Overclocking 940XM

Soloist, it's best to keep TDC (A) somewhere within about 85% of TDP. Your gap is awfully large.

 

I don't think they did anything except raise the default multipliers and increase the TDP to 62.

 

Soloist: Thanks for posting those logs. The way your average multiplier slowly drops during a longer wPrime run is interesting. ThrottleStop 2.89 has a new setting in the Options... window called Force TDP/TDC. I don't think the M17x can benefit from this setting but on your M15x, could you try dropping that down to a low number like 16 and do another 1024M wPrime run and see what happens. Dropping this to a low number forces the TDP/TDC settings to be updated in the processor more often in case they are being reset like the M11x R2 likes to do. By default on the XM CPUs, I think this Force option is set to 1000. If lowering this setting doesn't make any difference to the multiplier fading away during a run then my guess would be that as your CPU heats up, its power consumption slowly increases. As power consumption of your CPU tries to go up, you hit the turbo TDP limit and turbo boost is reduced to keep your power consumption in check.

Some CPUs have more current leakage compared to others so they consume more power and end up running hotter. On most CPUs this is not a major problem but on the Core i with turbo boost tied directly to power consumption, it's possible that this is reducing your turbo boost multiplier when you try to maintain full power for longer periods of time. I'm not sure what software accurately tracks real time CPU power consumption (HWiNFO32 ?) but maybe monitoring CPU power consumption might give you a clue here to what's going on.

A desktop Core i7 quad core CPU uses 130/110 for TDP/TDC so you're not likely going to hurt the CPU by trying to go higher but going too high could hurt the power delivery circuits on the motherboard if they can't keep up or it could cause your power adapter to reset if you are trying to draw too much power out of it.

At default settings this CPU likely runs 100% correctly so Dell probably won't be too eager to do anything about this problem you are having since it is running within the Intel design spec.

This might also explain why some laptops trigger chipset clock modulation a lot easier than others do if power consumption varies from one CPU to the next.

Edit: I asked my friend rge over on XtremeSystems about his experiences with power consumption vs heat when testing his Core i7-940 desktop chip. He's very knowledgeable and has done a lot of testing over the years for me so I thought I'd share. I think this helps explain the reduced multiplier when the core temperature goes up.

 

^ This is what happens. The force TDP/TDC did not change anything. Your story explains why I would get much faster results in wPrime when I just started my laptop up. Since the CPU isn't hot yet, it ran 25x across the board during 32mb test np at all, and the 1024mb test it would take about 15 seconds before it dropped. The runs after that showed instant dropping, as the CPU was already warmed up.

It appears my limits are about 105W with 85A, at that point the temps would hit 98-99 and multiplier be 24.5 avarage. I've settled with 95/80 as after many minutes the tems would not go above 95-96. Multiplier at those temps and TDP/C would avarage 23-flat.

So, I guess overclocking this beast would mean breaking out the liquid nitrogen? ^_^ I still have one question though; would using throttlestop (at the settings I have now) have any effect on performance in games? Seeing as many games don't fully utilize quad cores and I have never seen it hit full load in a game. In the end my only goal was getting some more FPS out of my games, and if this only improves benchmarks and programs like Sony Vegas, then doing this is pointless.

Of course, seeing as you guys cleared many things up and have been very helpfull (thanks!), the least I could do is keep doing tests and logs that you guys may want to see for further investigation.

 

At least now you have a better idea of what is limiting you and causing the turbo throttling you are seeing. Games don't fully load a CPU the way wPrime does but you can also adjust wPrime when testing so it is only loading 4 or 2 threads. This would simulate a load more typical of the average game. Your CPU won't get as hot and power consumption will be decreased so you won't be hitting the TDP wall like you are now. With less load you can work on increasing the 1, 2 and 3 core turbo multiplier limits in ThrottleStop and do some more testing to see if the multiplier is able to hold without fading away.

There are going to be some games that are not CPU limited where using ThrottleStop isn't going to make a lot of difference. Try adjusting the light load turbo multipliers higher and then go play some games and see if running ThrottleStop is useful to you or not.

Has anyone noticed this issue with the Extreme ES processors that most of you are using? Hard to believe that the ES CPUs are better than the OEM at maintaining full turbo boost when fully loaded.

