Slashdot Mirror
AMD Athlon64 4000+ Underclocking
Bios_Hakr writes "PC Stats is running an article on their experiences underclocking an AMD 4000+ processor. Their goal was to try and reduce the voltage requirements and lower the heat output. They benchmark using 3dMark01, 3dMark05, as well as SuperPi. From the article: 'This got us thinking though; what about under-clocking? Most modern processors and motherboards can just as easily run under a rated speed as it can run over... but is there a point to this? Well possibly.'"
The point is.....?
If you want low power you can buy systems specifically designed to perform well on low power supply.
Deleted
Their next article: how to remove 2 cylinders from your Ferrari's V12 engine.
Microsoft operating systems and software accomplish this without all the work.
Intolerance for ambiguity is the mark of the authoritarian personality.
Unfortunately... performance and heat seem to move proportunately. I would love to see (or hear) a silent server room. Hmm... maybe with embedded systems getting more powerful this will one day be a reality.
If you are designing a system for high reliability, under temperature extremes and such (military environments for example) underclocking is the way to go - you can minimize power and heat loads as well as potentially avoid timing instabilites that occur when you push a processor to the performance margins.
I thought what they are testing is the whole point of AMD Cool'n'Quiet technology, but they don't even mention it in the article! Nice try reinventing the bicycle. I'm already underclocking my Athlon 64 right now, thank you.
Couldn't you just buy a slower processor? Why buy a more expensive processor just to have it match a slower (read: less expensive) processor's performance?
I could see where the power drain and heat would be reduced for use in a laptop or other mobile device that runs on batteries. Currently chip makers have two separate cpu's for desktop and laptop but might save in production costs to manufacture just one for use in both systems.
Look it up on Google, you could have a near silent server room today. AMD and Intel suck, literally (power & AC).
Deleted
n/t
PC Stats actually took this idea from me. I took the same processeor, unplugged the computer (to lower power consumption) and then stuck it in a freezer (to lower temperature). Wow! Also nil power consumption (the freezer uses some to keep it cool) and the temp is really low! Next I'm going to try this with my laptop because the bottom of it gets really hot!
Underclocking is a real good thing, if you want to save your electricity bills, plus, you dont want to run your computer faster when there is no need to. Consider a case where you use your amd64 for both browsing and occasional scientific computations. You want to run it at its full speed (and consume how much electricity it wants to) only when you use it for computations. Otherwise, clicking on the back and forward buttons of firefox doesnt need a 4000+ system. It is very important to understand that the underclocking is as important as overclocking, given the fact that most people in the corporates want the fastest computer on earth for sending emails and solitaire
Is it really a surprise that the proc will run stably underclocked?
They've been doing this at the factory for some time now.
You would think the whole point gets moot - the system certainly won't be quiet. (I believe there is an actual need for quiet systems eg in recording studios etc - which make the article interesting, but not great).
I have a really elegant proof for Fermat's last theorem. If this sig was only a bit longer...
My gut instinct tells me.... Maybe!
If you want low power you can buy systems specifically designed to perform well on low power supply.
The article is about researching how to build such systems out of cheap commodity parts, unlike the proprietary, often Windows-only parts found in laptop computers.
I think it is somewhat useful information. While most people are thinking about how much faster you can process, many of us are looking to reduce the noise of fans blowing. I recall when 800MHz was a fantastic speed... hell, for that matter, 300MHz was pretty nice too depending on how far back you go.
And are we really using all of those cycles? Not really. Right now, a system's performance (IMHO) is largely the responsibility of the quality of RAM, Video and system board stuff. After all, what "feels" fast must be fast. If I've got a slow hard drive, then it's a slow system and if I can accellerate the video, then it's a slow system. What good is 4GHz if you've got a slow everything else... and by the same token, if you've got a fast everything else, a 2GHz processor is probably plenty.
I decided to underclock some 1U systems (~XP 2500) to ensure that they would never overheat (longevity was more important than performance).
It works perfectly: a drop of 20% in core clockspeed greatly reduced the heat output, the core temperature dropped by almost 10 degrees C.
