPU intensive tasks that will enter into every day use. Tasks like video editing, which currently is quite acceptable on top-of-the-line CPUs. Also, video encoding could probably be done quite a lot faster.

However, we're now talking very, VERY cpu intensive tasks. We're talking about tasks which really isn't interesting for everyone. Tasks that regular home computers now are able to fullfill.

The point is that regular people don't NEED much more power. Earlier on, we had to sit and wait for the computer to complete various tasks. These days, the tasks are just done 'whammo'.

Of course, I wish the compile time of KDE 3.2beta2 would be a bit quicker. It's not important for me to have it finish in less than a second though .. and it sure as hell isn't important for my mum, my dad, or 99% of the computer-using people that such specialist tasks finish quickly.

Computers are approaching the needed speed. The NEED to get them faster isn't there any more. They're actually getting _fast enough_.

Re:Is the processor clock rate trend coming to an

[Amiga Lover]

The clock rate of the CPU went up madly through the 90s but the wind appears to have gone out the sails a little. Is the actual speed of the CPU still climbing but they're doing this without adjusting the clock rate?


I'm hoping that because of this 90nm barrier (or pause, what have you) that the advent of dual core chips actually comes around this time. There have been many promises & comments from the G3 750FX a few years ago up through to today, of chip manufacturers turning to Dual Core CPUs.

I'd rather have one large chip with two 3.6GHz cores than one chip that attempts to get out a mere 400MHz more for 4GHz being delayed again & again.

Of course, there's always dual CPU motherboards...

Re:Is the processor clock rate trend coming to an

[Anonymous Coward]

One would be much better of with an Athlon XP, Duron, or a slow P4 as a budget processor.

It's a budget laptop. You don't get a lot of CPU choices when you're looking for a new laptop that is Linux-friendly, sub-$900(US) and comes with a DVD/CDRW. It does its job, so I'm not complaining.

On a brighter note, I can upgrade it to a P4... But I'd rather get it fitted with a working 802.11g card so I can surf Slashdot while away from my desk.

Ahhh... I can just imagine sippin a hot black cup of coffee while munching on a tasty pastry in an overpriced internet café. The fantasy also includes a really hot babe who is just overwhelmingly turned on by my savvy Linux laptop, and excellent karma... Oh wait. You people really don't need to know all this.

Re:John Kerry's Monstrous Record on Civil Libertie

[Anonymous Coward]

Err.... eh?

Why not use squidguard?

[FyRE666]

Just add a rewrite rule to your squidGuard config file as so:

rew slashdot {
s@it.slashdot.org@slashdot.org@
}

and you'll never have to see the yellow scheme again. In fact you could add it for any scheme you don't like on Slashdot... Simple!

Apple' ROADMAP? (interesting)

[Anonymous Coward]

Motorola adds dual-core G4 to PowerPC roadmap (June 2003!!! Today failed!!!)
New imac in September
Apple's problems.
Mac Rumors: PowerPC RoadMap from Motorola ...
Mac Observer: PowerPC RoadMap from IBM ...

open4free ©

This shows a benchmarking problem

[RhettLivingston]

Normal applications are not as likely to drive a processor into thermal throttling as a benchmark is. It sounds like benchmarks are going to need to be rewritten to either be short enough to not cause thermal throttling or to spread the benchmark out so that the CPU has a chance to dissipate heat buildup caused by the artificially intensive benchmark code.

Re:This shows a benchmarking problem

[DMUTPeregrine]

This isn't that funny. Most people are running Word, Excel, and IE. Possibly Solitaire and AIM. They don't need all the CPU power, so the benchmark is useless. Anything better than a 400mhz Pentium II is probably enough, if not excessive.

Re:Is the processor clock rate trend coming to an

[vadim_t]

It's a funny thing that I might actually have been right a few years ago. I posted here on slashdot a little calculation I did that showed that as clock speed goes up, the speed of light becomes a very limiting factor.

IIRC, at 5 GHz, clock cycles would be spend simply on waiting for the data to propagate through the path between the RAM and the CPU.

While I don't really know about CPU design, I'd say that this is a possible reason for stopping clock improvements, since as speed gets higher, latency would keep growing.

How Fast is Fast Enough?

