Slashdot Mirror
Is Your P4 Working At Half Speed?
"More information can be found in Intels Pentium 4 Thermal Design Guidelines (check out page 23)."
Several readers have submitted news of this clock-throttling, one aspect of the P4's built-in temperature sensor system Intel calls "Thermal Monitor." One thing to point out is that the same design guidelines document goes on to say that "the clock modulation feature of Thermal Monitor is disabled by default ... OEMs are expected to enable the thermal control circuit while using various controls and outputs to monitor the processor thermal status." Other things being equal (even if they never are) is there some reason to prefer a chip for not having this capability? If someone forced me to accept a free and loaded P4 system, I'd rather it be cool down at 750MHz temporarily than toast at 1.5GHz.
pentium4 comes directly from God, and questioning God's precision is blasphemous
Yup, linuxppc is pretty cool.
Although that's not really acceptable for a released, commercial processor, I must say that it sure beats overheating!
:)
Why can't processors dynamically adjust their clock speed based on temperature in the first place? Transmeta does this somewhat, but it'd be nice if my chip could overclock itself, insofar as that is safe.
---
pb Reply or e-mail; don't vaguely moderate.
pb Reply or e-mail; don't vaguely moderate.
You can read about the temp specs in this document: (in pdf.. sorry.. that's all I could find)
http://www.amd.com/products/cpg/athlon/techdocs/pd f/23794.pdf
It says 90C is the maximum die temperature. I guess that means you should start worrying if it hits 80C. My 900MHz T-Bird seems to hover between 60-70C. That bothers me, but I'm not sure what else I can do about it. I have a good heatsink that should keep it a lot cooler than that, but for some reason it doesn't seem to be working. I also added 2 extra fans to my case to help with airflow, but nothing seems to work.
It's not enough to bash in heads, you've got to bash in minds. - Captain Hammer
There is nothing wrong with having a thermal throttle. In fact, it's a good feature.
The real problem shows on page 25 or the Intel doc. That's where they reveal that the rather hefty heatsink requirements are targeted to be adequate for only %75 of full power.
In other words, if you do compute intensive work, you will engage the thermal throttle on a regular basis unless you install an even bigger and heavier heatsink with even better fans.
Most interestingly, it turns out that the full utilization dissipation is more like 73Watts. In other words, it's not actually significantly cooler running than the Athlon, in spite of marketing's sincere wishes to the contrary.
So, the only real 'advantage' over the Athlon is vapor.
Notice however that his statement of 100% utilization was accompanied by "I track its performance and I can assure you that it has not ever slipped into the throttling"
He is not saing that he expected the cpu utilization to drop top 50%, He is saing that he is running his apps at 100% cpu utilization and didn't notice any performance drop.. meaning that his cpu never kicked into the throttle mode. This 100% is meant to say that he is STRESSING his cpu.
You can have a dual p4; you just have to have one of them disabled at any given time...
Actually, I wonder if they could build a motherboard where, if one CPU got too hot, it would switch over to the other one.
Well lets just strap a box-fan on the side of the thing!!
Hehehe looks like someone did..
Your Working Boy,
- Otis (GAIM: OtisWild)
Okay, so this clears things up a little.
So, basically,
1) if you have proper cooling, it won't ever come into play (which sounded like it was the case anyway, since none of the benchmarks seemed to show this behavior)
2) it is still somewhat underhanded to advertise the part as having a power consumption of 54W, making the P4 seem as though it consumes less power than the K7.
3) it is still pretty silly for AMD _not_ to have some kind of thermal protection (though again, if you have proper cooling, it shouldn't be a problem).
Now that we've got that cleared up...
The enemies of Democracy are
It's not a rumor, the thing exists! A 4-way Foster (codename for the Xeon part) system called P-Shasta (which I got to work on! =D) was demoed last year.
As to why it ain't out yet, I can't tell you. Sorry.
The enemies of Democracy are
Remember, 100% utilization at 1.5Ghz is going to look the same as 100% at 750MHz. The useful info would be to run a program alongside that could monitor the current clock of the CPU. Either that or keep track of your work submissions to make sure they don't drop off.
What's the difference between a Java applet stuck in a non-terminating loop chewing CPU, and one that's, say, searching the key space for a distributed code-cracking problem? As far as the Java runtime environment, nothing that it can identify.