TODO: 753) write sig.
They listed the drop in temperature from 33.5 to 26.9 as a 20% drop. However, they didn't mention the ambient temperature. If you take 20 degrees, then this drop is more like 50%. That would also mean that it was consuming well under half the power. (I'm assuming watts->degrees is exponential.)
As a secondary matter, the person who got me interested in BSD, as a rule, made his servers with whatever was the cheapest AMD-K6, underclocked to 350MHZ. Bulletproof boxes with long lifetimes. I'm sure there are still some churning out the bits around this town.
On all of my personal must-stay-up servers, I get a processor that is too beefy for the task it's to do, then clock it down. It's usually rock solid and runs very cool. In some cases I've been able to get by using only passive cooling and still keeping the processor very cool, making the system solid, cool, and nearly silent.
If you want a vision of the future, imagine a youtube comments section scrolling - forever.
...running a server on 486DX80 downclocked to 66MHZ. Original fan+radiator replaced with a radiator alone. From an Athlon.
(several more power savings in the system - like not using a CD drive, and the power supply runs just fine with its fan switched off. So, a fanless config.
45 5F E1 04 22 CA 29 C4 93 3F 95 05 2B 79 2A B2
If they want to underclock a 4000+, they could just swap me my 3000+.. I wouldn't complain.
--
That's what CPU power management is for: it allows you to select in software what tradeoff between performance and power you want. I believe most (all?) of my desktop machines have it built in.
Claiming percentage change in temperature in Celsius is meaningless...it's not an absolute temperature scale like kelvin.
In this context, talking about a 20% drop in temperature in degrees celsius makes no sense for comparison purposes. They go on to state that "a 43% drop in voltage producing a 20% drop in heat seems more reasonable", but this is assuming that the temperature drop corresponds to a equal reduction in heat output.
- Brian.
-
- - You can't take something off the Internet! That's like trying to take pee out of a swimming pool.
They have measured the (absolute) celcius temperature of a well-cooled system, without quoting the ambient temperature. Then, concluding that the temperature hasn't droppped much, they assume the power hasn't dropped much.
The correct measurement is the *difference* in temperature between the CPU and the ambient air. Power dissipation is linearly proportional to this.
Cos by buying a 1U system with a slower CPU in the first place you could have saved money as well as reducing the temperature and power requirements.
Deleted
From 99-03 I always underclocked my AMD CPUs for gaming, the minimal requirements necessary for playing everyones favorite Counter-Strike, and my undying urge to make my CPU last ages kept my system running smooth, my latest underclocking was from a 2800+ 2ghz AMD down to a 1.87ghz miracle machine that ran on a minimal fan system, never overheated and my record uptime was 3 months and 2 weeks ;D
0.09 nanometers?? Wow!
Yeah, I know what you meant. Sorry I couldn't resist...
What laptops need is a variable voltage and speed controller. A bit of calibration software would map the lower-boundary of core voltage versus speed (maybe as a function of temperature, too) and then use that calibration data in daily operations. The machine would constantly regulate both core voltage and clock speed.
Two wrongs don't make a right, but three lefts do.
From the article:
"To put it another way, for a 66% drop in speed there was a 20% drop in temperature. This makes a bit more sense if you look at the numbers in terms of Voltage not speed; a 43% drop in voltage producing a 20% drop in heat seems more reasonable."
Utter non-sense.
Why do people not understand that temperature is a relative measurement? and that 0 deg C does not really mean much? especially measuring percentage?
dropping from 33.5 C to 26.9 C is not really a 20% drop in temperature, if you convert it to Kelvin then it's only some 2-3% drop? The correct way to quantify this is to measure the drop against the ambient temperature, which I assume is somewhere around 20-25, then the drop of would be a lot more significant.
If going by the author's logic, does he mean that if we could somehow expect to achieve 100% drop in temperature by not applying any currents and get a freezer block? and thus violating laws of thermodynamics and creating a warp in space-time continuum and leading to the end of universe?
Darn, I knew I shouldn't have turned that computer off...
The article is about researching how to build such systems out of cheap commodity parts, unlike the proprietary, often Windows-only parts found in laptop computers.
