Jump to content

Core i5 2430M clock speed random drops


Recommended Posts

So I recently bought a new laptop, namely the MSI GE620DX. It has the Core i5 2430M cpu and Geforce GT 555M (2GB) gpu. Games run pretty well, especially for it being a laptop, but every now and then I get stuttering/lag in games. At first I was unsure what's causing it, but I ran RealTemp during gaming and let it log the cpu speeds along with temps. Temps are fine, but the clock speed randomly drops to 798MHz for about 5 seconds and then back to normal. This occurs every 15-20 mins and can get quite annoying during gaming. I've updated the BIOS and even the EC firmware. The last one should have done the trick, but unfortunately it didn't. I have no idea how to check the EC version, but the BIOS is the latest version according to CPU-Z. I've googled it and couldn't find any other solutions. I also tried disabling the intel speedstep from BIOS and that actually fixed the problem, but it's a laptop and I don't want it to run on 2,4GHz when it's on battery. Furthermore, I would still like to have the turboboost during gaming.

Any kind of ideas and help are appreciated!

Share this post


Link to post
Share on other sites

I double-checked everything under advanced power options, seems fine. I also use gaming mode from MSI S-Bar software, which should keep the speed 2,4-3GHZ at all times and it actually does, but still gets this 798MHz spike in there every now and then during gaming.

Share this post


Link to post
Share on other sites

I am sure you have already done this, but have you tried making sure that all your power saving functions are off? Both in windows and the intel control panel?

That's a good way to kill the battery. :P

 

 

OP - if disabling Speedstep fixes it then it sounds like a BIOS bug. Perhaps you can call the manufacturer to see if they have a fix in the works?

Share this post


Link to post
Share on other sites

MSI says that the newest BIOS and EC should fix it, no further fixes nor updates to be found. I also found some info about the GPU overheating and thus triggering the CPU downclock. I'll test it in some games and report back with temps.

 

EDIT: Tried DiRT 2 & 3 and BurnOut Paradise. Couldn't get the GPU temp over 87C, which shouldn't be too high. In DiRT the game would stutter at 84C already, but in BurnOut Paradise it went till 87C without a problem and started the stutter at around 85C and 15mins into the game.

Edited by PingoPongo

Share this post


Link to post
Share on other sites

The BIOS often underclocks the CPU to prevent overheating. This is often based on the CPU power instead of temperature. Check the vcore on the processor while gaming, and see if it corresponds to the underclocks?

Share this post


Link to post
Share on other sites

I am not sure how to check the vcore during gaming as CPU-Z doesn't support info loging. Maybe just run the game windowed and check manually?

I checked the system info from BIOS and both the BIOS and EC are the latest version. I also played burnout paradise for some hours and still had the problem. I had Open Hardware Monitor running and it logged max temps as follows:

CPU core#1 92C

CPU core#2 89C

GPU core 88C

CPU power draw 32W

I know it's a little hot, but it's a hot summer and the ambient is at least 26C (probably more).

I checked the voltage during Prime95 test. CPU-Z reported 1,181V during load and 0,755V during idle (speedstep clocked it down too ofc).

 

Also reapplying the thermal paste is out of the question as I have this warranty sticker on one of the screws. But if the problem is caused by temps then I could easily buy a laptop cooler. I just don't think 92C is extremely hot for a laptop CPU, maybe I'm wrong.

Share this post


Link to post
Share on other sites

I am not sure how to check the vcore during gaming as CPU-Z doesn't support info loging. Maybe just run the game windowed and check manually?

I checked the system info from BIOS and both the BIOS and EC are the latest version. I also played burnout paradise for some hours and still had the problem. I had Open Hardware Monitor running and it logged max temps as follows:

CPU core#1 92C

CPU core#2 89C

GPU core 88C

CPU power draw 32W

I know it's a little hot, but it's a hot summer and the ambient is at least 26C (probably more).

I checked the voltage during Prime95 test. CPU-Z reported 1,181V during load and 0,755V during idle (speedstep clocked it down too ofc).

 

Also reapplying the thermal paste is out of the question as I have this warranty sticker on one of the screws. But if the problem is caused by temps then I could easily buy a laptop cooler. I just don't think 92C is extremely hot for a laptop CPU, maybe I'm wrong.

 

Intel state the maximum CPU temp on your i5 to be 100C, but in my experience, Intel CPUs generally max out at 70C, but it really depends on what your laptop is designed to handle. I would recommend checking your manual if you are particularly worried about temperatures.

With regards to your issue, I would try running OCCT for a while (Atleast 30min), it will stress-test your system, and will log info such as CPU clocks and temps, voltages etc. When you stop the test, open the log folder and see if the underclocks correspond to VCore peaks.

Share this post


Link to post
Share on other sites

  • 3 weeks later...

Okay, well took me long enough, or too long even.

And yes, both the core clock and vcore drop at the same time to 798MHz and 0,750v respectively.

 

I would suggest that the reason for your issues are throttling after all.

You have previously stated that you can't reapply thermal paste due to a warranty sticker, so I would suggest the best thing you could do is buying a laptop cooler, or perhaps contact your manufacturer and ask if you could send them your laptop and they could troubleshoot the high temperatures (Could be caused by a build up of dust or a fan failure etc.)

Share this post


Link to post
Share on other sites

I would suggest that the reason for your issues are throttling after all.

 

Exactly. After reading the thread and thinking more about this problem, I can find no other reason behind the frequency drop other than thermal throttling.

In that case, RMA seems to be the only solution to me.

Share this post


Link to post
Share on other sites

Please sign in to comment

You will be able to leave a comment after signing in



Sign In Now
×
×
  • Create New...