Message boards :
Number crunching :
CPU cooling issues
Message board moderation
Previous · 1 · 2 · 3 · 4 · 5 · Next
Author | Message |
---|---|
HAL9000 Send message Joined: 11 Sep 99 Posts: 6534 Credit: 196,805,888 RAC: 57 |
However overheating other components in the process would ultimately be a bad thing. I just look at 7 machines near me in my lab and there are 5 different cooling setups. That is why standards are good. Because there are so many of them. SETI@home classic workunits: 93,865 CPU time: 863,447 hours Join the [url=http://tinyurl.com/8y46zvu]BP6/VP6 User Group[ |
OzzFan Send message Joined: 9 Apr 02 Posts: 15691 Credit: 84,761,841 RAC: 28 |
However overheating other components in the process would ultimately be a bad thing. LOL Agreed. Perhaps if the engineers all agreed, we'd have a single standard! ;-) |
RottenMutt Send message Joined: 15 Mar 01 Posts: 1011 Credit: 230,314,058 RAC: 0 |
i think we can all agree that the cpu needs to be kept cool. the other components on the motherboard surrounding the cpu are rated for high temperatures and only need to be kept below their ratting. when it comes to fans you will get higher cfm by pushing air thru a heat sink the pulling the air thru the heat sink. fans are real sensitive to any obstruction on the intake side. as far as conventional cpu heat sinks go, they are a solution (crappy) for the lack of a proper solution. most oems now push the air straight thru the case rather then blow the air around the case and hopefully the hot air makes it out. unfortunately this is the design of all after market cases. |
Richard Haselgrove Send message Joined: 4 Jul 99 Posts: 14653 Credit: 200,643,578 RAC: 874 |
Wow, that seems rather absolute-ish. So you've talked to every other engineer of stock computers? Seems funny to me because every CPU fan I've ever received pulled the air off the CPU. It's only recently that CPU fans seem to be blowing the air onto the CPU. If this is only a recent thing, how could it be that every egineer is at odds with my findings? You prompt me to lift the lid on the Dell P4 I use as a footrest while typing here :-) No CPU fan at all. A big finned heatsink, enclosed in a duct/shroud, leading to a dedicated case fan exhausting through the backplate. That seems to be a pretty good solution, though I would wonder whether it could cope with the heat production of a Core2 duo or quad. At least these P4s seem to throttle themselves down, or in extremis shut off completely, when things go wrong with the cooling system: no permanent damage is caused, and they restart as normal once they've cooled down. My only other experience is with AMDs of the previous generation - whatever ran in parallel with the 386 and 486 ranges. I've seen examples where the cooling fan failed (whichever direction it was supposed to have been moving the air in), and the CPU continued to generate heat at seemingly full power. That led to a scorch-discoloured CPU socket, and a very expensive repair job. |
jason_gee Send message Joined: 24 Nov 06 Posts: 7489 Credit: 91,093,184 RAC: 0 |
i think we can all agree that the cpu needs to be kept cool. the other components on the motherboard surrounding the cpu are rated for high temperatures and only need to be kept below their ratting. From an Engineering standpoint, I think this one comes the closest so far to showing where the bones of contention lay here. What's easy to forget with an engineered system, good or bad, and I've seen plenty of both in repairing, upgrading & building PCs of both the home/office workstation and server variety, is that factors like system airflow and cooling come down to design optimisation versus 'fit for purpose', cost & profit margins, as well as maintenance and usability considerations. So as with software, there are many possible approaches, & what matters is what works for you & your design/changes rather that what works well on PC XYZ of a different design under different usage conditions. i.e. 'your mileage probably will vary' Motherboard design & layout certainly plays a big part, and there's been a trend lately to putting whopping heatpipes & heatsinks on the north/south-bridge & mosfets (incidentally typically rated to 150 degrees C!), which apart from looking pretty, does make the options more flexible with respect to case airflow design. As Power supply requirements have grown, there has also been a trend to go to cases that mount those in the bottom receiving cool air, and CPU heatsink designs that crossflow. In nearly all situations, mixing design approaches to end up with multiple factors conflicting, as opposed to design from a holistic standpoint, will lead to some or another thing not doing what 'it should'. I don't think any of us here perform computational fluid dynamics modelling of our systems, combined with testing every style & model of heatsinks, fans, case layouts, with different cost & performance goals etc, so I think in this situation heavy doses of 'whatever works' need to be applied. Jason "Living by the wisdom of computer science doesn't sound so bad after all. And unlike most advice, it's backed up by proofs." -- Algorithms to live by: The computer science of human decisions. |
OzzFan Send message Joined: 9 Apr 02 Posts: 15691 Credit: 84,761,841 RAC: 28 |