(The UK) Personal Computer World have an article in the current edition about using the 'Pentium M' processor in desktop machines. Mobos *are* available that support these (Abit IIRC); as well as adaptors that allow certain ordinary P4 motherboards to accept a Pentium M.
For my money, these look like a good compromise between the micro/nano-ATX mobos that take up little space and require no fans (no heatsink?), but have slow and un-upgradeable processors, and your average P4 system that sounds.... like a jet taking off.
As power requirements are low, I assume a low-powered silent power supply would do fine. I'd certainly consider this option for my next PC; I'm sick of the amount of noise my computer makes.
"Slashdot - News and Chat Sites Deviant". (Click "homepage" link above for details).
Use CrystalCPUID to manage your AMD64 CPU's speed and voltage rather than the default Cool 'n Quiet power management (set your Power Scheme to "Always On" to disable that, definitely leave the CnQ driver installed). On most HP AMD64 notebooks we've found that you can usually safely set the core voltage at about 0.2V below stock at full speed. Judging by the AMD Thermal Design Guide, that's enough to cut power consumption nearly in half. I swapped in a Mobile-class Athlon 64 3200+ into my Pavillion zv5000z in place of the stock DTR-class chip and have been running 1GHz at 0.8V, 1.6GHz at 1.025V, and 2GHz at 1.225V for months. That puts the full speed power consumption at slightly above AMD Turion ML levels. For the stock DTR chips, 1.3V at full speed is popular.
Of course, in average use, the standard AMD Cool 'n Quiet behavior of running 800MHz at 0.95V while idle will give you battery life that's almost as good as an undervolted setup. 3-4 hours of battery life with a 12 cell battery is common, versus a fraction of that for the poor bastards who bought the P4-based zv5000 series (HP wisely dropped Intel CPUs from their zv6000 line). Undervolting does wonders under heavy CPU load though.
MobileMeter is my favorite way to monitor CPU speed and temperature, and Hot CPU Tester Pro verifies that I didn't go too far.
..and, indeed, thought that the importance of ambient temperature was the whole point of his post....
Why you underclock your processor if the amazing thing is to overclock and see if you can cold them without loosing it.
http://www.michel.eti.br
Doesn't using CPUFreq with a powersave governor have the same effect as underclocking the CPU in the BIOS? What's the advantage of doing it the way suggested in the article?
I'll probably be modded down for this...
My system is never idle. It runs seti@home and/or folding@home 24/7 in the background. So I don't think the power saving features will work for me if they depend on the processor being idle. I bought a Dell 500SC for home. It has been rock solid, but the fan is very noisy, and the DMA on the secondary IDE is busted (chipset bug). When I upgrade, I don't care about bleeding edge performance, I want it to be quiet. Wouldn't you know, after I bought the 500SC, Dell came out with the 400SC, which I've installed at several customers. That thing is quiet as a mouse. Sigh. I thought about switching and telling them, "See, 500 is better 400!"
The benefit for heat reduction (and less thermal sensitivity in an enclosed space) is often a worthwhile tradeoff over a server I can just leave alone and expect to run for years at a time.
So... it would be a Transmeta processor?
Isn't this what CPU Frequency Scaling is for? I personally use the ondemand scaling governor made available in linux kernel 2.6.10, I believe.
The parent post is right. Without ambient temperautre, the results are meaningless.
Looking at the voltage levels, it is obvious that slowing the clock twice, allowed for about 4 times less power! The article didn't even mention that. The experiments were good, but the guys that did them failed completely in their conclusions.
what I don't see very often is reviews address all the other sounds in a case, like the damn hard drives. I never hear my fans, system is water-cooled with two 120mm fans at 5v, but all nite all i hear is GRINNNNNNNNNNNNNNNNNNNNNND of the damn hard drives. Why doesn't someone address this issue and do a REAL review on how to get rid of hard drive sounds? Sure silentpc has done a few, but everyone else is like "yeah, i hear like, a fan, sometimes, so i'm gonna run my new 4000+ processor at 800mhz".
talk about unoriginal....
my karma will be here long after I'm gone
Thats dumb, just get water cooling if you want a silent system... Or get the Zalman case that is totally fanless, yes w/c and case cots more than underclocking, but thats stupid, people pay more for for more power, why would you underclock it?