If I understand you correctly, the problem that you are suggesting that Java should solve is a variant of the halting problem, which is practically undecidable (it's theoretically possible for storage-limited machines, but not in any practical sense). In other words, it can't be done.
Java can be faulted for many things, but this doesn't appear to be one of them.
Go you big red fire engine!
Any sufficiently advanced technology is indistinguishable from a rigged demo
--Andy Finkel (J. Klass?)
I think intel made the announcement to drop prices to cover for this embarassing admission - so you still loosely 'get what you paid for'.
Sounds fair to me.
Further than who is prepared for? Who are you to make this judgement? Who are these 'masses' you have so little respect for? Who is moral?
Your post bespeaks of a quiet and pig-headed elitism that does you ill. I suspect you are probably a troll, and a stupid one at that.
In addition to this problem with the P4 having nothing to do with the advance of tech in general, the advance of technology will never be halted as long as humans exist. If you feel that technology should or should not be used in certain ways, you'd better convince people you're right instead of hoping in vain for the advances to cease.
Need a Python, C++, Unix, Linux develop
Some people (usually overclockers) have been doing something similar to this for quite a while. The way it's worked until now is that they run a program (waterfall for windows, or automatic under Win2k and {of course} Linux) that issues the idle command if the processor is otherwise idle. This saves energy, reduces heat, blah, blah, blah.
The problem is then exactly the same as here:
You've got a super-fast CPU, but you can't actually use it for long, intensive tasks.
(i.e. 5 minutes into a game of Quake, the system will become unstable)
The solution? *shrugs* Hard to tell. Wait for a CPU that won't do this. (AMD!) and don't overclock so much.
Note: I love overclocking. I've overclocked every system I've had since my 483/33 My current main system won't overclock more than 5MHz. Life sucks for me. But I'll overclock my next system.
Wombats: The Bulldozers of the Bush.
Nathan Brazil?
No, reading Slashdot isn't THAT taxing on my cpu.
Unless you're using Mozilla.
I have always wondered whether it would be possible to design chips with integrated heat sinks, or air holes to allow better circulation of air around the chip. This would certainly help in the cooling process.
Jumpstart the tartan drive.
/. is mangling my posts this morning, so I'll keep this simple on this third try...
Go to Linux Today, and read the announcement about Alan Cox's 2.4.3-ac7, which just came out today.
Maybe it will help you; maybe it won't.
Maybe Slashdot will accept this post; maybe it won't.
--
Sheesh, evil *and* a jerk. -- Jade
> I think you should ammend your "many people have burnt or cracked their chips" with "people who do not follow AMD's cooling recommendations or improperly try to force a non-socket A cooler on to their chips".
Do you know where I can find AMD's temperature recommendations? I just built my first system with sensors on the motherboard, and I got the lm_sensors stuff working with the 2.4.3 kernel Saturday, so now I can watch the temperature on my desktop with gkrellm. But I don't know what I'm looking at. It usually hovers around 50C, but sometimes climbs as high as 57C when I've run an all-night number crunching job.
At what point do I start worrying?
--
Sheesh, evil *and* a jerk. -- Jade
"This is not a bad thing, but a good thing."
Ummm, I beg to differ. It *is* good that the CPU doesn't burn itself out - yes. But it is bad that the CPU kicks into this mode on common applications.
If Intel is selling this chip as a performance solution then I want to be able to run it at 100% for 4-6 hours, for a kicking game of Q3 at a LAN party. I want to be able to do FP work 24/7 while I render an animation. If their CPU can't do these perfectly reasonable things they should mark it as such.
Look at it this way - if Intel was selling a CPU for use in WebTV (or some other embedded application) and the user was able to do something, rapidly opening and closing a window for example, that overheated the CPU and it slowed down - that'd be fine. It'd still function as a WebTV box, and while the user was screwing with the CPU it wouldn't be doing anything else.
Contrast that with them selling a CPU that they claim is a performance monster, better than any other x8 CPU. People expect to run their x86 machines at 100%, many servers at my work have been up for months and have been at nearly 100% load the entire time. When I installed a bunch of dual celerons a few years back as a rendering farm, they were at 100% for 18 months, running FP rending - about the hottest code you can get.
If the P4 is sold to a market that expects this performance, it had better perform.