[Gigantic1]

Sorry, but I look at these speed numbers and say to myself, "Who Cares?" I can't speak for those of you who are serious Doom Players, or for those who like to process video, or mathematically search for new planets, but a while back I believe I saw something interesting: PCs had become fast enough for most people. What do I mean? Well...I was working on a Telecom Development Project where we did some demos. PCs with 1.2 GHz Athlons and a Gig of RAM were receiving real-time, streaming, high definition video (e.g., a movie) while being able to play a video game and support MS Office, Not bad. The same PCs could also support the movie video, MS Office and person-to-person video, too: you college guys could watch a movie with your buds and do your homework together, too. Unfortunately, this stuff won't be widely available for a while. Sorry. My only complaint were that the PCs max RAM was 1.5 Gigs - I'd like to have at least 10 Gigs please so I could also do some advanced Photoshop while I watch my movie - and talk to my girlfriend over high-speed telecom. Ain't that enough for most people? I mean if I can get a 3 MHz AMD or Intel and about 10 Gigs of RAM (and a good sound and video card, of course), what more can I need - for now or the future? If I'm mistaken, please let me know - I don't mind being corrected. Thanks. Gigantic1

Re:How Fast is Fast Enough?

[PingPongBoy]

That's definitely not enough. I would need at least a 1000 GHz machine to simulate sex with my girl while I work.

Re:How Fast is Fast Enough?

[Anonymous Coward]

Get a couple of these.
http://www.netcell.com/

Or it could be that....

[aminorex]

By focussing on clock speeds the
article's author has failed to notice that Intel's
new flagship line (the 7xx series) is continuing to
improve while their higher-clocked but end-of-life
Pentium 4 line is topped out.

heat and faster speeds ...

[Anonymous Coward]

too bad the basic wafer is still round and
they're still trying to squeze out as many SQUARE
chips from one wafer.
it's prolly not a chip design problem anymore but
more a wafer design problem ...
the basic chip design will have to address heat
disapation at the very basic level of the
transistors, meaning that maybe in a two core
setup there's acctually a hole in the chip
where the "whatever-heat-conducting" liquid
circulates thru the chip itself and maybe even
around the chip-to-mobo connector pins ...

but as you can see, a hole is another waste of
the wafer ... *sigh*

Re:heat and faster speeds ...

[dragonp12]

There are good reasons for having a round wafer, pity I can't quite remember any of them :-\
Possibly something with the way the light goes through the mask or something :-P

Re:heat and faster speeds ...

[joib]

Round wafers are easier to manufacture. Keep in mind that the the Si chips are made of is monocrystalline. When growing the wafer it is grown from the center outwards. When it hits the wall of the reactor vessel, it will probably break the crystal structure, and whatever growth that continues after that is not usable as there is a grain boundary. With a circular wafer you hit the edge at the same time.

Also, the wafers used today are what, 300 mm in diameter while the chips are something like 10x10 mm, so there's not much material lost anyway. And the leftovers are simply sent back to the wafer factory to be remanufactured into new wafers, so there's no material lost. Not that it would matter anyway, since Si is among the most abundant materials on earth.

Re:heat and faster speeds ...

[Anonymous Coward]

The wafers are cut from a single crystal of silicon. The crystal is made by spinning a 'seed' crystal in a vat of molten silicon and extracting the seed at just the right rate so the liquid Si solidifies in a perfect crystal arrangement. Since the growth is in all directions on the same plane, the result is a cylinder of silicon with a cone at the end. This is the only way to economically make Si wafers.

Math is hard.

[Anonymous Coward]

"Performance improvements ranged from 2.29 to 3.22 times more pages per second. In percentage terms that is 229% to 322% faster."

No, it's not.

Obviously, if something is faster by a factor of 2.0, then that's a 100% increase ("faster" aka "double"), not 200%.

Re:Is the processor clock rate trend coming to an

[open_source_dweeb]

While I don't really know about CPU design, I'd say that this is a possible reason for stopping clock improvements, since as speed gets higher, latency would keep growing.

Which is where the L2 cache plays an important role. The combination of both the speed and size matter. In fact, the newer 3.0 GHz Prescott core with 1MB L2 is typically slower than the older 3.0 GHz Northwood with only a 512K L2. This is so because Intel had to slow down the Prescott's L2 to 70% of the Northwoods in order to keep the core from overheating, to the extent that it more than offset the advantage of the larger size.