Computers built for industrial temperature ranges are routinely underclocked. We have three underclocked Pentium 4 systems in the Overbot
quite a few years ago I was given a Cyrix PR133+ motherboard/processor from a friend that bought it in the states, it was terribly unreliable, crashed all the time, he didnt want the hassle of returning it because of all the shipping etc etc so he gave it to me, I underclocked it to a CyrixPR100+ and man that thing was stable, I used it as home server for years, it had 120+ days of uptime no problem (thats amazing for a cyrix)
--
The article states:
Unless the computer and participants were in a frozen room (at 0 degrees celcius), their analogy is flawed. The amount of heat generated is directly preportional to the temperature INCREASE above the ambient temperature. Let's assume that the test occurred at "room temperature" (70F deg or 21C deg). The chart would look more like the one below:The article should have stated:
"For a 66% drop in speed, there was a 53% drop in added temperature."
"a 43% drop in voltage produced a 53% drop in in heat seems more reasonable."
My observation from that data above:
"A drop of only 400MHz (17%) and 0.15V (11%) showed a significant drop in the amount of heat generated (25%)."
The systems on themarket designed for low power (trmansmeta, et al) are desiged for minimal power, at the great expense of cpu horsepower. The article is more about building a system with the greatest possible horsepower without needing noisy or failure prone cooling methods. Perhaps you want a fast computer that's going to go into your bedroom or home theater, where noise is an issue (my TIVO keeps me awake at night sometimes, so I can sympathize).
/.ers have entioned that you should just buy a lower speed CPU. This doesn't work, because often a cheaper, slower CPU (from an older core) uses just as much power as the faster, newer ones. An athlon 4000+ dropped to 3000+ may use a lot less power (and geerate less heat) than a stock athlon 3000+. You can stick a fanless CPU cooler on the processor and a single 2000 rpm exhaust fan on the back. Viola, quiet computer, and less fans ready to die and bring down a mission critical system.
Many
I used to have an underclocked PC acting as a router and currently have another that is used as an internet guest machine or for bittorrent downloading and hosting. They're old hardware and it's worth cutting down the speed by about 1/3 so I can remove the fans. I have to turn on the monitor to check if the machine is actually running. I save electricity costs, and I'm not adding to an already noisy room.
The 3D benchmarks seem a bit retarded though. If you're gaming, you should have the sound turned up enough to cover fan noise.
A couple of years ago, I took part in a project to build desktop PCs for low income families with high-school age kids.
We made about 200 total boxes. When we got towards the end, we had about 3 dozen donated motherboards that could only be set to a max of a 450 Mhz processor and PC 100 DIMMs for RAM, but had 2x AGP slots, and a bunch of P2-500 and 600 CPUs, PC 133 RAM, and single speed AGP video cards to match up. We ended up turning out a bunch of machines that were all basically underclocked in several ways at once.
These have tended to last surprisingly well, compared to the rest. This was charity, and there was no way to regularly survey all the recipients afterwards, find out who had gotten other tech support, installed other hardware and so on to make it a scientifically accurate conclusion, but just as an impression, there have been no power supply failures or burned up CPUs or RAM in that group, and at least a few in each of those categories for the others.
Who is John Cabal?
My xp2000 ran at a deafault 1.75v. I decided to start droppping the voltage to cool it down. I dropped by .05v, tested for a day, then dropped another .05v until i reached 1.4v, at which point my system crashed once or twice. So I put it up to 1.45v, and it's been like that for over 2 years now.
The temperature of my cpu dropped almost 10 degrees celcius, and i've had uptimes as high as 3 months (on a winXP system, mind you). So while I don't think under clocking is a good idea, lowering the voltage is awesome if you wanna lose some heat.
I am extremely glad to see "underclocking" gaining in popularity.
I design systems that run applications across hundreds or thousands of servers. Many of my applications are bound by items such as connections, long before processor becomes a bottleneck.
