CPU upgrade affects heat?

I am looking to upgrade to a processor of the same wattage and same size, it is about 34% faster though according to passmark cpu benchmarks. Will this in effect lower the heat my computer produces at idle and web browsing use by around 34%? It runs really hot at about 70 degrees at low usage being a w7j, do you think it would lower to below 60 after an upgrade?

 

When you say wattage I assume you mean TDP? Not exactly the same as what you said. That said I would not look for lower temps per se but the good news is I would not look for higher either.

To be blunt, no your thoughts of lower temps are not likely as the TDP attempts do make adjustments for real world usage and thermal requirements.

No your strategy is not good. You have a bad thermal solution or need to clean your notebook. Or maybe apply a thermal grease. Or maybe this is the way your system runs?

What model again? Not getting it from your post, what CPU?

 

I don't think it will reduce temps. What you probably need to do is clean the vents. Take some compressed air and blow out the heat sink. Put a tooth pick in the fan so it doesn't spin. Or if you chose to upgrade the processor. When you are installing the new processor, blow out the heat sink. My favorite thermal paste is Antec Formula 5. t lower my temps by 8C.

Here's a picture with a crude drawing I made.
Thats supposed to be a can of air by the vent Blow the ar through the vent, and dust come out the fan.

Here's a full tear down guide: http://forum.notebookreview.com/asus/69536-w7j-f-disassembly-instructions-4.html

 

Ah thanks guys, I thought maybe since it uses the same electricity to do a higher cpu speed it would use less electricity overall compared to a weaker processor in doing the same task.

I'll try blowing it out, I've heard a w7j is supposed to run hot but it kinda sucks anyways. Thanks again.

 

can you name the specific cpus you are referring to.

 

I have a w7j, it is currently using a T5500 at around 70 degrees and I am looking to get a T7600.

http://ark.intel.com/Product.aspx?id=27253
http://ark.intel.com/Product.aspx?id=27257
http://www.cpubenchmark.net/cpu_list.php

 

Thats a good upgrade if it doesn't cost over $75. You can buy the t7200 on eBay for $70-75. The temps will stay the same, since they have the same TDP(34watt). You can lower temps by cleaning out the fan/heat sink and using good thermal paste like Antec Formula 5.

 

Undervolt your CPU. You'll notice a big drop in temperature.

See this thread: http://forum.notebookreview.com/hardware-components-aftermarket-upgrades/235824-undervolting-guide.html

Some results you should expect once you're done undervolting your CPU:

T2370 upgraded to T7500. Starts cold at 38 C, runs at 49-52 C on idle undervolted (fan's at lowest speed), runs at 62-65 C when not undervolted .

P8400 upgraded to P9600. Starts cold at 27 C, runs at 40-45 C on idle undervolted (fan's low, sometimes off), runs at 50-54 C when not undervolted.

Check CPU usage on idle. It should be 0 to 1% depending on your background services.

 

So since the heat generated is the same would that mean that a faster processor with the same TDP always has an increase in heat in order to generate those faster speeds? Is it always like a trade off?

 

Nope. It won't generate anymore heat.

 

No, a faster CPU of the same generation will always consume more power than a slower one, despite having the same TDP. How much more heat will depend on the cooling capacity of the notebook.

 

so running at max processor speed on each processor they generate the same heat, wouldnt the faster one process more data? Or running at max would the faster one generate more heat in order to process more data?

 

Yes both processor will generate the same amount of heat. I doesn't go by: higher speed=more heat. It goes by goes by: generation, nano meter size, and TDP.

The processors for your laptop are both 34 watt TDP's. So same heat.

 

Again, TDP doesn't correlate to heat output or power consumption... A faster CPU within the same generation will generate more heat since it will consume more power. The T5500 will definitely consume less power than the T7600 and hence produce less heat (assuming if you do the switch you apply the exact same thermal compound in the exact same manner).

 

As above.

Thermal design power - Wikipedia, the free encyclopedia
CPU power dissipation - Wikipedia, the free encyclopedia

The figures are often somewhat fudged. Even if 1.8ghz and 2.4ghz processors(random numbers) shared the same TDP rating, there would be considerable differences in heat output, assuming both ran at the same voltages.

I'd really like to see an industry standard for measuring this. Something along the lines of:-

CPU at rated maximum speed, and voltage.
Heatsink of known size, with fan of known size running at typical speed in a simulated laptop case.
Controlled temperature and humidity.
Processor temperature after 30 minutes at full load on each core.

As things stand, the 65nm Celeron processors, running at 1.86ghz and 2.4ghz, both share the same TDP - whilst using the same voltages. 2.6ghz T7800 and 2.2ghz T7500 also share the same TDP - despite being the same C2D core, stepping and CPUID string. The two will not output the same amount of heat. At a given voltage, using the same core, more cycles makes for more heat.

 

All you guys have pointed out good points. And I would say correct in the limited scope of what you all are saying.

From my limited OC experience yes increasing clock will slightly increase heat. Slightly. Increasing voltage is where the real increase happens. Voltage is increased to increase stability. With notebooks this is where we are limited so only cycles which as I said I agree with but slightly. Remember a faster CPU completes a given task faster so it drops to a lower activity ergo less heat sooner than a lower clock CPU that takes longer to complete the task and get in a lower state.

TDP is not an absolute. Daytona linked some useful links. I have read them so many times in the past did not bother reading again. But TDP has no equal standard Intel has theirs AMD theirs. But with in the MFG there is a standard. So if Intel to Intel especially with in same gen should be a damn good comparison.

TDP is not running Intel burn in test and seeing which runs hotter. Yea the faster might run hotter but if was a real workload it would also produce more. TDP is an attempt to simulate what might be real world usage and the required cooling capabilities. CPU's do not other than a power virus "stress test" always run at 100%. If the higher clocked does stuff faster then lowers usage allowing cooling at the end of the day what needs more cooling short of a 30 minute stress test.

Like I said I think you all posted correct info and had good points. Just not sure you are considering everything.

Also I am not sure I am considering everything either.

 

I had an Asus G50VT-x5. It had the base P7450 Processor. It idled at 37C. I upgraded to a P8700 and aplied Antec Formula 5 to it. The Temps dropped to 29C idle. Thats a huge difference. I bet if I had the same thermal paste as the original, the temps would have been the same. The P7450 is 2.26GHz the P8700 is 2.53GHz.

 

So due to finishing the processing faster at around the same wattage it is able to cool off for longer periods? So the upgrade would cool it down you think?

 

If when you replace the processor you use good thermal paste. Like Antec Formula 5 or Arctic Silver. And clean out fan/heatsink, then yes the temps should drop.

 

Apart from those of us doing SETI, or climateprediction, or all the other distributed computing projects that use idle CPU time to do some useful work...

 

Call Of Duty: Modern Warfare 2 made my P8700 run 98-100% the whole time.

 

And the upgrade ran at what? With out that, that info is useless to me for comment.

Edit: And what was your temperature differential?

 

It reached a max of 79C. 21C below max. What I meant was there are several programs that make your processor run at 100%. I didn't have MW2 until after I upgraded, so I don't know the temps of the old processor. But they where both 25 Watt TDP. I played CoD 1 on the old processor, and I think it when to the high 60Cs. But i'm not sure.

 

So the differential is 79c or 21c. I am asking a very serious question. Not sure what you are saying or your point?

 

You where saying that processors rarely run at 100%. But several games make the processor run at 100%.79 was the max at 100% on MW2. THe max temp of that processor is 105C, so its actually 26C below max. I might be missing something. Sorry

 

You only fail to set up the experiment. But cool my friend.