If the only way to get a chip to overheat was to execute a tight loop of CPUID instructions, or something else that had no real purpose, then I would accept it kicking in thermal limiting. That's beyond reasonable operating specs (as in, performs no function, exists just to stress the product.)
To use a car as an example - if I buy a car, I expect it to be capable of highway speeds for prolonged times. It's not okay that the company profiled drivers and decided that 99.2% of the time, a duty cycle of 70% 30mph, 20% 50mph, and 10% 60mph was all that was required. I'd expect the car to be capable of handling 65mph 100% of the time, for 8-10 hours. If it couldn't do this, I would like a thermal limiter where it warned me gracefully instead of blowing up, but I'd still take it back as defective.
Copied Verbatim from www.hardocp.com
Now that you guys have had a couple days to get excited about what Bert McComas had to say about the P4 clock throttling itself, I thought I would throw in my two cents. First off, here is the statement that is stirring the pot so well.
Intel's Thermal Design Guide has revealed that the absolute maximum power dissipation of the 1.5GHz P4 is actually 72.9 watts. This is 33% higher than the published system design specification, and essentially identical to the 1.33 GHz Athlon. In order to prevent the CPU from exceeding 54.7 watt, thermal throttling is used. If performance critical applications drive CPU power above its artificially low 54.7 watt limit, the CPU is halted with a 50% duty cycle (alternating 2 microseconds on; 2 microseconds off) until it cools down. This effectively turns your 1.5GHz processor into a 750MHz processor - just at the moment you demand peak performance. On the other hand, you will probably still be able to check your email at 1.5GHz.
While I don't know Bert, I have had the pleasure of meeting him and you have seen his links here on the [H] many times. On this occasion I think Bert has been sucking the crack pipe a bit too hard or either must have been in a terrible auto accident and had his cranium lodged in his rectal cavity and did not notice before he wrote the above statement.
We have been running an over-volted overclocked Pentium4 with the factory heatsink installed now for some time. It has been running here beside my desk folding proteins for Stanford University now for a solid month and has stayed at 100% CPU utilization. I track its performance and I can assure you that it has not ever slipped into the throttling that Bert speaks about above. If Bert's apps are running at 50%, it is because he does not have the sense to put a heatsink on the CPU or either he is operating his P4 system in Hell. Bert is taking an Intel safety device and demonizing the P4 with it. Here is what Intel PR George Alfs had to say about Bert's statements.
Hi Kyle,
You can run benchmarks all day on a Pentium(R) 4 processor with the benchmarks unaffected by the thermal protection circuitry. The key is to have a robust heat sink and thermal solution. With the heat sink setup we designed for Pentium 4 processor systems, I have yet to see thermal protection kick in.
George.
I have to fully agree with George's statements and have a few things to add. Also, I think that "robust" need not be in his statement.
What Bert may not know is that some mainboards have an adjustment in the BIOS that you can set yourself with the temperature that you want throttle to. (On our systems we have left it at default and never messed with the settings on the particular board that is Folding.) Yes, YOU can turn this on and off and fully control it on some mainboards. If you want a shield in between you and a burned up processor, set it low; if you want to forego the safety feature, set the temp high. I know that MANY of you wished that AMD had the courtesy to include a feature such as this instead of leaving their Athlon and Duron CPUs totally unprotected.
We have never seen nor heard of the CPU throttling being active on any person's CPU and certainly have not experienced it ourselves (unless we FORCED it to happen) under conditions more strenuous than 99.9% of the P4s in the field will ever encounter. I do not suggest that DIYers or hobbyist go the P4 route if they want to buy a system for themselves, but bashing it on this front is simply bad journalism and transparent to many people.
We here at the [H] have a lot of respect for Bert McComas' work and think he should step back up to the plate and possibly rephrase the statements in regards to this issue. Bert, we love you man, but you were just totally out to show Intel in a bad light this time, or were simply not thinking through the issue properly because you are being misleading and it looks to us as if you were trying to do it purposefully.
In case of Emergency, Curl up in the Fetal position, and lick a Bible for comfort!
Koolance is a company with a "mainstream" water-cooled case.
And it's quiet because water cooling is not only used for the CPU, but also hard disk, power supply and graphics card.
Of course, I've seen 4 way powerpc systems that have no cooling required...