No CPU fan at all. A big finned heatsink, enclosed in a duct/shroud, leading to a dedicated case fan exhausting through the backplate. That seems to be a pretty good solution, though I would wonder whether it could cope with the heat production of a Core2 duo or quad. Thankfully one of the selling points of the Core 2 series was that they ran more efficiently (higher IPC) and cooler than Pentium 4's. I would bet that a Core 2 Duo would cope with a passive heat sink solution just fine. The Core 2 Quad though might be a different story. At least these P4s seem to throttle themselves down, or in extremis shut off completely, when things go wrong with the cooling system: no permanent damage is caused, and they restart as normal once they've cooled down. That same ability is supposed to be present in all of the Core 2 chips and newer, as well as most newer AMD chips. My only other experience is with AMDs of the previous generation - whatever ran in parallel with the 386 and 486 ranges. I've seen examples where the cooling fan failed (whichever direction it was supposed to have been moving the air in), and the CPU continued to generate heat at seemingly full power. That led to a scorch-discoloured CPU socket, and a very expensive repair job. I'd be willing to bet you're actually referring to the original Athlon CPUs without overheat protection. They were infamous for burning up and literally melting CPU sockets if the CPU fan failed. AMD's 386/486 line didn't run nearly hot or fast enough to do any scorching or socket-melting. |
jason_gee Send message Joined: 24 Nov 06 Posts: 7489 Credit: 91,093,184 RAC: 0 |
Well this thread is a pretty good example of how most Engineering meetings turn out (that I've been to anyway), and why most off-the shelf systems end up coming in below user expectations, despite the best intentions from all parties :) "Living by the wisdom of computer science doesn't sound so bad after all. And unlike most advice, it's backed up by proofs." -- Algorithms to live by: The computer science of human decisions. |
kittyman Send message Joined: 9 Jul 00 Posts: 51468 Credit: 1,018,363,574 RAC: 1,004 |
Well this thread is a pretty good example of how most Engineering meetings turn out (that I've been to anyway), and why most off-the shelf systems end up coming in below user expectations, despite the best intentions from all parties :) LOL.....an engineering meeting, then, eh? "Freedom is just Chaos, with better lighting." Alan Dean Foster |
J. Mileski Send message Joined: 9 Jun 02 Posts: 632 Credit: 172,116,532 RAC: 572 |
|
kittyman Send message Joined: 9 Jul 00 Posts: 51468 Credit: 1,018,363,574 RAC: 1,004 |
I have several 120mm sidefans running next to the GPUs........ Somebunny does not believe that extraterrestial cooling is normal.......LOL. "Freedom is just Chaos, with better lighting." Alan Dean Foster |
HAL9000 Send message Joined: 11 Sep 99 Posts: 6534 Credit: 196,805,888 RAC: 57 |
At one point in time I had used a vacuum cleaner motor to pull air through a case. It was a big excessive in noise and power usage though. SETI@home classic workunits: 93,865 CPU time: 863,447 hours Join the [url=http://tinyurl.com/8y46zvu]BP6/VP6 User Group[ |
AndyJ Send message Joined: 17 Aug 02 Posts: 248 Credit: 27,380,797 RAC: 0 |
Well this thread is a pretty good example of how most Engineering meetings turn out (that I've been to anyway), and why most off-the shelf systems end up coming in below user expectations, despite the best intentions from all parties :) Yup, I agree. I was told once by an engineer, (my boss at the time,) and I kid you not, "spontaneity has it's time and it's place, you know." Off topic, I know, but this thread could do with a little lift. :-) Kind regards, Andy |
Jim_S Send message Joined: 23 Feb 00 Posts: 4705 Credit: 64,560,357 RAC: 31 |
Play Nicely Folks. (Post Cleaning.) I Desire Peace and Justice, Jim Scott (Mod-Ret.) |
Grant (SSSF) Send message Joined: 19 Aug 99 Posts: 13745 Credit: 208,696,464 RAC: 304 |
During the build up here, all my PC cases have been hot too the touch. During the dry, they're just warm. Here in Darwin we don't have Summer, Winter, Autumn or Spring. Just the Wet season, the Dry season & the Build up what's sometimes refered to as the build down. Sorry, but i disagree. Millions of systems use that cooling method, year after year, without failure. The ATX cases were designed so that the PSU fan was responsible for cooling the system. The motherboards in such cases are designed to make use of the CPU fan to help cool their onboard regulators. And it works, as the systems around the wolrd still running prove. The PSU was an integral part of the cooling system for the AT case design. It is an integral part of the cooling system for the ATX case. It's no longer the sole device for cooling a system, but it remains a major one. I've been doing Seti for a few years now. All my systems have run 24/7, they have all employed stock cooling- ie PSU fan being the main method of cooling the system with the heat from the CPU being blown over the motherboard components in ambient temperatures usually in the mid 30°s (celcius) & the RH often 85% or higher. I have not had one system fail. Of course, we could agree to disagree......... What we disagree with is the statement that the air from the CPU blowing on the motherboard resulted in system failure. I keep pointing out this is the way the ATX system is desgined, there are 100s of thousands of systems out there using this method & are all working without failure as a result. Sure, they could change the layout of the motherboard so that the CPU cooler fan is mounted vertically on the heatsink (like many high performance air coolers) & it is in the midle of a duct that pulls air from outside the case & expells it to the outside of the case & so keeping the system temperature down & improving the cooling of the CPU. However such a re-design would be rather dramatic. Since the BTX form factor died a slow death, even though it was a much more minor redesign & provided better cooling than the ATX form factor i don't see it occuring. The fact is the present system works. So i think agreeing to disagree is about the best we can hope for here. Grant Darwin NT |
Bill Walker Send message Joined: 4 Sep 99 Posts: 3868 Credit: 2,697,267 RAC: 0 |