At home I run a small farm of overclocked P4 3.0s for a DC project. I have a mix of both cores, and it is obvious that the newer core is slower and has less headroom for overclocking .

Re:Is the processor clock rate trend coming to an

[PingPongBoy]

The AMD roadmap says it all: "As market requires". If the market says give me 5 GHz and I'll pay anything you can bet 5 GHz will be on the shelves. Right now you can buy sub $500 supercomputers that sit relatively idle. Word processing, db query, e-mail, web surfing, solitaire - most of the world goes no further.

The next market force is competition. If AMD looks like it will be selling a 4000+ Intel will match that.

Processors capable of this speed are most likely possible. There's no way Intel can sell and support 3.6 GHz without having perhaps seen 2 to 2.5 times as fast in stable operation under extreme cooling. In the lab where they can really reduce the feature sizes and power consumption who knows what is really around the bend?

Has anyone noticed the recent trend in laptop computers? It's all marketing and forcing consumers to buy crap. No floppy drive - so how do you boot if you want to install legacy stuff? Ultra wide screens but I've seen screens at 12" and the laptop has all the interface goodies (CD-RW, parallel port, PCMCIA, USB, speakers, 10/100, modem, headphone/microphone jacks). The widescreen monsters don't have any new interfacing, but are super heavy and don't necessarily have long battery life. The idea is to hook the suckers who will see the error of their ways and buy a new version several years later for the sake of lighter weight and longer battery.

which to purchase?

[mmikey121]

I have been waiting for several months to figure out which cpu to purchase? I am software developer but also gamer. I believe the current choice is the amd 64 fx53 however, this technology does not use the pci express features. Should i be buying intel to get into the pci express technology? It seems that intel is the only one providing the pci express video slots but based on performance charts for gamers the amd line is providing more value. Should I not worry about buying an outdated system that still relies on agp and purchase amd? I seriously need some help with this decision.

Re:which to purchase?

[Anonymous Coward]

If you have a definite card in mind for PCIx, buy that. If not, consider the fact that there are only very few PCIx cards available now. Until your next machine normal PCI and AGP will be better supported AND cheaper. I hope this helps.

Re:which to purchase?

[Zed2K]

AGP is going to be around for awhile still. If you are even remotely thinking of getting one of the new ATI or nvidia video cards go AGP. It is impossible to find a pci express version of the card. You can find older cards that are now pci express but not the new 6800 or x800 series cards.

I would place a bet that AGP will still be around and strong this time next year and maybe even the year after. There is too large of a user base to just drop it entirely.

Re:Is the processor clock rate trend coming to an

[PingPongBoy]

Compile speed is one thing I could use. I don't do a full compile that much, but I don't like waiting.

As programmers we can encourage the demand for faster processors by writing software that will do more for users if only they would spend more money. Users have to have the desire to achieve goals that they cannot reach without the help of some blistering computation. What could that be?

Futuristic scenarios like AI, virtual reality, true speech recognition, true handwriting recognition, automated car driving, etc. demand processor speeds out of this world not to mention software that works extremely hard. People will buy such software if we programers prove it's plausibility. Then we'll have our fast CPUs.

End of Life on 32bit

[tacocat]

When will they start building chips that have no support for 32-bit software?

I mean, we don't really support 8-bit today and I'm not really sure if we even have support for 16-bit software in todays CPU's. So when are we going to rid our selves of the legacy throw backs?

Re:End of Life on 32bit

[Henriok]

I mean, we don't really support 8-bit today and I'm not really sure if we even have support for 16-bit software in todays CPU's. So when are we going to rid our selves of the legacy throw backs?

CPUs will never lose its 32 bit support because 64 bits are bad for a application if it isn't necessairy. The binaries get larger, the pointers get larger and the application won't run faster.

A good operating system, and a good processor will allow for a mix of 32 bit and 64 bit applications. Use the technoogy we're it's needed.

All 64-bit PowerPCs are 32/64 bit processors running both with no performance penalty.
Same goes with Sparc and 64 bit x86 processors.
It's _not_ the same with Itanium though, as 32 bit applications will take a considerable performance hit running on Itanium.

I can't say how other 64 bit platforms behave (Alpha, MIPS, PA-RISC) but I would guess that they behave like good 32 bit processors if needed.