As a case example, I will have an application that utilizes 55% of the proc across two processors. I use two processors to keep response time down (multi threads). Intel gives me a new processor. I get to spend more money to power the new processor, but now I get the amazing advantage of the new, faster, more power hungry machine now being 30% utilized.
More money down the drain, but I am not getting much for it. The worst abuse of this is static content web servers. I run into connection issues and network latency issues long before I run out of processor.
With the new HE processors from AMD, I can turn down the processor clock and cut my power consumption by as much as 50% across the board. This translates into real savings on power and cooling infrastructure.
I've no idea what they are up to, these days. Probably like 1.5ghz, which is about the equivalent of a 1ghz Pentium III. You can cool it passively with ram sinks.
For any CPU's manufactured on 90nm or some fine geometry, there's not much point in under-clocking.
The leakage current from the transistors makes up the bulk of the power draw. It can be 90% or even higher.
(The static current drawn is when you just turn on the power and don't apply any clocks or toggle anything).
So at best, even if you clocked it to 0MHz, you'd save maybe 10% power if you're lucky. That's it. It's not worth it.
You can get better power savings if you completely shut off portions of the chip, so this might be a strategy for low-power CPU's, and that's why the standby might differ a lot from the peak power drawn. But if you're actually running the chip 100%, forget it, underclocking is a dead end.
A fair comparison of processors would be to calculate how much processing power you get in some benchmark per Watt of power dissipated: Fan noise for getting rid of CPU heat should be lowest when the least heat needs to be gotten rid of. Probably the processor with the best performance per Watt at full speed also has the lowest disspation per Watt at lower speeds.
Thank you for letting me share this old-timer drivelling on slashdot.
Panurge has posted for the last time. Thanks for the positive moderations.
I have all these old IBM PC games that only work at 4.77 Mhz...
Write Only Memory: Another pointless blog.
They spent a lot of time doing the tests and still missed spelling out one essential parameter: ambient temperature...
They complained that by dropping the CPU clock rate from 2.4GHz down to 800MHz they lowered the CPU temperature by a flimsy 6.5 *C. Of course, if the ambient temperature is 24 *C, then they reduced the temp increase (which is what matters) from 9 *C down to 2.5 *C, a 3.6X improvement!
Of course there is merit in underclocking; they just picked a relatively cool CPU to start with.
The benefits of over or underclocking depend upon the application. If you're a guy that wants to squeeze that last .1 FPS out of his Doom 3 box, sure push it to the limit. But if you are more concerned about stability, you might want to underclock.
... well, let's just say that reliability and stability are an issue. I underclocked processors for years, even back in the 386 days, just give myself a little extra temperature margin. I did have some situations where even that wasn't enough and I had to go to active cooling, but that's another story.
For example, when you stick a PC, even a rack-mount industrial PC, into a plant floor environment with an ambient temperature over a hundred degrees
Thing is, for a typical industrial control or data acquisition application, you usually don't need performance. Stability is more important, and underclocking can help improvate that.
The higher the technology, the sharper that two-edged sword.
"So you're good wtih computers, do you know how to make them run faster?"
"No, but I am quite good at slowing them down."
I built a WinXP machine that runs SnapStream's BeyondTV PVR application. I use a Zalman HSF (1600 RPM), 1 case fan, and the power supply fan as the only air management in the system. The video card is a GeForce 4 MX with a passive cooler. It also has a pair of 7200 RPM IDE HDs in it.
I was running into some lockups and stuff with a XP2400+ CPU in the system.
I took it out, and purchased a Athlon XP-M. I use Crystal CPU ID to dynamically change the CPU multiplier on the fly. It works great.
Most of the time the CPU is idle when it's not recording or doign playbacks. So it runs at 800 MHz. Conserves power and helps keep the temperature in the case down which is most important to me.
A bunch of years ago I experimented with underclocking. I had an old AST 'Bravo' 286 machine and there wasn't much interesting to do with it. (I was running Slackware on a constellation of 386sx and better machines to fool around with networking). It had a socketed 'crystal block' TTL oscillator. I had a bunch of other oscillators around so I started plugging them in.
