I had some time to do some various testing of my MSI GT70. I repasted my CPU and my GPU because I wanted to be shure it runs as cool as possible. I highly recommend doing your own paste job or let the resellers do it for you since MSI don`t know/doesn`t care, what they are doing...
See how my 680M looked like here before repasting (MSI factory that did this): http://forum.notebookreview.com/gaming-software-graphics-cards/697157-too-much-paste.html#post8952607
Not shure if this is interesting to other people, but atleast it gives you some idea how 680M run and how 3610QM run with proper paste.
All of this have been done without the turbo fan mode I have on my GT70. Usually it was on the lowest fan speed with occasionally ramping up the speed, but it only lasted for a little while before settling down again. Pretty satisfied with the "noise" of this notebook. Its a pleasant "WHOOSH" sound because the fan is so big. :thumbsup:
73C for the 680M and 60C for the CPU in average while gaming on 1080p with Ultra. And 34C while browsing the internet and watching some short movies. Pretty good temperatures![]()
3DMark Vantage:
![]()
http://www.3dmark.com/3dmv/4420494
3DMark11:
![]()
http://www.3dmark.com/3dm11/5047314
Just browsing the internet/watching movies:
![]()
3DMark 11:
![]()
![]()
Skyrim
![]()
![]()
-
-
That's great ! I'd be interested to see the gaming/benchmarking temp with the turbo fan on ! Temps are a little high for stock clocks imo even though I know the fan doesn't kick off a lot
-
What do you mean?
Skyrim temps are slightly better than the Clevo temps where they have a fan just for the GPU. And this is with lowest fan setting almost all the time and the Clevo is noisy as hell.
Mind you that this graph shows exactly what the temperature is at any given time. 99% of the people who use monitor software just watch the temperature that is there on the screen when they have the time to exit the game but never use log functions like I did. Pretty shure that this is normal if not better than normal
The 3DMark temps are higher than what you would see in games since the fan only kicks in at the very end of the tests since they are so short.
I will see if I can do some rounds with the turbo fan mode on -
My academic sense is telling me that all these graphs and numbers are meaningless because you didn't say anything about ambient temp.
-
You`re right: around 22C +/-1 in my house
-
I`d love to see other GT70 owners do some runs with their notebooks to compare temperatures with me. Maybe do some comparisons with 7970M owners, and GT60 and GT780R with 580M.
Download HWMonitor Pro (free 30 day preview before you have to pay) and use the log function and it will create these beautiful graphs for you. Just make shure you have the GPU running since it won`t log the dGPU if the HD4000 is running.
CPUID - System & hardware benchmark, monitoring, reporting -
I've got the GT70, but it's the barebone version fixed together by PWNPCs and pau1ow
MSi 1762.
Interested in my temps? -
That would be great
-
Hi,
Here is my benchmark results and some temps.
+24C temp in house, btw.
MSI GT70 0NE-276US custom body: Stock CPU/GPU, Killer 1202 Wi-fi, 24Gb RAM, HDD in optical bay AND IC Diamond 24 Carat Thermal Compound on both CPU/GPU !
So, I guess it'd be nice to provide some temps in comparison to fully stock box and see how additional paste affects temps.
3DMark11 scores:
Performance
Extreme
Temps:
3DMark11 - Performance
3DMark11 - Performance - CPU
3DMark11 - Performance - GPU
3DMark11 - Performance - Fan on
3DMark11 - Performance - Fan on - CPU
3DMark11 - Performance - Fan on - GPU
3DMark11 - Extreme
3DMark11 - Extreme - CPU
3DMark11 - Extreme - GPU
-
These are the temps I was expectingfrom the 680M @ stock clocks! Well done.
-
FYI:All scores/temps above and below are on stock OS, drivers (302.75), Firmware and VBIOS. No OC at all.
FurMark 1.10.3 scores and temps (1080p 15min Burn-in test):
3D Vantage scores and temps:
IntelBurnTest 2.54 scores and temps (Stress level - Maximum):
The same but with Fan on:
And finally, OCCT 4.3.2 Power Supply test (15 min, FullHD fullscreen)
The same but with Fan on:
-
Sorry wrong thread.
Don't have 3DMark scores yet, but my temps were 81c (GPU) after an hour of playing Far Cry 3 maxed settings 1080p. -
Thanks for the data deskdollars. Now I have something to compare against +rep
Few questions popped up in my head when I looked at your 3DMark11 temps. Your was shorter than mine. Was it without the demo? I ran demo + the benchmarks. Could you do one with demo so we can compare equal sessions?
I just finished testing Vantage and have so we can compare temps from that. Was yours with turbo fan on or off? It looks you have yours on or the fan at a bit high speed because from the graph the temperature is dead on at one temp while mine fluctuate a whole lot more. Or my fans are way more slower than yours. I had mine running at the lowest speed almost the entire session. Is that normal? What about yours?