Re:End of Life on 32bit

[Anonymous Coward]

Ask Intel.

The day they get past the Pentium-M architecture (long after Pentium 4), and introduce the 64-bit Celeron cheapo model sporting either EPIC or a 64-bit-only x86 derivative, then we'll probably already have a 64-bit MS Windows or MS Linux as the majority OS...

AMD alone can't get us there.

Re:End of Life on 32bit

[sibtrag]

All 64-bit PowerPCs are 32/64 bit processors running both with no performance penalty.

That's because the 32-bit PowerPC used for many years was a sub-set of the initial 64-bit PowerPC design. This is quite different from the x86 world where an initial 8/16-bit processor has had 32-bit and 64-bit extensions grafted onto it.

Not that there isn't a penalty of some sort for the 32-bit support...lots of "blank upper word if MSR_sf is off" hardware.

But, it's worth it because applications which don't require 64-bit pointers often run faster when compiled in 32-bit mode (less paging, smaller stack, etc). But, we in chip design use plenty of applications that require 64-bit mode.

GPGPU.org

[Anonymous Coward]

If you already haven't, check out http://www.gpgpu.org/ for some of that :)

While it's not exactly a project to produce a holistic "offload driver", the research they do and share is probably a good starting point toward such a package...

They do a boatload of suprisingly different kinds of stuff on the GPU already!

Re:Is the processor clock rate trend coming to an

[wass]

Speed of light limit has been a known issue for a long time. At 4 GHz, a photon in vacuum will travel about 3 inches between clock cycles. Add in the actual index of refraction of the stripline leads, and it's probably more like 2 inches of travel.

I was talking to my friend about this the other day, and we think that eventually they cannot go that much faster (well, maybe have a SMALL core of the chip that can go faster), and they'll start stacking in parallel instead. Ie, massively hyperthreaded processor cores. So maybe in a few years we'll see 6 GHz chips with 8 or 16 hyperthreaded processors?

We're physicists, though, not engineers, maybe there are some other clever ways to keep pushing the envelope?

Correction

[Epistax]

www.chip-architect.com

Overclock!

[Anonymous Coward]

My lowly P4 2.4C is running overclocked at 270 MHz FSB (3.24 GHz) using Vcore of 1.65 Volts, air cooling via a Thermalright SP-94 heatsink (that I really should lapp) and Artic Silver 5 compound: 55C at load with 30C ambient, CPU fan at minimum (~2600 RPM). It benchmarks faster than the 3.6E, 3.4C, etc.

Using Sandra to benchmark, and Prime95 to torture-test for a minimum of 24 hours without errors.

The whole heat problem, etc

[Epistax]

As technology improves, devices get smaller. As devices get smaller, leakage (that is, wasted power) gets much larger. As this happens, more power is required to run the chip and more heat is generated. This creates the throttling. To achieve a higher clock speed you need the improved technology. We now have the situation where leakage is so high (I won't give a percentage but it's a big one) we've got real issues.
One possibility is to keep throwing more power at the problem and keep cooling it off. This is the easy stupid way. The hard way is either something revolutionary (that is, a totally different technology), or instead a slow incremental improvement of leakage through manufacturing techniques and perhaps some theory.

Re:End of Life on 32bit, wrong...

[Doppler00]

All x86 CPU's today STILL support 8 bit and 16 bit software, amazing enough as that sounds. This mostly has to do with the CPU registers that exist.

Example, AH is an 8 bit register and AX is the 16 bit counter part. EAX is 32 bit, I forgot what new 64 bit registers are called... but yes you will find these instructions in old software and it can usually still run.

Re:Is the processor clock rate trend coming to an

[Anonymous Coward]

Moore's law still likely applies, though, as you can probably get the chip you bought for half the price now.

Re:Is the processor clock rate trend coming to an

[bofkentucky]

One possible solution is using a material with a lower index of refraction than the current copper, been awhile since physics but silver meets this requirement IIRC.

Re:Is the processor clock rate trend coming to an

[dragonp12]

I fail to see how the index of refraction affects current technology at all...at least, not until photons are actually used to transmit information.

Re:Is the processor clock rate trend coming to an

[Anonymous Coward]

I got a 1GHz Athlon a few years back and it doesn't seem to be much behind the latest Athlons (especially when I count my athlon's overclocked speed).