Later,
Thad
The Bolachek Journals
The PIII Systems Programming guide has a section on thermal control for the P6 core as well. Read chapter 12: System Management 12.14 Thermal Monitoring. On the P6, software can directly control the duty cycle or a default behavior can be programmed in to take affect when the system goes over a certain temperature. The on demand clock modulation can be a lot worse than 50%, the values range from 12.5% to 87.5%... Oh, BTW this stuff is almost always controlled by the BIOS so running linux won't change its behavior.
Does this mean that people who over-clock thier P4 systems are auctually likely to SLOW their cpu down?
No, reading Slashdot isn't THAT taxing on my cpu.
When things get too hot at work I switch into my brain-cooling mode. I spend 0.2 microseconds working then switch off to slashdot reading mode for 2 seconds. This modified duty cycle allows time for my brain to cool down. Once frustration levels are within acceptable tolerances, I switch back into 100% work mode.
Read the end, he isn't serious. Give him a Funny or 3.
-----------------------
Nicotine free Amish .sig.
Perhaps there's a reason to prefer a chip that doesn't get so hot in the first place. I heard Motorolla and IBM make one, and that some company has released a Unix variant that runs pretty well on it.
Introducing, in the 21st century, a chip without SMP capabilities, and that will still require you to buy a whole new mobo?
You mean the Athlon? Or are you running quad Athlons are your Socket 7 motherbaord from 1995?
The article says that a thermal diode is responsible for triggering the throttled-down performance. But then it also says that the throttling happens due to the power consumption. These are different things: power consumption only causes the temperature to rise if the cooling is not slurping off excess heat fast enough.
Anyone care to comment on this seeming discrepancy?
Assuming that it really is thermal throttling, I would love to see what a good tech site like Tom's might be able to determine about the throttled down CPU when using various heatsinks. If that feature is really there then you should expect more powerful heatsinks give the same temperature as lesser heatsinks, but higher performance.
In other words, it is possible to see this as a feature, not a bug. You get 1.5G when the processor is capable of it. You get half that when you are running hot; but with good enough cooling you should always get the highest performance possible.
"Overclocking" may go away, replaced by "overcooling".
-- props to Wreck (leonard@dc.spam.net) --
Reality is just a clever Hack, and the Planck constant is the refresh rate.
Back when computers were sold to engineers, they came with a spec sheet with the environmental specs: allowable ranges on power, temperature, air pressure, and humidity. Compaq still provides them. Here are the numbers for a PIII desktop machine:
Note that any environment that isn't air-conditioned and humidity-controlled will probably go outside those ranges at some point in the year. So some form of overheat protection is essential to prevent component damage. A slowdown is better than a shutdown, which is better than a crash, which is better than a meltdown. You want thermal throttling, fan speed control, and emergency overtemp shutdown on anything used for more than Quake. Fortunately, those features only add a few dollars today.
Surprisingly, Compaq's laptops have a 95F temperature limit, which is on the low side for a portable device.
There used to be a saying in railroading: "Never buy equipment from a supplier who's in a better climate than yours". All the good railroad suppliers were in places like upstate New York, where they experienced snow, ice, rain, heat, and thunderstorms. Computing could use more of that attitude.
This has been said of Intel over and over again. They were deceptive liars when they released the Pentium III, which was almost exactly the same as the Pentium II, except for the CPU ID (and perhaps other minor differences), and that was coupled with the privacy concerns of the CPU ID, which was going to ruin Intel by erroding trust amongst consumers.
The Pentium Pro was also going to ruin Intel, as it was so expensive and didn't seem like it'd ever be worth the money for something that didn't perform much better. And so on... people have said more or less this same thing most of Intel's new processors. Ok, maybe this time it really will happen, but much more likely is that history will repeat itself yet again.
In fact, the only time Intel's ever really had any major trust problems was when the FDIV bug hit, and when they finally did the right thing and offered to replace any FDIV-bug chip for free, consumer/business's trust was almost fully restored.
This certainly isn't the first time there's been a slow-down in the market.
Moore's law has been predicted to have run dry many many times. Right now doesn't seem like such a good time to be forcasting the end of Moore's law, since short-term incremental improvements (1.7 GHz up from 1.5 GHz on the P4) and long-term improvements (IA64, async "clocking", even finer geometry transistors in the lab, etc) are in the making.
Just as predicting "everything the can be invented has been" didn't work in 1899, it's incredibly short-sighted today.