 

have you replaced the paste on the CPU yet with something better. You can get a 7-10 degree drop from a good paste job and as high as 15 degree drop. That would make sure you are not getting close to thermal throttling what you seems to be experiencing when you push the CPU hard for long time.

 

I don't believe that it is traditional thermal throttling that is reducing his multiplier. Thermal throttling isn't designed to start until the core temperature hits 100C. When a CPU gets hot, its internal resistance increases so power consumption goes up. On the Core i7 XM, that puts you over the turbo TDP/TDC limit and the multiplier will start to rapidly cycle so your average multiplier drops lower and lower.

Re-doing the paste and doing anything else you can think of to reduce the core temperature should help reduce this turbo throttling problem. During normal gaming, it might not be a problem at all.

You can check the thermal status of your CPU using RealTemp. If it shows OK in the Thermal Status area, that means thermal throttling has not happened once since you powered on your laptop.

Download Real Temp 3.60 | techPowerUp

I believe HWiNFO32 also reports this info from the CPU but I haven't tested that program yet.

 

Yep, HWiNFO32 reports the Thermal Monitor status as "OK" or "Throttling" as well.

 

Thanks Unclewebb, I'll look into RealTemp. I think what I will do is push the multipliers of the first two cores up to 28/29 and leave the last two on 25/26. I've ran some 3DMark06 runs and it only used the first core. When it got to the CPU test it crashed with I had a 30x multiplier set. The temps wouldn't even get close to 80.

So yeah, now I am running with 29-28-25-25 so that way it can run nicely in games when not under full load.

Little off topic: when you run a simple 3DMark06 benchmark, are there any settings I need to change to be equal to how everyone else benches? I would guess telling it I have 8 threads / 4 cores, but I have an unregistered basic version, so I don't think I even can. It's vesion 1.2.0

 

Thanks Mumak for confirming that.

To test your 1 core multiplier, just run a single thread of Prime95 and that will maximize your average multiplier. If you can run this without any errors with the 1 core multi set at 29 then you will probably be OK during gaming.

You won't get the full 29 multiplier because other tasks will occasionally wake up a second, third and fourth core but the average should probably be over 28 in a test like this. You can also try some 2 and 3 threads of Prime95 for light load stability testing.

I think most users are running the unregistered version of 3DMark06 at its default settings.

 

I know that this is a really old thread, but I'm also interested in OCing my 940XM. The problem is, using all of the settings seen previously in this thread (95W/75A and 29 28 28 27), I can't get below 12.4 seconds in wprime 32m. What exactly am I doing wrong? Also, what are the chances that this will harm my computer and what are the stock settings for the CPU in case I want to change them back.

Lastly, when Throttlestop is only monitoring, does it still maintain the settings I implemented, or does the computer run on stock settings when I'm not using Throttlestop? I know that I'm incredibly nooby, but I figure I have to start somewhere Any help with this would be greatly appreciated

 

Your CPU is throttling, make sure to enable Chipset Clock Modulation in Throttlestop and set it to 100%.

You can't really damage your system with this, but the heat may become a problem. Worst thing that can happen is that the CPU reaches 100°C which results in an immediate shutdown to prevent harm.

Throttlestop shouldn't change anything when in monitoring mode. You have to activate TS before the CPU values get manipulated.



I recommend starting with some lower settings and then increasing them as soon as you see that it works fine. Every chip performs a bit differently when set to such extreme values.

When the TDP/TDC values are to low the multiplier won't go to the value you set it. In this case you have to adjust these numbers.
Best thing is to monitor the values on TS while doing a wPrime 32m on 8 threads.


Last but not least make sure there's enough room between your system and the desk, this improves the airflow.

Let me know about your results.

 

"Warning: Changes happen immediately"

​

Both multiplier settings and power settings (TDP/TDC) take effect immediately and remain in effect until (a) reset to the default (or other) values or (b) reboot. Whether Throttlestop is enabled or only monitoring makes no difference. Closing TS makes no difference. So, be careful; when you set and accept a value, it goes live immediately and stays there.

Be sure you are using wPrime w1.55 to have results comparable to those seen here. The newer version (2.06?) returns slower values, but has flaws that make it unacceptable to the serious benchmarkers.