Looks though that our CPU`s are pretty similar give or take if you look at the average but your GPU is running hotter than mine. But also here yours is a straight line but mine fluctuate...Weird
3DMark Vantage:
Cloudfire CPU:
deskdollars CPU:
Cloudfire GPU
deskdollars GPU:
-
Hi Cloudfire,
Looks like you're trying to compare wrong screenshots.
You took the latest ones from OCCT Power Supply test (which is most excessive test), please check above once again.
Will post 3Dmark Vantage test results w/o and with Fan on. BTW, initially was posted w/o fan on.
Tests with Fan on are marked additionally.
As for 3DMark11 you're right, that was without demo. -
Oh sorry about that. Didn`t see that. Updated my post with the right screenshots
Looks like our temps on GPU and CPU are identical but my CPU spike is 6C higher -
No worries, it looks like that initial temperature on GPU might also affect max temps during the test.
I'll try to start from 35C as in your test to get the comparison even more accurate. -
My max temps when gaming, OC'd are 82C so far. This was playing with 900/2200 clocks. I will check for stock clocks.
Furmark pushes my system regardless of clocks etc, eventually reaches 90s (with unlocked vbios full utilization). My original vbios would have a max of 78 clocks for furmark, but those scores were fake because nvidia GPU would be downclocking while testing. Gaming though, is very much real with 100% utilization and hours of gameplay. It helps to play with vsync to have even lower temps. -
Could anyone point me out to the link of modded VBIOS? pau1ow, can you?
-
-=$tR|k3r=- Notebook Virtuoso
Also, you may want to check out Svet's Fermi/Kepler VBIOS Tuner - [ver: 2.36f], which now supports the GTX-680M. Svet is a global moderator, and infamous modder in the MSI forums.
-
One does not simply obtain modified vbios. You must past the tests, and become one with the GTX680m. Until then, you will remain in stock clocks, forced to be ridiculed by your peers obtaining 3dmarks of 8k while you cry to your sleep with mere 6k.
Joking aside, there are many vbios, it depends on what you are looking for. Popular ones include Saltius 900/2400 with slight overvlotage. Gaming wise they seem to be fine, but at least for me, they ran too hot for my taste. -
It records a graph for each of the cores... I cannot comprehend this -.-
Oh dear god -
-=$tR|k3r=- Notebook Virtuoso
Agreed, the mods are a bit toasty, which is another reason I suggest MSI's WIN 8 version vbios. I'll post a link for this later in the owner's lounge.
-
Thanks -=$tR|k3r=- and ryzeki for pointing me out to the right direction.
U'll be surprised to know that I'm not a gamer, and 8k or 6k in 3dmark11 means almost nothing to me.
I'm not sure that I'll go with modded VBIOS but still would like to keep it for future. Yeah, I know, there will be more of them, but still
The most frustrating thing is that TDE is not properly working, at least as intended, I guess, and that's why flashing modded VBIOS along with updated firmware might help to release overclocking features.
---- Offtopic off ---
Updated screenshots are comming...
EDIT: -=$tR|k3r=-, I think you have mixed out firmware and vbios in your post. Do you suggest to flash win8 firmware and modded vbios on top? I guess Ken mentioned that he got rid of throttling on latest Win7 firmware, did he? -
-
-
I'm going to update to 310.70 drivers.
Does anyone interested in 3dmark11/Vantage scores/temps on stock firmware/vbios on 310.70? -
Yea I'm interested in any benchmarks
-
Temps during benchmarks mean almost nothing, since the load is uneven and you throttle back clocks and utilization between tests. It is best to test with actual games, with both vsync on and off to see max temps and typical vsync temps.
Some games, which produce over 60fps, will tax the system less during normal use as long as vsync is on, resulting in low temps in general. Other games will quickly raise your max temps and keep your fan at high speed, but temps I haven't seem them get over 80s even with demanding titles and 100% utilization, OC'd -
-=$tR|k3r=- Notebook Virtuoso
-
But for Win8 models I can overclock out of the box without using ThrottleStop or modded vBIOS.
So like -=$tR|k3r= said you can update BIOS, vBIOS and firmware from Win8 model however, you will need to change boot mode in BIOS from UEFI to Legacy if your BIOS is updated to Win8 models otherwise your laptop will not able to boot to Windows. -
Hello Ken, thanks for information.
Just to confirm.
Do I have to update all three things marked as for Win8 models - BIOS, Firmware and VBIOS? Or just one of them?
BIOS from http://www.msi.com/product/nb/GT70-0NE.html#?div=BIOS - E1762IMS.70T.zip (I have 276US model, so RAID, Backlight keyb and Win7 on board)
Firmware from http://www.msi.com/product/nb/GT70-0NE.html#/?div=Firmware - 1762EMS1504.zip
VBIOS from http://gentechpcforums.com/Images/MSI/MSI_GT60/680m.rom - (WIN 8 models)
Is that correct?