The base machine was as slow as an AT gets, it was a 6 MHz. 286. I plugged in a 1 MHz oscillator to make it a 512 KHz '286 machine. It actually booted up, very veeeery slowly. You could count the actual steps as the BIOS did the traditional 'step the floppy drive to one end and back' sequence.
Very nice!
Then I tried some even lower-value oscillators. I have block oscillators down to a value of 32.764 KHz. The machine wouldn't boot up at all at lower frequencies.
This is because the memory on the motherboard, and indeed the registers inside the CPU themselves, are dynamically refreshed. If the chip isn't run fast enough, it crashes.
There are processors that can run down to zero hertz, with an all-static CMOS design. The Intersil/Harris 6100 processor has this characteristic. You can use a knife switch as your clock if you wish.
This is such a smart post. Thanks for this!
Use the Silentmaxx HDD-silencer Rev. 2.0
No noise and no need for a fan cooling the HD to prevent heat-death.
Combine this with RAID-1 and you've got ultra reliability.
- -- Truth addict for life.
Why does every Pc benchmark have to use bar graphs? In this case some nice trendlines could have been made. I admit that with a general pc comparison they make sense, but when you are comparing only one computer you can establish a real relationship between the variables you change and those that you measure.
2*31*37*263
Seriously, I started reading their web site and this horrible "brrruzzztt!" sound came from my speakers. At first I thought it was some idiot on a CB radio broadcasting noise that leaked over into my system or something. Then it happened again, and as I scrolled down the page, I saw that lame-ass "swat the fly!" advertising on the left-hand side of the screen! I even swatted it once, hoping it'd just show me the stupid "you win a prize" page and shut up -- but nope. When I went back to the original page, the fly started buzzing again. I bailed out on it.
My MSI K8N Platinum NForce 4 runs "Cool & Quiet" as a feature. Thus my AMD64 is running at 1000MHz as I type this but will dynamically crank up the mulitiplier should increased grunt be required. Commensurate with this the fans pretty much shut the fuck up and I guess if I fixed those settings passive cooling would be fine.
Given this technology has been around for sometime now I'm not really sure what the point of the article is.
Plays violent online games as: Nerfherder76
Use a USB external drive (if your system can boot from USB) and keep it in a closed drawer.
after underclocking to 800Mhz, it was still good enough for word processing and surfing. I'm getting me one of those!
On a long enough timeline, the survival rate for everyone drops to zero.
Once, I installed a Linux box acting as an Appletalk/NFS/FTP/Samba server for a company doing video/audio productions.
The processor was a Cyrix P166, more than enough to saturate the coaxial ethernet.
So I decided to underclock the processor a lot (I think something around the half of its original speed).
After two years, the company called me. During a storm, lightning hit their electrical grid and fried one of the HD in the server. When I opened it, I discovered that not only was the HD dead, but the CPU fan had failed for about 1 year (this is an estimation done using the amount of dust on it). Still the heat dissipator was keeping the CPU cool enough!
Thanks to underclocking!
Well, here in Vostok I need the overclocking just to keep my fingers from freezing while typiiiiiiiiiiiiii
You can get both at the same time. When working with old carbuerated V8 racing engines (specifically, SB and BB Chevys, because this is where my experience lies in) you can get more horsepower when you lean the air/fuel mixture up (smaller jets in the carb). You see what happens is that the leaner the A/F, the hotter the temps get. The hotter temps create more pressure on the tops of your piston (because you get power out of pressure in an IC engine) and viola! more power with less fuel comsuption. the only problem is making sure you don't lean too much because you'll end up melting your pistons.
Cheers.
It's Athlon64 3500+ Venice socket939 2.20 GHz, Model 2F, stepping E3, Max. 67.0W TDP.
To see ADA3500DAA4BP from http://www.amd.com/us-en/assets/content_type/Downl oadableAssets/K8_Electrical_Spec_Rev_ENG.pdf
I'm programming "general purpose x86-64 + SSE3 only" on Linux-2.6.12 & FreeBSD-5.4.