Start at reasonable levels and go slowly. Default multipliers for the 940M are 25-24-18-18 (Intel/M15x) or 25-25-25-25 (M17x-R2 o/c with most recent BIOS), both with 62W TDP/62A TDC. The M17x cannot operate at 25x across the board with 62W/62A power settings, so it is really window dressing. It takes about 90W/80A to support the 940XM running at 25x flat. At those speeds, wPrime will complete the 32M run in about 7.6-7.7 seconds. At default power settings (62/62), it takes about 9.8-9.85 seconds (an effective 19.4x rate). Watch temps, don't go any higher with either multipliers or power - particularly power, power is where the danger lies - until you learn more about TS and its effects. You can fry motherboards and PSU's done wrong.

 

Oh, you're right Rev, my fault... The only effect of "Turn on" are the other options getting active.

 

What's your opinion about rev. Q4AP (940xm).
Is it safe to buy it?

 

That is the exact same stepping as the retail processors so there shouldn't be any significant difference, other than the price.

 

I've got 920xm oem (dell m6500)
Tried to oc with TS and now looking for the best settings.

Could you please tell me what should be ticked?

Code:

clock mod
chipset clock mod
set multi

What these functions mean?

Also I don't know if TS has to be turn ON or OFF?

There is my first results
@22x133
TDP/TC 85/70

3dmark06 test.

logs from trottlestop

 

The functions are probably best described in the ThrottleStop Guide, available here: Throttlestop Guide. Dell uses Chipset Modulation as the prime method of CPU temperature and power control in most models. I have no experience with the Precision line, but assume the same applies there. Dell and Intel also use multiplier modulation (i.e., a forced reduction of the effective multiplier level) as a CPU protection mechanism to prevent severe overheating. The Intel fail safe scheme kicks in at 100C for the 920/940xm's, but Dell reduces that in most models. So, which options should be selected (checked) depends on the model and what you're doing. The selected options only take effect when ThrottleStop is turned on, so most people select all three, and then use the On/Off option to control whether those that apply are active. Note that the Set Multiplier option should be set to Turbo (increase the multi number as high as it will go, then one more click puts you into Turbo). Whether TS needs to be On or Off depends on the settings and the use. As long as the computer can accommodate the multiplier and power (TDP) settings used, there is no need to turn TS on; it is needed when the settings bump up against the temperature and power limits, mostly temperature, set by Dell in the BIOS. Because that is not always apparent, many users just leave it on so that it is there if needed.

It's hard to tell much from the TS log of 3dMark06 without knowing what part of the test was being monitored. You would expect the multiplier (FID) to be at 22x instead of 20x +/-, but you can't tell why it's not. There is no Chipset Modulation or Clock Modulation. If the TDP/TDC were active at 85W/70A, there should have been no power limitation holding it back. The fact that it starts at 20x +/- with temps in the high 60C's suggests temps were not a factor, though you see a slight bleed-off as temps increase. My guess is that the multipliers got reset, but the TDP did not, but I'm not sure.

We need a more controlled test bed. Download wPrime v1.55 here: Downloads | wPrime Multithreaded Benchmark (Note that there are several versions; scroll down to v1.55, the one recognized by benchers and used here.)

When you start wPrime, click on Advanced Settings and put "8" in the Thread Count box. Then click save to return to the main screen. Start Throttlestop and set the multipliers to 24-24-24-24 and the TDP to 85, TDC to 72. For this run, check all options as shown below and turn TS On. (At the settings used, you should not need the anti-throttling devices, but use it anyway for now.) Then run the 32M test in wPrime. You should get a time score of about 8 seconds. Try it and tell us how it works.

 

Thank you for a detailed explanation.

I'll do more test and show the results.

BTW There is a problem with fan's work. They turn the fastest gear very lately so the temps are high during OC as you can see. I can't find any software which I could manually set the fans to run at full speed.
Using 15324 trick I can't exit to Windows (desktop). When I press ENTER the screen blinks on and off with different colours.