And one more thing, in case I have to update all three parts, do I have to follow some strict order? What should be updated first?
Anyone did this before on Win7 models? -
Okay I should post my benchmarks and temps tomorrow
just figured out where the logs were stored lol
-
lol I knew you were smart enough to figure this out ! Well done !
-
-
-
Althoug the vBIOS isn't necessary. -
-
You are most welcome. Good luck.
Okay back on-topic, so I did the test and logged the whole thing.
3DMark11 (slightly lower than my regular clocks due to my last session in my suite)
NVIDIA GeForce GTX 680M video card benchmark result - Genuine Intel(R) CPU @ 2.60GHz,Micro-Star International Co., Ltd. MS-1762 score: P7282 3DMarks
CPU (to all wondering about the other cores, I can upload them if you want, only 3 *C difference between them)
GPU
-
As promised, scores and temps with 310.70 Release drivers (not Beta!).
On stock bios, vbios and firmware.
Briefly - temps remain the same, slight drop in 3dmark11 and slight up in 3dmark vantage.
3DMark11: Score, Temps, CPU, GPU
3DMark Vantage: Score, Temps, CPU, GPU
-
@jaug1337bro
You should really do some stock clocks benchmarks ! No one has made any overclocked tests, also indicate the clocks of both CPU and GPU you're running !
Very impressive temps.
3dmarkVantage being a little more demanding, it'll be interesting if you also do it! -
Just need to finish studying for my exam tomorrow and get it over with, then'll I'll do it -
Folks,
I'd like to summarize temps we've got (both with Cloudfire).
I have laptop with IC Diamond 24 Carat Thermal Compound CPU+GPU applied by GenTech PC.
Cloudfire has the stock one (at least I think so).
As far as I can see -=$tR|k3r=- mentioned that Arctic MX-2 is applied to all stock MSI GT70 laptops, but in fact I couldn't find any proof of that.
-=$tR|k3r=- let me know if I misunderstood you statement from initial post http://forum.notebookreview.com/msi/678611-official-msi-gt-70-w-gtx-680m-owner-s-lounge.html
Anyway,
3DMark11
Stock
CPU - 79C
GPU - 79C
IC Diamond 24 Carat Thermal Compound
CPU - 67C (-12C)
GPU - 75C (-4C)
IC Diamond 24 Carat Thermal Compound with Fan
CPU - 64C
GPU - 69C
3DMark Vantage
Stock
CPU - 83C
GPU - 76C
IC Diamond 24 Carat Thermal Compound
CPU - 75C (-8C)
GPU - 74C (-2C)
IC Diamond 24 Carat Thermal Compound with Fan
CPU - 67C
GPU - 67C
Cloudfire, would you like to try to do some isolated CPU and GPU tests to identify max temps on your laptop?
For example for CPU I've got max temps using IntelBurnTest 2.54, and OCCT 4.3.2 Power Supply test helped me to reach the ceiling on GPU.
http://forum.notebookreview.com/msi...temperatures-3dmark-scores-2.html#post8981896 -
Awesome,
+rep when I get home.
Wow, Icdiamond is quite impressive, and with the fan those are really good temps. -
Decided to go with almost full OC on both CPU and GPU. Just for fun... but things got serious lol
need to get 30k now I MUST
Enjoy,
3DMark Vantage
CPU (clocked to a delicious 3,9GHz, see screenshot) -regarding the temp, it was the HIGHEST recorded core temp, so the others are lower
GPU (clocked to a delicious 1037/2550 OV vBIOS, see screenshot)
-
Any hint on how to OC CPU?
-
-
WhatsThePoint Notebook Virtuoso
The difference in cooling ability between all the premium thermal compounds when properly applied to the chip and heatsink is minimal.It's about 2 degrees C.
Thermal Compound Roundup - February 2012 | Hardware Secrets -
Next thing is that there are plenty of other reviews, e.g. Thermal interface: testing the 28 current models of thermal paste
As you can see there is a signifficant difference -10C - -12C between different thermal compound options.
I would agree in case of premium compounds, but not for all of them. -
-
WhatsThePoint Notebook Virtuoso
I have no experience with the Coollaboratory Liquid Pro(maybe my next Amazon order) but I've done a lot with the MX and Tuniq lineups.
The MX-3 is the best of the MX group in my notebooks and desktop and my 1st hand Tuniq experience is that the discontinued TX-3 was the best and TX-2 not far behind.My TX-4 experience is not very good.The stock MX-2 used by MSI isn't bad but I've found the factory application to be excessive.
I was planning on doing some IC Diamond testing a while back but the shipment was lost by the USPS.
I also have had good result from Artic Silver 5 when tested 14 days after application.
Some GT70 0NE temperatures + 3DMark scores
Discussion in 'MSI' started by Cloudfire, Dec 5, 2012.