I have both a 1Ghz Athlon (PC133 RAM) and an AthlonXP 2600+ with DDR333 here in the office.

There's a big performance difference between the two. (About 2.6x on the CPU and 2.5x on the RAM bandwidth.)

However, since the time that the 2800+ was top of the line, the fastest chip in AMD's line is somewhere around 3500 or so. Only 50% faster in two years.

My next system will have to be a dual-CPU system.

Re:Is the processor clock rate trend coming to an

[WuphonsReach]

Of course, there's always dual CPU motherboards...

That's the roadmap that I'm going to be following for any workstation that requires it. Downside is the cost, but from all reports it makes the operating system and everything much more responsive. Still, you know what they say, the only thing faster then X is *two* of X.

I'm hoping for dual-core as well... but unless they break the 90nm issue, are they really going to have room on the die for a 2nd core?

3000+ Hear Hear

[enterpriserx]

Personally I'm looking forward to dual-cored processors myself. What is the heat rating on the Pentium 3.6Ghz? The heat suggests that the 3.6 would last a fairly short time if we scaled it to older Intel Pentium 4s. Personally I can hit 3.56Ghz with air cooling on a 2.8C Pentium 4, pretty nice. On the other hand it runs at at 50-60C, I assembled a 3000+ on a K8T-FSR, unfortunately MSI still hasn't worked out the fine art of PCI/AGP bus locking with VIA chipsets (at least not on S754), so I'm consigned to running at 2250Mhz, but on stock air cooling, it is doing pretty well, CPU temp is 35C in average. I'd say, AMD is doing something right...

Re:Is the processor clock rate trend coming to an

[the_2nd_coming]

a 1.5 GHz system with Max DDR 2700 RAM and a SATA 10000 RPM Drive with a 16 MB cache will be as fast as the fasted intel CPUs in such a set up, unless you are doing some Processor intensive operations on a continuous basis.

Re:Is the processor clock rate trend coming to an

[the_2nd_coming]

which would be why only morons buy a CPU based on clock speed.

Re:Is the processor clock rate trend coming to an

[bofkentucky]

IIRC the electrons flowing along a copper wire encounter a resistance that is defined by an index of refraction.

Which CPU is better for gaming?

[antdude]

I am planning to upgrade from my aging Athlon XP 2200+ CPU after September 2004. I have not decided which brand and speed to get for CPU. I am into gaming like DOOM 3, Far Cry, etc.

Should I go AMD Athlon64 or Intel P4? I heard P4 Prescott CPUs have issues like heat which is bad because my room can get up to 85 degrees(F) during heat waves. Any recommendations welcomed. :)

Re:Which CPU is better for gaming?

[Anonymous Coward]

To be honest, your best choice is probably a desktop based on the Pentium-M (Dothan). By the time frame you are looking at, Dothan should be running at 2.13 Ghz with a 533 Mhz FSB. Plus, it has a low power draw (21W) and already out benchmarks the P4 and the Athlon64 in pretty much everything.

Re:Which CPU is better for gaming?

[antdude]

Isn't Pentium-M going to be very expensive? I do not buy the newest stuff out there because of 1) price 2) unknown how good the hardware is.

Re:Which CPU is better for gaming?

[Zed2K]

Get a northwood P4. The actual differences between amd and intel are mainly benchmark numbers. Real life differences are not that large.

If you get a 775 based prescott you will have a hard time finding a pci express video card to go along with that new processor. I just bought (last week) a 3.4Ghz northwood core P4 instead of a prescott and a 875 chipset motherboard to go along with it.

A lot of folks here will probably say go AMD and give a reason like "its 64bit". Don't let that sway you. Currently its useless unless you want to install a beta version of windows which would be pretty stupid too. If you can wait Intel is dropping prices on Aug 22nd.

Re:Which CPU is better for gaming?

[antdude]

Thanks. I am still going to keep my ATI Radeon 9800 Pro AIW (128 MB) video card. I am keeping my old three IDE HDDs, PS/2 mouse, etc.

Re:Which CPU is better for gaming?

[mmikey121]

You dont' see the pentium class in workstations. Pentium M's are primarily for laptops.

Re:Which CPU is better for gaming?