 

There's a great little utility named HWiNFO64 (..32 for the 32-bit version) that provides fairly sophisticated fan control in the M15x available here: HWiNFO, HWiNFO32 & HWiNFO64 - Hardware Information and Analysis Tools. Its functions are not always apparent. To invoke fan control, go to the Sensor tab or screen and click on the small fan icon (adjacent to Logging Start). You can set the CPU and GPU fans (separately) manually at any level or define a temperature driven fan scheme to suit your purposes. It works beautifully in the M15x, but depends in part on the Phoenix BIOS used here if I understand it correctly. The author (Mumak) actively supports the program and is willing to try most anything to accommodate user needs. HWiNFO32/64 Discussion. Try it; if it works in the Precision, it will take care of that issue. Several members here know much more about the HW64 fan control app than I do and are happy to help.

 

HWiNFO/32/64 Forum • View topic - Fan control icon Dell M6500

 

Sorry, Aldam. I wonder why.

 

I would suggest doing some wPrime testing so you can get a better feel for what ThrottleStop settings do what when overclocking. You can make adjustments as the benchmark is running and watch what happens to the average multiplier that ThrottleStop reports.

http://www.wprime.net/Download/

The log file looked like the CPU was hitting the TDP or TDC limit. That should not have been happening if you selected the higher limits you posted and pressed OK when you exited the TDP window. Either that or your OEM CPU is consuming significantly more power compared to the ES CPUs that most users in this forum have been using.

Some wPrime 32M benchmarks will give you an idea of how to improve your CPU speed and some longer 1024M benchmarks will let you know how you are going to be limited by heat. Doing what you can to improve CPU cooling is the best way to maximize performance from these hot but fast 920XM and 940XM CPUs.

The Revelator: I'm not sure if you need it but I finally added per profile Turbo Ratio Limit settings for the XM CPUs. Beta testing looks good so far.

ThrottleStop 3.20 beta 2
http://www.mediafire.com/?4ug155h06073w5v

 

So I set the multipliers to 24-24-24-24 and the TDP to 85, TDC to 72.
Eventually also set fans at ~4k rpm (I did it using 15324 trick with hidden service menu).

First run 32m benchmark. Result ~12sec. Temperature of cpu ~70C
next run 1024m for a minute. Temperature max ~81C

 

Something is holding back your CPU, maybe it can't draw enough power to go reach the multi you set.
Is your BIOS up to date?

 

Something badly out of whack here.

First, be sure that you are using v1.55 of wPrime for testing purposes. There are newer versions; you don't want any of those.

Second, a 920xm at stock settings (24-23-17-17, meaning 17x flat in wPrime) will complete the 32M run in ~11.2 seconds, so at 12 seconds it is not even operating at stock levels. Clearly your 920xm is not getting set to 24x even though it appears so. Be sure that after setting the multipliers to 24x, you click Apply and then OK. It's easy to go straight to OK, which leaves the multipliers at the prior (saved) level. Same applies to TDP/TDC. First Apply, then OK.

If these do not remedy the situation, I would recommend downloading the Intel Processor Diagnostic Tool to check the health of the CPU.

​

 

I just quickly checked the M6500 specs, this beast comes with a 210W PSU, you shouldn't have any power issues with this system, it's designed for high power draws.

You are using your stock PSU, right?

 

Thank you for your help.

Today I'm going to check the health of the CPU with Intel Processor Diagnostic Tool.

My PSU is even 240W so it can't be a problem.
I'm using v1.55 of wPrime. My cpu is i7-920XM SLBLW so it should be oem version. When I set ratio and TDP/TDC I first Applied, then OK. I also have the newest bios.

When I run 1024m test the TS shows x22 multi, but when I run 32M TS shows different multi from 17 in the begging to ~20 in the end. The ratio fluctuates the whole time.

BTW in wPrime I click on Advanced Settings and put "8" in the Thread Count box and then save.

 

In the Turbo Ratio Limits window set a 4 core multiplier of 22. Run wPrime 32M and then keep increasing the TDP / TDC limits until your CPU can maintain the full 22 multiplier for the duration of the test. Run the test a few times until you know what it is going to take. If heat isn't too crazy then try doing the same thing with the 23 multiplier and the 24 multiplier. You might find that you have to increase the TDP / TDC settings far beyond what most other users need. That would be a sign that either you have a crappy, power hungry CPU or maybe Dell deliberately uses more core voltage on these motherboards for improved full load stability.