Speech recognition isn't too hard to imagine today. While it isn't likely to become the primary way of interacting with the computer (ala Star Trek), it will certainly become a high-demand feature when it's refined and cost effective. Among other benefits, speech recognition may really open up the possibilities for people communicating with one another by email and discussion forums (like this one), as a great portion of the population has reasonably good speaking skills, but typing messages is "hard work".
It's also not too hard to envision future software parsing natural language, at least with some level of success in understanding the meaning. Today's computer interfaces aren't much more sophisticated than caveman's point-and-grunt (well, maybe except for geeks/programmers who can use the command line). Today's successful user interfaces tend to build their success by arranging objects to be pointed at... but it's easy to see with the massive growth of available information on the web that point-and-grunt doesn't scale well. Quite a lot of research has gone into this dream. In fact, the aim of languages like XML are to facilitate computers being able to "understand" the information, so that new methods of interaction can be built (well, there's other shorter-term benefits too) When/if natural language parsion becomes a useful interaction technique, it will be very compelling (aka a "killer app") and today's computers will seem as ancient as black-n-white television (or perhaps an old Apple ][).
There's many other amazingly short-sighed quotes lurking in sociology's post (hard to believe 3-4 people mod'd it up as insightful), but perhaps the best is "Computers are at the base of all our technological advances." Perhaps that could be said of the written language or maybe even the printing press.
There's plenty more to be commented upon, but, dear moderators, please take a moment to ask yourself how insightful is a viewpoint with very limited historical perspective that predicts no advanments in the future? Sounds to me like the wishful thinking of a luddite.
PJRC: Electronic Projects, 8051 Microcontroller Tools
In related news, President Bush's latest gaffs were explained by stating that his Pentium 4 powered aritificial brain went into thermal throttling mode.
The Secret Service is now charged with the responsibility of making sure his Peltier cooler is always firmly attached to his forehead.
Just because it CAN be done, doesn't mean it should!
This story was posted and discussed at length 3 days ago...
And that has what to do with the article? You're comparing a PIII vs an Athlon, neither of which is a P4. Besides that, there's a great number of things that can affect performance. The processor is but one of them.
Page 26, they have a neat little graph that says that if the chassis has 70% of some optimal cooling capacity, the throttling won't happen. So since this is will keep an overheated processor from being cooked, I can only see this as a positive.
I demand a million helicopters and a DOLLAR!
The comments here make it sound like running a simple photoshop blur will make the CPU slow to a crawl in order to avoid going tits-up in under 5 minutes.
For all we know, the only way to cause the clock throttle to choke up is running the chip heavilly with no fan in close proximity to a blast furnace under direct sunlight for three days straight...
Just how hot is "too hot", and how much do you have to abuse the chip to get there? Until that question has been answered fully, I am going to view this as a lot of panic over nothing.
Information wants to be anthropomorphized.
The guys at [H]ardOCP addressed this issue. Look under Sunday, April 15, 2001 - Ed 2. "We have never seen nor heard of the CPU throttling being active on any person's CPU and certainly have not experienced it ourselves (unless we FORCED it to happen) under conditions more strenuous than 99.9% of the P4s in the field will ever encounter. I do not suggest that DIYers or hobbyist go the P4 route if they want to buy a system for themselves, but bashing it on this front is simply bad journalism and transparent to many people." Sorry, I don't know of a way to direct link to an article on their site...
I had one.. heated the whole house.
---
"Watch these suckers jump when I get Administrator."
Ahem. I am still waiting for real time raytracing to render asymetric multiprocessing via peripheral pseudo-"GPUs" obsolete.
---
"Watch these suckers jump when I get Administrator."
This wont happen if GOOD heatsinks/fans are used. These processors get hot. I wish AMD did this, as AMD is smokin. But for the price, you cant argue on AMD. Though many AMD users have either burnt or cracked their chips, so having protection like this isnt always a bad thing. just use good heatsinks and you'll probably be just fine and dandy.
Linux: Because a PC is a terrible thing to waste.
James Brents
That may not be too good either. Having too many fans can cause bad circulation pushing air back and forth to each other, doing nothing at all. Check out a pic at this site towards the bottom how "ATX Standards" show how to cool off the system...
.kb
Honestly, Out of the two types of cases I have, I have a In-Win Q500A and two Enlight Cases (Full and Mid-Tower), the Enlights stay super cool, the panel is always cold to the touch, the In-Win burns up on the side...I love my Inwin case though...its just always HOT..