[Anonymous Coward]

Who said he was using Windows? You hopefully realize Linux runs many of the same games now, and 64-bit Linux distributions have been in production for over a year now. Windows still has a ways to go...

What is with the 64-bit fetish?

[DragonHawk]

"64bit CPUs to become the norm (seems to be happening)"

Why do people care about this so much? A 64-bit address space only matters if you have more then 4 GB of RAM. That's a crapload of RAM, any way you look at it. In some cases, moving from 32-bit to 64-bit can even slow things down, because now you're spending time managing twice the address space. I'd much rather have a dual-core 32-bit chip then a 64-bit chip.

Re:What is with the 64-bit fetish?

[yarbo]

This has been brought up many, many times before. 64 bit processors also have 16 GPRs, which is a HUGE improvement over 8.

Re:What is with the 64-bit fetish?

[tricorn]

Personally, I want 64 bits (and have it, in the form of Alphas and Mac G5s) to emulate 60-bit processors.

It will also speed up Java using 64-bit integers.

There are many programming tricks that can use long word sizes to do interesting things using bitwise logical operations.

Start of clue on 32-bit

[DragonHawk]

"When will they start building chips that have no support for 32-bit software?"

"64-bit" is not "better" then "32-bit" just because 64 > 32.

The size of an address word determines the memory space you can address, and most people are simply unlikely to need more then 4 GB of virtual segment size for at least the next ten years. I'm convinced that most processors will have a 64-bit address word because obviously most people think like you do and believe a bigger number automatically means a better product. That does not, however, change the reality that a 64-bit address space is a complete waste for most people.

Re:Start of clue on 32-bit

[HuguesT]

Right now you are correct, but apparently Longhorn will require 2GB just to run.

Most computer sold today have 512MB in them and people do use that amount of memory and start swapping soon after startup. Once you have the memory you realize you can have all your apps loaded up and switch between them, it really does change your way of working.

No matter how much memory you have Linux will use it up and it does make things faster. If you have enough RAM Linux caches everything you read from the HD and then read from the cache. Try a `du *' twice from your home directory to see what I mean. Notice how the second time is so much faster?

People are getting into video editing. It's not mainstream but it won't take 10 years to catch up.

I've been needing more than 2GB for a long time, I work in image analysis, my code has been running on 64-bit machines (alphas and sparcs) for about 10 years now. One machine at work is a 4-way opteron with 32GB in it and we are using the lot.

I can't wait to have more than 4GB on my desktop. Right now the price of the memory is the problem.

Re:Is the processor clock rate trend coming to an

[MojoStan]

The Celeron is a severly crippled chip, unlike the Duron, which is a respectably performing budget processor. It only has 128KB cache, which is CPU sucide on a P4 core...One would be much better of with an Athlon XP, Duron, or a slow P4 as a budget processor.

He said he had a Celeron laptop. For a budget laptop processor, I think the best choice for most people is the Celeron M. It shares the same architecture as the Pentium M (Banias and Dothan), but has "only" 512KB L2 cache and no "Enhanced Speedstep."

Tom's Hardware reviewed a Celeron M notebook and, unlike the old P4-based desktop Celeron (128KB L2 cache), it is not at all crippled. Here's a link to the review: Does Everything Have To Be A Centrino? Intel says "No"!

Unfortunately, Celeron M notebooks aren't as cheap as notebooks with cheap desktop chips. The HP Compaq nx9020 (1.3GHz Celeron M) "starts at" $800, but that's with Intel integrated graphics, CD-ROM, and 128MB of shared memory. But for those that want a budget laptop with high performance, low power, and thin-and-light dimensions, a Celeron M is probably well worth the extra money.

cpu draws mainly outside air?

[pwarf]

You are assuming a homogenous case temp.

As many people place an intake fan right outside the CPU to feed the CPU outside air (with a slot fan or normal case fan), so it is quite plausible that the CPU temp is under the average or representative case temperature. Especially since the exhaust from the CPU fan is usually directed into the case and further heated by hard drives, GPUs, etc.

Granted, the output air temp cannot be lower than the intake, but there is no reason to assume that the intake temp is the case temp. I wouldn't be surprised if there are some overclockers feeding an air conditioner output directly into their computers. Extreme, but less risky than homebrewed water cooling.

Trusted Computing may take some of the blame?