Is there any option in the bios that can be set to adjust the CPU voltage on the M6500? My best guess is that for whatever reason, the power consumption of your CPU is higher than the XM ES (Engineering Sample) processors that many other enthusiasts in this thread have been using. Higher power consumption is what triggers the turbo throttling that you are seeing.

If you run some tests then you can compare to how much TDP / TDC others need to run wPrime 32M with a steady 22 multiplier.

 

FWIW, I'm using the OEM version of the 920xm.

 

look at this

TDP/TC 100/85!
Score 11,3 sec !
Temps below 80C

EDIT
I have problem with Net framework 3.5. Can't install it. Could it be the problem with TS and wPrime?

Next 24-24-24-24

TDP/TC 100/87
Temps imo low.

 

Now you're cooking. That M6500 is a power eater!

 

But temps are low imo. I really don't know what is going on.
My score never goes below 11 sec .

With 90/77 it can maintain only ~21

 

ThrottleStop or wPrime shouldn't cause a problem with installing .NET framework.

I was going to tell you to keep cranking up the TDP / TDC but I think The Revelator knows what can happen if you do that. Motherboards can only take so much of this free performance before they complain.

You would need to run your CPU in The Revelator's board to try and figure out whether you simply have a power hungry CPU or perhaps the motherboard bios is being generous with the core voltage and creating extra power consumption and heat. When running fully loaded, a small increase in core voltage can make a big difference to power consumption and when turbo throttling begins.

 

Is your BIOS up to date? And the chipset driver as well?

 

This is where my processor knowledge hits the wall. How can a 920xm be running at 23x or 24x across the board, but take 11+ seconds to complete wPrime's 32M test? Could something about the M6500 be causing wPrime to misread the timing?

 

When the baseclock throttles... there are sometimes protective functions in the BIOS which can slow down the bus speed in certain circumstances.

@aldam: Can you try to monitor you baseclock speed?

 

Bios is the newest one.
I run 32M @stock 920 and got 12,66-13,1 sec.

I dont know how to monitor it. I looked at cpu-z and the multi was 19- 22

 

I couldn't install net framework before TS and Wprime was installed.

I run also 3dmark06.
stock 920 - 14700 points
@22 TDP/TC 85/70 - 16200 points

So there was a gain in performance.

 

There's a "bus speed" entry in the first tab of CPU-Z, and you can also see it on top of Throttlestop, where it displays the CPU frequency, it displays the multi x bus speed = frequency. For example 26 x 133.0 MHz = 3457.9MHz

The bus speed should stay at 133 MHz all the time. I don't know how often its value gets refreshed.

 

Bus speed is 132,7 - 133 during the test

 

Something doesn't sum up... can you try it with a longer test and see whether the bus speed drops after a minute or so? Make a 1024M test. You don't have to finish the test, only keep an eye on the bus speed in CPU-Z for about a minute.

 

stock and OCed bus was 132,7

 

aldam: What does the Task Manager show for CPU Usage % when your laptop is idle. Your CPU doesn't seem to be using the 2 common throttling methods that other Dells use. You can try checking the Clock Modulation and Chipset Clock Mod boxes and setting those to 100.0% if you haven't tried that yet.

 

It is ~3%. It is taken by CIS.
I've already checked Clock Modulation and Chipset Clock Mod boxes.

I'm going to reinstall my os. Now I use win7 which has been cut with vlite.

 

I played BFBC2.
24-24-24-24
86/73

Code:

