• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

Fx 6300 oc question

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.
I wanted to update BIOS but B15.0630.1 version that is available for download is released before my mobo gone to market. Since I thought it might couse some issues with OC, current version is F1. Whatever I do it seems I am not able prevent drops in prime and core failure remped up vcore all the way to 1.44 tried reducing multiplier using different LLC-s. Only few things cross my mind now BIOS and Fresh installation of Windows could any of those be causing those problems
 
Last edited:
Intel burn test showed the most increase in temps up to 58c after about 15min "Max setting". Now all others wouldn't go above 48c, overdrive and prime all were on 4.2. How I can check individual core temps? Every single software I install wont show up all cores only single one tried to set it up to show but it seems I cant...

"All others" What are you referring to? All other stress testing programs?

Okay, I misunderstood what you were saying in an earlier post about core temperatures. I thought you were reporting on the temps of individual cores and at the time I thought to myself that with the FX series of CPUs you only get one temp - the package temp. There is only one temp sensor on the FX die. So in Overdrive, for instance, when it shows six core temp graphs five of them will be repeats of that same temp sensor reading.

What program are you primarily using to read the package temp? CPUID HardwareMonitor?

Language is a bit of a challenge at times so please work at not blending too many thoughts into one sentence.
 
I wanted to update BIOS but B15.0630.1 version that is available for download is released before my mobo gone to market. Since I thought it might couse some issues with OC, current version is F1. Whatever I do it seems I am not able prevent drops in prime and core failure remped up vcore all the way to 1.44 tried reducing multiplier using different LLC-s. Only few things cross my mind now BIOS and Fresh installation of Windows could any of those be causing those problems

This is confusing to me. Why did you want to roll back your bios to an earlier version? "Updating" means installing a newer version. "Flashing" the bios is a more general term referring to changing the bios version, whether back to an earlier one or forward to a newer one.

Do you have the latest bios installed already on your motherboard? I would think F1 would be the bios version installed when the board was originally released. You should be trying the newest one that is available.

- - - Updated - - -

I looked on Gigabyte's website and there is only one bios available at this time, F1. Perhaps the company will make others available in time but the fact is they will be moving away from supporting socket AM3+. It is now a dead technology since AMD is soon to come out with socket AM4 for the new Zen CPU about to debut. You would just have to check back on the website every now and then.

I would not assume your instability problems are bios related. They are probably either due to low voltage or high temps. Keep a few things in mind:

1. Press the F5 key while in bios to return all settings to default/auto
2. Set your memory frequency to "XMP"
3. Leave your vcore on auto and leave your processor speed at stock or auto for the time being.
4. Run the 1 hour AMDOverdrive stress test. Let's make sure the system is stable with no overclock.
5. Post back with a picture attached that shows what the core temps are Overdrive after about 30 mintues of the stress test. This would be under Status Monitor>CPU status. We need to see what the "thermal margin" is.
6. Also include a pic of the frequencies and voltages under Performance Control>Clock/Voltatge.

You have never attached any pics with your posts in this thread. Do you know how to do this? First, in Windows Accessories find the Snippin Tool. Use it to put a frame around the images you wish to attach and also to save them to disk. Then click on the Go Advance button under any new post window. That will load the forum's built-in attachment tool and the rest will be obvious.

A picture is worth a 1000 words, especially when dealing with different languages.
 
Last edited:
Sorry about my English grammar skills xD.
I tried Overdrive. prime95 and intel burn. Overdrive and Prime never have gone over 48c in package, in other hand intel burn test would go mad at maximum torture test up to 58c and if froze winddows after 20min in it. I use HWMonitor for temps, cpu speed control, vcore etc.
I thought there are separate sensors for each core I have misunderstood you, all temperatures I was referring were from cpu package in HWMonitor.

Those BIOS "updates" were confusing to me thats why I asked but by using some logic its useless to rollback, just was curious.
I have returned all to stock when I noticed instability, tested with increasing vcore playing aroung with LLC reducing speed... Thats why I am lost, I managed to OC on old **** board
but cant on this one, that kinda odd isnt it. Would windows reinstall help?
I will do as you said and post results tnx on info

EDIT: Well I dug trough UEFI and it seems I cant find anywhere in Memory Settings XMP profile. Maybe my memory cant be set to XMP? I can only increase speed up to 2000...
 
Last edited:
I'll wait for the pics.

AMDOverdrive does not give package temps. It gives thermal margin temps which is the distance in degrees of how close you are to TJmax (the thermal throttling threshold).

As we have discussed before, your CPU cooler will be a limiting factor in your overclock. I would recommend not using Prime95 or IntelBurnTest as stress testers for the reason they do, as you have found out, generate extremely high temps. For this reason many people are abandoning those two in favor of other stress testers that do a good job of checking for stability in terms of real life computing but without generating so much heat.

For stress testing, just stick with the AMDOverdrive stress test for now. Two other good ones that don't create such high temps are AIDA64 Extreme and Realbench. Many people are using those instead of Prime95 and IBT these days. Remember, we are trying to confirm stability in your overclocking, not trying to test your cooling capacity.
 
IMG_20170120_181901.jpg IMG_20170120_181916.jpg

Prfoles? wtf... also when I get in there is no XMP at all??:confused:
Can I do it without xmp?

By the way thermal margins were fine steady at 25c
 
Let's check the memory settings (and other bios settings) another way. Please download and install a free program called CPU-z. Run the program and attach pics of these three tabs: CPU, Memory and SPD. The "Memory" tab will show what actually are the memory setting sin bios. The SPD tab will show what the memory manufacturer's recommended timings and voltages are at various frequencies and for the XMP setting. We can compare the two to see if your memory is running in XMP mode. But one thing you need to know to make sense of all this is that the memory frequency reported in the CPU-z "Memory" tab will be half of what it actually is. This is because it is reporting the DDR bus frequency rather than the DDR2 frequency. So in t he pics below, if you double the 688 you would get my actual memory frequency which is 1376 mhz (I have the 1333 mhz memory slightly overclocked).
 