[Alsee]

How much of Intel's development force was moved over to Trusted Computing development? That could be strangling some of the CPU development roadmap.

-

Re:Is the processor clock rate trend coming to an

[Proc6]

Huh? Widescreen is a "good thing" (tm) have you ever used one? They're quite nice. My wife's new notebook was very inexpensive, like 1 grand at most. 1280x800 widescreen, DVD, PentiumM, 256MB, 20gig, built in wireless, iRDA, etc, etc. Everything you'd ever need except a floppy. Im not about to cry that floppies are going away, and if you really need one you buy spend all of what, $20 on a USB one? I would say cutting costs by dropping items 1 in 100 people will need is generally a good idea. Btw, her notebook is about 6 lbs, the average weight of any low cost notebook, very portable, and it runs 5.2 hours on the internal battery. Pretty solid. Maybe youre just looking at junk notebooks.

Re:Is the processor clock rate trend coming to an

[aSiTiC]

The permittivity of the oxide surrounding the conducting striplines must but be taken into account as well. For many oxides used on motherboards and in silicon this values ranges from 4-7. Take the square root of this value and divide your photon speed to get the max value on an oxide with a perfect conductor. This is rather limiting too.

# of registers is independent of word size

[DragonHawk]

"64 bit processors also have 16 GPRs, which is a HUGE improvement over 8."

The number of registers has nothing to do with the word size. The 32-bit SPARC has 24 general-purpose registers, for example. The 64-bit Alpha has 64 GPRs (32 integer and 32 FP). There's no reason we have to have a 64-bit word size just to get more registers. Nor is there any reason to limit the number of registers to 16, for that matter. That smells like 8086-induced brain damage. If we're going to go for more registers, why not make it 32 or even more? Registers are something that most applications will actually use.

I can understand wanting more registers, but why needlessly couple that to the word size?

Re:# of registers is independent of word size

[yarbo]

In the x86 world, that's the only way to get more registers. x86 isn't best, but it's the most common desktop computer processor, and it gets a pretty nice performance boost from going to 64 bit in many applications.

btw, word size is tied to number of registers. Each executable instruction that makes use of registers has to have the register number as part of the instruction (at least from the designs I've studied (LC2 and MIPS)
(posted from a 32bit PPC)

Re:Is the processor clock rate trend coming to an

[wass]

Sending an AC electronic signal (ie, any signal with any non-zero frequency bandwidth) IS photonic in nature. You're not just sending electrons down a pipe, you can look up for yourself the electron drift rate in even the best conductors to see how incredibly slow that would be.

Photons are the mediating particles of electromagnetic force, and it's definitely this force that couples two electrons together, or the electrons to the 'holes' in the doped semiconductors, etc etc. An elementary description of current in a wire is akin to a tube filled with marbles, you push one in, and one comes out at the far end. This interaction between the 'marbles' would be mediated by photons. Of course metals and semiconductors are far more complicated than this picture, but it's a rough start.

It might sound weird to you (it did to me at first), but when you send a 100 MHz signal down a coax cable, you are really sending photons. They're rather low-frequency photons confined to a waveguide, but they're definitely photons.

Re:Is the processor clock rate trend coming to an

[Anonymous Coward]

Signal propagation delay is the main reason for pipelining. The rising edge of each clock cycle dictates when the storage elements(called registers) sample their inputs and hold that value. The actual adders, shifters and other computation functions(logic functions) are done between these storage elements and have a certain delay. The delay is a result of MANY factors including physical distance of the wiring, the capacitance, transistor sizes and thier operating voltages.
A way to increase clock speed is to add registers at more points inside the logic functions. That way you can run a faster clock because the signal does not have to propagate as far before being detected by the registers.
The main problem with high clock speeds is heat, not wire delays. Adding more registers generates more heat simply because they have transistors which switch every clock cycle. There other semiconductors that can run MUCH faster(SiGe) but aren't used because the heat is too high.

Re:Is the processor clock rate trend coming to an

[vadim_t]

Well, what I mean is, wouldn't you eventually bump into a limit regardless of how high is the clock speed?

Say, if you made a 200 GHz CPU, it'd probably spend a lot of its time waiting for stuff to arrive from memory. Ok, you have a cache on the CPU, but that won't help for long since at 200 GHz this cache will be exhausted a lot faster. You could put a bigger cache, but then you enlarge the size of the core, and it will probably not take very long to have exactly the same problem internally.