08/06/11 20:17:27  24.00   18.9  100.0  100.0       0   71
08/06/11 20:17:28  24.00   19.6  100.0  100.0       0   71
08/06/11 20:17:29  23.42   21.4  100.0  100.0       0   70
08/06/11 20:17:30  23.52   20.0  100.0  100.0       0   71
08/06/11 20:17:31  23.72   22.2  100.0  100.0       0   70
08/06/11 20:17:32  23.18   20.8  100.0  100.0       0   70
08/06/11 20:17:34  23.94   20.7  100.0  100.0       0   71
08/06/11 20:17:35  23.18   20.9  100.0  100.0       0   70
08/06/11 20:17:36  23.70   20.3  100.0  100.0       0   71
08/06/11 20:17:37  24.00   20.2  100.0  100.0       0   70
08/06/11 20:17:38  24.00   19.1  100.0  100.0       0   70
08/06/11 20:17:39  24.00   20.9  100.0  100.0       0   71
08/06/11 20:17:40  24.00   18.7  100.0  100.0       0   70
08/06/11 20:17:41  23.83   21.1  100.0  100.0       0   70
08/06/11 20:17:42  23.08   21.1  100.0  100.0       0   70
08/06/11 20:17:43  23.22   22.3  100.0  100.0       0   73
08/06/11 20:17:44  24.00   20.6  100.0  100.0       0   70
08/06/11 20:17:45  24.00   23.5  100.0  100.0       0   73
08/06/11 20:17:46  24.00   29.7  100.0  100.0       0   73
08/06/11 20:17:47  24.00   31.7  100.0  100.0       0   74
08/06/11 20:17:48  24.00   30.4  100.0  100.0       0   74
08/06/11 20:17:49  24.00   30.1  100.0  100.0       0   74
08/06/11 20:17:50  24.00   31.8  100.0  100.0       0   78
08/06/11 20:17:51  24.00   30.3  100.0  100.0       0   74
08/06/11 20:17:52  24.00   29.6  100.0  100.0       0   74
08/06/11 20:17:53  24.00   29.6  100.0  100.0       0   74
08/06/11 20:17:54  23.18   23.9  100.0  100.0       0   72
08/06/11 20:17:55  24.00   25.2  100.0  100.0       0   75
08/06/11 20:17:56  23.87   16.5  100.0  100.0       0   75
08/06/11 20:17:57  24.00   19.9  100.0  100.0       0   73
08/06/11 20:17:58  24.00   20.0  100.0  100.0       0   74
08/06/11 20:17:59  24.00   21.3  100.0  100.0       0   71
08/06/11 20:18:00  24.00   21.5  100.0  100.0       0   72
08/06/11 20:18:01  24.00   19.5  100.0  100.0       0   74
08/06/11 20:18:02  24.00   29.4  100.0  100.0       0   75
08/06/11 20:18:03  24.00   25.7  100.0  100.0       0   73
08/06/11 20:18:04  24.00   20.7  100.0  100.0       0   74
08/06/11 20:18:05  24.00   23.8  100.0  100.0       0   74
08/06/11 20:18:06  24.00   21.8  100.0  100.0       0   74
08/06/11 20:18:07  24.00   23.6  100.0  100.0       0   73
08/06/11 20:18:08  24.00   21.1  100.0  100.0       0   74
08/06/11 20:18:09  24.00   21.3  100.0  100.0       0   76
08/06/11 20:18:10  24.00   23.4  100.0  100.0       0   74
08/06/11 20:18:11  24.00   21.2  100.0  100.0       0   74
08/06/11 20:18:12  24.00   23.1  100.0  100.0       0   73
08/06/11 20:18:13  24.00   21.7  100.0  100.0       0   74
08/06/11 20:18:14  24.00   23.0  100.0  100.0       0   74
08/06/11 20:18:15  24.00   22.3  100.0  100.0       0   74
08/06/11 20:18:16  24.00   23.2  100.0  100.0       0   77
08/06/11 20:18:17  24.00   22.8  100.0  100.0       0   74
08/06/11 20:18:18  24.00   25.6  100.0  100.0       0   74
08/06/11 20:18:19  24.00   25.8  100.0  100.0       0   74
08/06/11 20:18:20  24.00   23.3  100.0  100.0       0   74
08/06/11 20:18:21  24.00   24.1  100.0  100.0       0   75
08/06/11 20:18:22  24.00   21.6  100.0  100.0       0   74
08/06/11 20:18:23  24.00   22.1  100.0  100.0       0   74
08/06/11 20:18:24  24.00   27.2  100.0  100.0       0   75
08/06/11 20:18:25  24.00   29.1  100.0  100.0       0   74
08/06/11 20:18:26  24.00   34.1  100.0  100.0       0   75
08/06/11 20:18:27  24.00   21.0  100.0  100.0       0   74
08/06/11 20:18:28  24.00   22.3  100.0  100.0       0   76
08/06/11 20:18:29  24.00   23.8  100.0  100.0       0   73
08/06/11 20:18:30  24.00   21.4  100.0  100.0       0   75
08/06/11 20:18:31  24.00   23.8  100.0  100.0       0   73
08/06/11 20:18:32  24.00   21.6  100.0  100.0       0   74
08/06/11 20:18:33  24.00   22.5  100.0  100.0       0   74
08/06/11 20:18:34  24.00   21.5  100.0  100.0       0   75
08/06/11 20:18:35  24.00   22.1  100.0  100.0       0   75
08/06/11 20:18:36  24.00   22.3  100.0  100.0       0   74
08/06/11 20:18:37  24.00   21.6  100.0  100.0       0   74
08/06/11 20:18:39  24.00   31.8  100.0  100.0       0   75
08/06/11 20:18:39  23.98   25.2  100.0  100.0       0   75
08/06/11 20:18:40  24.00   22.5  100.0  100.0       0   73
08/06/11 20:18:42  24.00   20.4  100.0  100.0       0   74