Two Wrongs Don't Make A Right-- But They Make Me Feel A Whole Lot Better
Ok, I see let me explain... The primary benefit is lower packaging costs in two ways... As I tried to state before (but didn't say properly), the packaging costs go up by about $1 per watt for each watt over 30... until you reach around 130W. At that point, air cooling will no longer suffice, so you have to use an alternate cooling method which costs lots more (this is always how I'e heard the rule stated--I don't knwo what the alternate method is). So, this techinque allows manufacturers to make processors which have cheaper packaging (saving money) as well as avoiding that 130w number (saving more money). Now, I understand your criticism of the technique. If it kicks in all the time, then it's probably not worth it. But that's not what this is targeted at. As has been stated, this technique never kciks in for 99.9% of people with a properly attached HS/fan, so it's more of just an emergency mechanism (in case the fan dies etc). Now, eventually (as has been published in the literature) people expect this mechanism to be applied more aggressively, so I'll briefly discuss that (but again, I don't think this applies in this case, where it is mostly just an emergency escape mechanism). Most applications are phase oriented... they do work for a while, then stream in a bunch of data, then more work etc. Well, as you probably know, during these data streaming periods, the processor is just sitting around doing nothing, so that would be an opportunity to cool off. Now, some applications (like the FP benchmarks from the SPEC suite) are highly highly optimized, so they spend their entire time in a small loop doing work all the time without ever pausing... Applications such as these which are constantly doing high heat operations (floating point operations generate a TON of heat) are going to trigger the mechanism. But, the majority of apps have this phased behavior where they intersperse low heat work periods between the high heat periods, and most likely, will usually avoid triggering the mechanism. Did I make the distinction clear? If you're interested, I could point you at some papers on the topic which do a better job explaining this than me (power isn't my area to be honest). So, to restate what I said, the benefit to the consumer is lower prices due to lower packaging costs, as well as possibly a higher performance processor in the general case. I don't know *anything* about layout, but I would imagine that you could use this technique also when the heat density of certain parts of the chip were getting to high. Where you get to place different components on the chip make a huge difference in the performance, so maybe you could use this mechanism to allow you to place, say, the fp function units closer to the register file, knowing that if the heat density in that region got too high you'd invoke this technique.
I must burn in hell, suffer and pay for my sins
But Gods the one who's losing, Satan always wins!
Come on people... you even made me log in to reply to this one... This is not a bad thing, but a good thing. Power has been a problem for cpus for a while. This is, in fact, actually a quite cool feature. Generally how processors are designed is you get some guy to generate the maximum heat producing code that he can find. It doesn't do anything useful, and generally consists of lots and lots of floating point instructions. Then, you find out how much this heat this program generates when run on your processor. Now, you design your processor to be able to tolerate the heat generated by this program. However, first of all, no program that does anything useful will ever generate as much heat as this test program. SO really, you're forced to design your processor packaging for a way overkill case just to be sure that you don't have your processor die when someone is doing legitmate work with an unoverclocked processor. It has been shown that packaging costs increase by about $1 per watt generated by a processor for every watt over $30, so you can see that developing your packaging for the worst case scenario can be quite expensive. The alternative, then, is exhibited in the P4... Build your processor packaging for less than the worst case, then use some form of thermal throttling to prevent overheating. This has two advantages 1) It lowers your packaging costs 2) It prevents processor death in the case of catastrophing failure (such as a fan dying). I expected that people would get up in arms about this feature, but really, most of you just need to learn about the most recent research in this field to see this is actually a step in the right direction. However most people on slashdot are primed to jump on Intel at every opportunity, so they interpret this in the worst possible light. And BTW, I'm getting my PhD in computer architecture, so I know what I'm talking about :P. There have been papers at all the major conferences for the past few years dealing with power issues, and I might work on one myself soon.
I must burn in hell, suffer and pay for my sins
But Gods the one who's losing, Satan always wins!
Yes, it is a good thing that the chip throttles itself down instead of melting down when it gets too hot. And apparently (contrary to the article), it won't throttle down if it has enough cooling. The issue is that Intel's specs are misleading as to what constitutes enough cooling -- they say 53W, but to actually get all the computing power you paid for, you need a 73W heatsink. It's like advertising a car as capable of 150 MPH and 50 miles to the gallon, but burying in the fine print that you get that gas mileage by turning off the V8 and running 10 mph with a lawnmower engine.