Of course, it can still be blazingly fast as long as stuff stays in the cache, but as soon as you started working with a big dataset it'd suffer a huge slowdown.

RTFL!!! The 3.4 is EE, the 3.6 is not

[megalomang]

... is just a sign of a too inefficient design for the speed

READ THE FINE LINKS!!! If you would actually look at the links (as would the submitter and reviewer), you would realize that the 3.4 GHz part is an Extreme Edition, whereas the 3.6 GHz part is not an Extreme Edition. So much for your lecture about the inefficient design...

Another Intel roadmap

[vincecate]

There is another interesting unofficial Intel roadmap. Since Tejas and Jayhawk have been canceled there is just not much new Intel x86 in the server, workstation, and desktop in the next 11 months. The other interesting thing is that Intel's first duel CPUs will first be in Itanium and notebooks. Since heat is something of a problem in current single core server, workstation, and desktop chips, it makes sense that these will not be the first duel-core chips. AMD plans to have high end duel core chips next year.

Also, people are right that in the SPEC results the 3.4 Ghz chip has a 2 MB level-3 cache and the 3.6 Ghz does not.

Why square?

[tehdaemon]

Others have pointed out that we are stuck with round wafers. The real question then is why are they still using square chips? I would think that hexagonal chips would pack more onto the round wafer than squares. Or perhaps triangles? (you want some shape that fits together without gaps, otherwise you waste more than what you would save)

All of the chip desighs I have seen have large square L2 caches, Is that the reason?

As for cooling, the transistors are in a very thin (measured in nanometers!) layer. Why not grind/polish off the rest of the SI (after it is made) and putting the cooling fluid right next to the transistors? (a few 10's of micrometers?)

Re:Is the processor clock rate trend coming to an

[NeMon'ess]

If you'd RTFA you'd know that Intel doesn't have anything faster than 3.6GHz planned until June 2005. Meanwhile AMD is likely to release their 4000+ rated chip this October. First quarter next year 4200+ running at 2.8 GHz is coming. If Intel can't counter, they will lose the high end of the market to AMD for as much as six months. This is worse for them than back when AMD reached 1.0GHz first and stayed 100MHz ahead of them for six months. Intel is really going to lag behind. In some benchmarks, the AMD 3700+ and 3800+ already beat the 3.6GHz.

Re:Is the processor clock rate trend coming to an

[JollyFinn]

Right now that holds for multiple reasons. A) both AMD and intel is going for dual core in a year or so. Desktop maybe later. B) They are moving to 0.65u manufacturing by then. Thats 4 times transistor density for AMD and twice for intel. Which result more stuff on processor die. So its 2 processors and lots of cache and other stuff or 4 processors... C) Probably a new architecture is introduced by then K9 for AMD. For intel It might be desktop Itanium. [Yes thats fast, unlike the common misconseption. Intel server line is currently one process generation behind the x86 line, probably because longer verification cycles for server CPU:s, and server systems, and extra reliability requirements. And die area is huge and so is power because its designed as high end server CPU not because of architecture. x86 is quite fast as software emulation on itanium, actually so much faster that itanium could get nice speed up from ditching hardware x86 compability that consumes power and slows down the clock speed slightly.] But intel should get something in 2 years either itanium, or Pentium M based multicore desktops. Or something.

Sorry to all non-scandinavians

[empaler]

Millenia er flertal. Det ville være ligesom at sige "Computere fra dette årtusinder". Du ville sige millenium.

Røvsnaps!

(This was just a language-nazi post, in case onlookers were wondering)

Re:Sorry to all non-scandinavians

[EuropeUnited]

Tack ni svenska vakttorn,
med plutonium tvingar vi dansken på knä!
Här - Danmark utskitet av kalk och vatten,
Där - Sverige, hugget i granit.

Danskdjävlar.
DANSKDJÄVLAR!!!

Hvad er nu det?

[empaler]

Det lyder sørme som et mindreværdskompleks! ..
Altså, bare fordi vi sad på magten i Skandinavien i et halvt årtusinde betyder det altså ikke at I er mindre værd. Øhm. Nej.
Det er der MASSER af andre årsager til!

Med det værende sagt så ser jeres lunarspam mere indbydende ud end vores...