08/06/11 20:18:43  24.00   21.3  100.0  100.0       0   71
08/06/11 20:18:44  24.00   18.0  100.0  100.0       0   74
08/06/11 20:18:45  24.00   17.6  100.0  100.0       0   72
08/06/11 20:18:46  24.00   22.9  100.0  100.0       0   73
08/06/11 20:18:47  24.00   26.5  100.0  100.0       0   72
08/06/11 20:18:48  24.00   46.4  100.0  100.0       0   77
08/06/11 20:18:49  24.00   46.5  100.0  100.0       0   78
08/06/11 20:18:50  24.00   47.6  100.0  100.0       0   78
08/06/11 20:18:51  24.00   45.3  100.0  100.0       0   78
08/06/11 20:18:52  24.00   45.7  100.0  100.0       0   78
08/06/11 20:18:53  24.00   44.8  100.0  100.0       0   80
08/06/11 20:18:54  24.00   47.5  100.0  100.0       0   78
08/06/11 20:18:55  24.00   49.9  100.0  100.0       0   74
08/06/11 20:18:56  24.00   47.3  100.0  100.0       0   81
08/06/11 20:18:57  24.00   50.8  100.0  100.0       0   79
08/06/11 20:18:58  24.00   41.1  100.0  100.0       0   77
08/06/11 20:18:59  24.00   24.6  100.0  100.0       0   79
08/06/11 20:19:00  24.00   24.2  100.0  100.0       0   77
08/06/11 20:19:01  24.00   23.1  100.0  100.0       0   77
08/06/11 20:19:02  24.00   25.1  100.0  100.0       0   77
08/06/11 20:19:03  24.00   22.5  100.0  100.0       0   77
08/06/11 20:19:04  24.00   25.0  100.0  100.0       0   77
08/06/11 20:19:05  24.00   22.4  100.0  100.0       0   77
08/06/11 20:19:06  24.00   23.9  100.0  100.0       0   79
08/06/11 20:19:07  24.00   24.1  100.0  100.0       0   76
08/06/11 20:19:08  24.00   22.5  100.0  100.0       0   77
08/06/11 20:19:09  24.00   25.1  100.0  100.0       0   76
08/06/11 20:19:10  24.00   22.5  100.0  100.0       0   76
08/06/11 20:19:11  24.00   23.2  100.0  100.0       0   76
08/06/11 20:19:12  24.00   22.4  100.0  100.0       0   76
08/06/11 20:19:13  24.00   24.7  100.0  100.0       0   76
08/06/11 20:19:14  24.00   23.4  100.0  100.0       0   76
08/06/11 20:19:15  24.00   22.8  100.0  100.0       0   72
08/06/11 20:19:16  24.00   22.8  100.0  100.0       0   76
08/06/11 20:19:17  24.00   22.3  100.0  100.0       0   76
08/06/11 20:19:18  24.00   25.5  100.0  100.0       0   75
08/06/11 20:19:19  24.00   22.7  100.0  100.0       0   76
08/06/11 20:19:20  24.00   22.4  100.0  100.0       0   77
08/06/11 20:19:21  24.00   22.9  100.0  100.0       0   76
08/06/11 20:19:22  24.00   21.8  100.0  100.0       0   74
08/06/11 20:19:23  24.00   23.9  100.0  100.0       0   75
08/06/11 20:19:24  24.00   32.1  100.0  100.0       0   75
08/06/11 20:19:25  24.00   31.3  100.0  100.0       0   75
08/06/11 20:19:26  24.00   28.8  100.0  100.0       0   75
08/06/11 20:19:27  24.00   31.0  100.0  100.0       0   76
08/06/11 20:19:28  24.00   32.3  100.0  100.0       0   76
08/06/11 20:19:29  24.00   32.8  100.0  100.0       0   76
08/06/11 20:19:30  24.00   32.7  100.0  100.0       0   75
08/06/11 20:19:31  24.00   29.0  100.0  100.0       0   75
08/06/11 20:19:32  24.00   31.2  100.0  100.0       0   76
08/06/11 20:19:33  24.00   27.7  100.0  100.0       0   75
08/06/11 20:19:34  24.00   32.0  100.0  100.0       0   76
08/06/11 20:19:35  24.00   29.5  100.0  100.0       0   76
08/06/11 20:19:36  24.00   32.0  100.0  100.0       0   79
08/06/11 20:19:37  24.00   31.1  100.0  100.0       0   76
08/06/11 20:19:38  24.00   28.5  100.0  100.0       0   76
08/06/11 20:19:39  24.00   30.0  100.0  100.0       0   76
08/06/11 20:19:40  24.00   27.8  100.0  100.0       0   76
08/06/11 20:19:41  24.00   30.9  100.0  100.0       0   76
08/06/11 20:19:42  24.00   29.7  100.0  100.0       0   76
08/06/11 20:19:43  24.00   29.7  100.0  100.0       0   76
08/06/11 20:19:44  24.00   31.1  100.0  100.0       0   76
08/06/11 20:19:45  24.00   27.9  100.0  100.0       0   75
08/06/11 20:19:46  24.00   29.9  100.0  100.0       0   76
08/06/11 20:19:47  24.00   28.1  100.0  100.0       0   75
08/06/11 20:19:48  24.00   31.6  100.0  100.0       0   76
08/06/11 20:19:49  24.00   30.2  100.0  100.0       0   76
08/06/11 20:19:50  24.00   31.1  100.0  100.0       0   77
08/06/11 20:19:51  24.00   28.7  100.0  100.0       0   75
08/06/11 20:19:52  24.00   28.5  100.0  100.0       0   75
08/06/11 20:19:54  24.00   29.7  100.0  100.0       0   75
08/06/11 20:19:55  24.00   27.9  100.0  100.0       0   75
08/06/11 20:19:56  24.00   30.4  100.0  100.0       0   75
08/06/11 20:19:57  24.00   29.0  100.0  100.0       0   74