I guess my P60 would kill at Pentium 4 750mhz. It's not that bad, by almost an order of magnitude -- that is, you'd need a 500 or 600 MHz chip to beat the throttled down P4. And it only throttles down if your cooling isn't good enough. But the main point is, unless your software has been recompiled specifically for the P4 (and I don't know if there are compilers that will do that yet), the P4 will give you less performance for more cost as compared to the 1.2GHz AMD. The release of the P4 in its present status seems very premature...
that sounds like why-your-intel-processor-is-bad-hype to me.
you may want to read this posted on hardocp.com:
Hi Kyle, You can run benchmarks all day on a Pentium(R) 4 processor with the benchmarks unaffected by the thermal protection circuitry. The key is to have a robust heat sink and thermal solution. With the heat sink setup we designed for Pentium 4 processor systems, I have yet to see thermal protection kick in. George.
(George = Intel PR George Alfs)
And here is a small excerpt from Kyle's comment:
We have been running an over-volted overclocked Pentium4 with the factory heatsink installed now for some time. It has been running here beside my desk folding proteins for Stanford University now for a solid month and has stayed at 100% CPU utilization. I track its performance and I can assure you that it has not ever slipped into the throttling that Bert speaks about above.
keep it simple.
Props to Intel! What a way to keep their system in shape! Heat sinks and fans are already at the highest level of heat displacement they could possibly get. Fans have not evolved for decades - they've only gotten bigger. With normal methods of heat displation, Intel needed to find a new way to get rid of the excess heat the P4s create, and what better way than to have them alternate their cycles as on/off? Sure, you might only get half the clock cycles, but the ones you do get, are two times faster, so you are getting even more performance than a regular chip! I can see this coming into wide use in the server market, where most caculations are quick file access requests, because the average user doesn't need pure horsepower ontheir file server, they need faster disk speeds. I think this new design might also save some power in West, where the bulk of those machines will be bought and used. Then, the other west coast states such as Arizona, Nevada, and Texas can sell their extra power to California. Keep up the innovating Intel! I can't wait to get myself a dual proc p4!
The post that started this whole thing comes from someone who misunderstood the table they were reading. The processor has a THERMAL sensor that kicks in when the processor overheats. The table lists 72 degrees C as the max for the P4 1.5 (Pretty dang hot) The table (see http://developer.intel.com/design/pentium4/datasht s/24919802.pdf page 70) recommends that those designing a cooling solution design it for at least 54.7 watts since that is a normal load on the chip. A thermal senser senses TEMPERATURE not the wattage the chip demands. So if the chip stays below 72 degrees C, no worries. AND... If you want to know how it runs, I'll tell you. SYSTEM CONFIG: P4 1.5 Boxed (stock fan), 256 Meg RDRAM, ATI RADEON64, Enlight 7237A34 case (stock)and D850GBAL boxed motherboard. The Fan1 and 2 connectors are temperature controlled, and the processor fan didnt even come on for the first 5 minutes from room temperature startup - wasn't needed yet... The case fan NEVER came on until we heated the inside of the case with a heat gun. (It came on at 49 degrees C) The stock Intel CPU fan ran at 2800 rpm while the processor was at a stable 34 degrees C. With the heat gun on the heatsink for a couple minuted I got the processor to register 52 degrees C and the fan was at 4500 rpm plus. After removing the heat gun, the chip colled back to 34 degrees C in about 1 minute. Now - here is the clincher for you naysayers... I ran 3DMark 2001 full test looped and never got above 45 degrees C. SO WHAT IS ALL THE FUSS ABOUT ?!? I think it is a great solution. Oh well- Some people just WANT Intel to be evil. I know this message wont even be a blip on anybodys radar screen, but hey - gotta say SOMETHING.
BTW- Another article citing SpecInt2000 scores distorts the truth in the same way they claim Intel does. They only give you the information they WANT you to see. The score they show was just one part of the test. On the rest of the test the spread was much more significant.
Just a final note: Always check their sources. Even when they say they are quoting from the manufacturers specs/facts, they sometimes interepret them wrong or distort them.
DONT BE BRAIN DEAD AND JUMP ON ANY WAGAON THAT DRIVES BY !!!