A64 3500+ Venice - 4 GiB DDR400 - Asus A8V - 250 GB HD RAID1.
-=-=- DaNgErOuS hAcKeR -=-=-
All they did was run a processor slower and measure how slow it was. Then they measured the temp. However, by their own admission, neither of those were relevant.
What they should have done was get a few CPUs and underclock each to the point that they can run fanless reliably. Then look which made the cheapest, acceptably fast, silent system. Putting that video card with a fan in it was a really dumb move too.
The MSP430, for example, a relatively recent microcontroller designed to have low power draw (microwatts), can be used with two crystals: a fast one and a slow one. The slow one is usually a 32kHz watch crystal. Lots of microcontroller systems will completely halt their clock when they're not doing anything (say, between keystrokes). It takes a while to start up again, but in many applications you can afford to wait tens of milliseconds for the system to respond to an event.
Maximum efficiency is a 15:1 mixture, maximum power is more like 12:1.
So you'd only see a simultaneous increase in both power and efficiency if you happened to start with a richer-than-12:1 mixture.
Introducing the PowerMac G1, starting at $2999. The world's most advanced processor, underclocked to amazing speeds. Advanced architecture: with extra inefficiency, our processors can do an entire FLOP! That's right, a floating point operation! Select from three blazingly slow models: 9 Mhz, 2Khz, or the world's most advanced desktop featuring dual 0.01Hz underclocked 2.7Ghz processors.
I have been underclocking an AMD 550 K6-2 for abt 3 years. Low heat output was a benefit, motherboard stability was the key reason for doing so. Pre 100 MHz bus boards were not very stable at higher bus speeds. Even though 83 Mhz was supported by this particular "offbrand" board, 66 MHz was the most stable, 75MHz was used for some time, but resulted in many "coffee coasters" of bad burnt CD's, due to a 4x4x20 Ricoh burner, lacking buffer under-run tech. (since replaced)at the time. This AMD 550 still operates at 350 MHz, for max reliability. And is housed in a modified 8088 processor era desktop case, with abt 9 fans! So cooling was not my prime concern.
From an end user standpoint, the other important point apart from energy is that the drop of 20C in temperature means that you don't need to run your fan at full speed.
With a nice fan and a rheostat or a motherbard that supports the quiet part of "cool and quiet", it means almost no noise.
I thought this was a pretty dumb article, although I did like the comparison of the heat-sink to Soviet archtecture. It was dumb because if you could produce a high-speed low-power chip this way, that's exactly what AMD and Intel would be selling for this part of the market, instead of the specialized chips they do sell.
"It's the height of ridiculousness to say for those 9 lines you get hundreds of millions."
I have my AMD Shuttle mildly underclocked to keep the fan at its lowest speed. However, this allowed me to tweak the memory timings up, so I didn't sacrifice any memory bandwidth in the process. Without controlling for memory bandwidth, this underclocking exercise is completely worthless. And he also didn't control for ambient. The expected temp. drop is a fraction of the temp. rise above ambient.
When you underclock a processor while holding memory bandwidth constant, there is a net efficiency gain, since more real work is accomplished with fewer processor cycles (many fewer misused cycles spent waiting on the memory interface).
I can't believe this topic made it on Slashdot, while the long article linked from Inq^Reg about reflow pipelines on the Netburst article didn't. Or maybe I can believe it and I don't want to.
How about pursuing a solution actually meant to address this problem: the newly announced Via C7 with crypto acceleration that outperforms either AMD or Intel by a ballpark factor of 100:1 in work accomplished / energy consumed. As far as can tell, Via has not yet published the Padlock documents online on Monty mult or the SHA extensions. That would be worth reading. This topic is crap at 37 degrees ambient.
I have a Pentium IV Prescott that runs pretty hot. To reduce the noise of the system fans and still run at a decent speed, I tried underclocking it. The drawback: it started measuring time faster, and all of a sudden it would say it's 8pm but the time was actually 7:50pm. Since I run a PVR in the computer, that isn't good news and I started losing some shows.
Just something to keep in mind if you try to underclock.