 

Real world testing looks good. Most games do not fully load a CPU the way wPrime does. There are some games like GTA4 that are very multi threaded and will significantly load your CPU but even playing that game, your multiplier will probably still be able to maintain the full 24.00.

When posting log files instead of using the QUOTE html tag in brackets, try replacing that with CODE and /CODE instead. It maintains the formatting and keeps the mods happy because it adds a scroll bar so it takes up less space.

 

Although I have an M17x-R2 this thread is relevant to my interests. I have an ES 920xm (though it is physically labeled as OEM :\ ) which I tested at 90/80 and 85/73 and 24x multi with 140 bus speed. I am getting around 7.6 secs for 32m test and for the 1024m test I don't even finish because once my temp hits 91c my notebook crashes, I get a blue screen. I have chip mod enabled. What are your guys highest temps while running the 1024m? Is my chip a piece of crap or do I need to re-paste? I used MX-2. Thanks

 

I had a similar experience. I ran the 1024M test at 25x (90W/78A)(32M @ 7.61 seconds) to approximate your 24x@140 run with TS active. With TS preventing throttling at lower temps, the CPU stayed on full power (25x) until the upper 90's, when there was limited multiplier throttling (~23.5), but temps continued to rise to 100C, at which point it crashed with a BSOD, exactly as Intel intended. Fans were running full blast. It's definitely time for me to repaste, but given the heat generated by these 920xm's under full load at high frequencies I'm not sure you can ever complete the 1024M test at or above 3.33Ghz without throttling in the absence of auxiliary cooling.

 

I see thanks for the explanation. So it seems it would be fine for gaming though and that my experience is rather normal.