Just discovered this thread. I think Jason has the best input. You are all trying to discuss many different combinations of hardware, packaging and cooling, which means there are many different "right" answers. Some basic engineering principals: Heat is bad for electronics. Moving air moves heat. This may cool some components, and heat others. Unless you want me to implement specific solutions for your specific hardware (rates available on request) I would suggest you get a program like Speedfan that lets you log temperatures over time, and start a controlled test program. Vary ONE thing at a time. This will find your best solution. The next person will find another best solution, for their package. |
JLConawayII Send message Joined: 2 Apr 02 Posts: 188 Credit: 2,840,460 RAC: 0 |
I would just like to say that whoever recommended against an aftermarket cooling solution and gave you that explanation was a complete imbecile. I really, REALLY can't stand "experts" that don't know wtf they are talking about. They make actual experts look bad by association, and mislead people that want and deserve good information. CPU temperatures can suddenly shoot up for a couple reasons. Typically either your case filters or the heatsink (or both) get clogged with dust. Dust in the filters blocks airflow, and dust in the heatsink acts like a huge insulator, holding in heat and keeping air from moving it away. However, unless you live in a coal mine or have 5 dozen cats it shouldn't be this clogged after 1 month of operation. Another reason it might suddenly shoot up is a bad connection between the CPU and the heatsink. This happens more often if you move the computer around, making the heatsink move around and damaging the thermal paste connection with the CPU. However thermal cycling (turning the computer on and off) can also have this effect if it was a poor connection to begin with. Taking off the heatsink, removing the bad paste or thermal pad, and reapplying thermal paste (like Arctic Silver 5) will fix this issue. Sorry if I repeated info that was already posted, I'm sure dust was mentioned half a dozen times. I know you said you're not a computer wizard, but I recommend still learning about these things so that you can more easily diagnose issues when they pop up, as well as know when someone's feeding you a load of crap like that representative you mentioned. |
OzzFan Send message Joined: 9 Apr 02 Posts: 15691 Credit: 84,761,841 RAC: 28 |
The PSU was an integral part of the cooling system for the AT case design. You need to go back and study AT case and AT PSU design. They never had a fan included in any AT PSU (except their own internal fan to cool off their own components) - ergo the AT PSU was never an integral part of the cooling system. Most systems that were designed with AT formfactors did not require active cooling because the systems never got that hot until some of the last systems (AMD K6-IIs and Pentium MMX systems). This is what prompted the move to ATX for better cooling - and yes, initially using the PSU for cooling. It is an integral part of the cooling system for the ATX case. It's no longer the sole device for cooling a system, but it remains a major one. Yes, it has been an integral part of the ATX design, but with today's multi-core CPUs it should no longer be required to cool a system - especially one running BOINC. The excess heat created by BOINC will kill most PSUs. I've been doing Seti for a few years now. All my systems have run 24/7, they have all employed stock cooling- ie PSU fan being the main method of cooling the system with the heat from the CPU being blown over the motherboard components in ambient temperatures usually in the mid 30°s (celcius) & the RH often 85% or higher. The PSU fan does not blow over the components. The PSU fan pulls the heat from the components and pushes it out the rear. As you can see from my signup date, I've been doing SETI quite a while myself, and I have had to re-examine my methods for cooling because of today's powerful systems and the heat they put off. What we disagree with is the statement that the air from the CPU blowing on the motherboard resulted in system failure. I keep pointing out this is the way the ATX system is desgined, there are 100s of thousands of systems out there using this method & are all working without failure as a result. And this is where we have the confusion: the ATX design does not work the way you think it does. The fan in the PSU pulls the air off the components and pushes it out the back. This is the way 100,000s, if not millions of ATX systems work. Pulling in that much hot air from the components + CPU + the PSU's natural heat itself is a recipe for disaster. The PSU was never designed to handle that much heat. This is why most modern systems have moved the PSU to the bottom of the case; because they were not designed to take in that much heat. I urge you to research this and I'm sure you'll find you're mistaken. So i think agreeing to disagree is about the best we can hope for here. Sorry, I'm not satisfied with that yet. I will yield to the fact that there are multiple ways to cool a system, but you must understand that you seem to have some fundamental misunderstandings of AT and ATX PSU design and their limitations. OzzFan |
Team kizb Send message Joined: 8 Mar 01 Posts: 219 Credit: 3,709,162 RAC: 0 |
|
zoom3+1=4 Send message Joined: 30 Nov 03 Posts: 65758 Credit: 55,293,173 RAC: 49 |
+1 for the last post with no name! At least He's not the Unknown Comic(Gong Show fame, the guy wore a bag on His head). The T1 Trust, PRR T1 Class 4-4-4-4 #5550, 1 of America's First HST's |
Team kizb Send message Joined: 8 Mar 01 Posts: 219 Credit: 3,709,162 RAC: 0 |
|
©2024 University of California
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.