Attachments

  • CPU-z.PNG
    CPU-z.PNG
    28.4 KB · Views: 22
No, you don't have to run your memory in XMP mode but XMP will give the best performance. I'm sure there is a way to do this. We just need to figure it out in your bios.
 
bzhibju.png sasaasddd.png sasasaasaa.png

I will run test without it then and post. I think my bios just derped out for some reason
 
By the way, if you have a flash drive you likely can capture and save bios/UEFI screenshots by using the F12 key while in the UEFI itself. It saves them to the flash drive. That will give better quality pics than using your smart phone camera. If you do use "F12" to capture bios pics I think you may have to "resample" them before you can attach them to your posts as I think they will exceed the file size limit for attachments otherwise. You can resample them with a free program like Irfanview. Resampling simply removes some pixels to make the file size smaller.
 
I will try using that. I saw that option but it says something about formats I think something about FAT32 but not sure rly. Anyway I will check it out if we gona need those.
 
It appears that your RAM does not have an XMP profile to select in bios. That's why it wasn't there. By the way, Kingston RAM is famous for not playing well with AMD systems. All the experienced overclockers on this forum will tell you that.

Other than that, it looks like your timings are what they should be for that frequency.

If your thermal margin in Overdrive was 25c that means your package temps was probably about 45c. I believe the TJmax for the FX series is about 70c.

NOW HERE THIS! Even though the TJMax is 70c it is well known that WHEN OVERCLOCKING AMD CPUs they become unstable before you reach TJMax. In fact, from experience we know that when overclocking AMD CPUs instability sets in at a package temp of 55-60c. So you have 10-15c to play with in your overclocking.

Here's my suggestion. Start increasing your core frequency ratio multiplier in 1x steps, run the Overdrive stress test after each increase. When you reach a multiplier that won't allow you to pass the stress test, add .05 to the vcore and retest. Always watch thermal margin temps.
 
Alright should I disable turbo and other power saving features or let them be? Here are results of test you asked for half hour in test including hw monitor if its gona help
dasdas.PNG hhjghjhg.PNG jjhgjgjg.PNG
 
Last edited:
I would disable Turbo, C6 and C1E. It's been so long since I overclocked an FX I hope I'm telling you correctly about disabling these settings.

And let me correct one thing I said. When your multiplier increase first results in a failed Overdrive stress test, add .025 to the vcore, not .050. .050 might drive temps too high.

Also, run the Overdrive stress test for the full one hour. Start increasing the multiplier in 1x increments. Currently, the base multiplier is set to 35x I believe. Increase it to 36x and stress test. If pass, increase it to 37x. When stress test fails, add .025 to vcore and retest, etc.
 
Yes I know what I should do if I get BSOD. I am facing other problem I mean test passes but there is decrease in speed during test. It depends on set speed of cpu example on 3.9 I get droped to 3.8 for few seconds, at 4.2 I get to 3.9 or 4.0. I noticed that before, but the problem is in games like GTAV and it knows to lag like hell with massive fps drops at times...
I tried several things before like LLC, vcore, droping multiplier and nothing solved it. Yes I still had temperature room to play with Thermal Margin would be at 25c as I said before
 
Read this: http://www.tomsguide.com/answers/id-2486092/graphics-card-run-gta.html

"Since GTA games on PC are notorious for FUBAR optimization, you'll need a really strong rig to play this with less stuttering and fps drops. Like, at least an i7-4790k overclocked to the ceiling, 16 GB RAM, and two GTX 980s in SLI to play it on medium settings @ 30 fps with as little lagging as possible."

Your PC doesn't come close to meeting those hardware requirements. An i7- 4790k CPU is about twice as powerful as your FX-6300 and you have one video card.

Concerning core frequency dropping during stress testing? Do you have turbo and the other power saving stuff disabled?

Also, please run the stress test with HWMonitor open so we can see what the motherboard socket area temps are maxing out at. Attach a pic of HWMonitor after stress test.
 
.... I play it on stock on medium without problems, without advanced graphical settings. 50-70fps in town outside sometimes more. Check youtube
 
So you are confusing me. In post #75 you say you get bad stutter and lag in GTAV while in you last post you say you have no problems on medium settings without advanced settings being activated. When do you get stutter and lag?
 
Found this and I think it might be helpful: http://www.logicalincrements.com/games/gtav

I note their official recommended CPU requirement is an FX-8350. Looks like your GPU is okay. Looks like it's one of those games that puts a lot of demand on the CPU as opposed to the GPU. So, you can expect stutter and lag at certain points during the game play without sufficient CPU horsepower. That's a classic sign of the CPU being overwhelmed.

I would suggest checking this with the Resource Monitor in Windows Task Manager. Get Resource Monitor up and running in the background and then play GTAV for a few minutes until you have encountered stutter/lag. Close GTAV and then check the CPU usage graph in Resource Monitor. I'm guessing you will see points where the CPU is maxed out.
 
Last edited:
I said on stock unstable OC-s know to screw gaming performance, every game is different in way of rendering things etc so not every game will behave the same way with unstable OC or the configuration. On bad OC game will go full ****** and not stutter rather get stuck for few seconds and as time passes by it will get worse, but as I said not all of them are made the same. GTA is perfect for cpu testing since it uses lots of it more than GPU, and RAM
I use MSI Afterburner cpu isn't maxing out, those requirements are overblown same as many other games.
I just passed on 4.1 had to increse vcore for 2+ meaning 0.050. Now its time for some real time tests to see live performance.
 
Back