SOLVED FX 6100 is acting on me. trying to OC.

Hi fellas,

I have successfully overclocked quite a few Phenom II's and I was running Phenom II x2 555 unlocked to x4 B55 and overclocked to 3.8 easily on GA-870A-UD3. Yesterday, I got new FX 6100 from one of my friend and replaced old B55 with new FX 6100.
Computer is running Fine, tried to Overclock but its acting on me.
Currently its running at 3.8 (x19) but once I go for 4.0 it does not post, computer turns on, all fans are running, no light on MB but no display.

Next thing, I turned off Turbo, C1E, C6, Hardware thermal Control and cool & Quiet still CPU-Z shows lower clock speed while not stressed? I was expecting it to stay at max Clock speed even in idle after turning all green stuffs. Is there something I am missing?
Will upload CPU-Z tabs and HWMonitor screenshots in a while with P95 at 3.8GHz.
thanks

Drosera
PS: I know, 870A MB does not have Heatsink on MOSFETS but it has 8+2 phase and I am keeping eyes on MOSFET temperatures as well. This is in fact pretty good board for Overclocking, if needed I can add aftermarket heatsink on MOSFETs.

Check for EPU and APM in Bios and make sure you're set to performance in windows.

Johan45 said:

Check for EPU and APM in Bios and make sure you're set to performance in windows.

Click to expand...

Running P95 while typing now and can see multiplier is fluctuating betn x15 and x19. Volatge on HWMOnitor is changing between 1.21 to 1.29 while running P95. Computer is set to High Performance (windows 8). Could not find EPU or APM on BIOS, what are those actually and where is it supposed to be on BIOS? any idea? It does not make any sense to run P95 as multiplier is acting here. thanks.

From what I have seen said on the net, the 8xx chipset boards are not going to have APM in bios and that may become an issue with overclocking an FX procesor. Are you running an FX bios for that motherboard?

RGone said:

From what I have seen said on the net, the 8xx chipset boards are not going to have APM in bios and that may become an issue with overclocking an FX procesor. Are you running an FX bios for that motherboard?

Click to expand...

I would Say "YES". Latest BIOS for this MB is FEE and supports FX 6100, 6200, 8100 etc but not 6300.
FX 6100 is in MB's CPU support list.

thanks

That's a bummer

I have seen a few boards that never seemed to release in Rev 1.0 but few. Gigabyte website shows the five Revisions of the board as below. Only the last revision says anything about supporting the AM3+ processor. So your results are going to vary anyway because of the 8xx chipset and then on what is the Revision of your motherboard.
Rev. 2.0
Rev. 2.1
Rev. 2.2
Rev. 3.0
Rev. 3.1 < the only one with any mention of some sort of support for AM3+ processor.

RGone said:

I have seen a few boards that never seemed to release in Rev 1.0 but few. Gigabyte website shows the five Revisions of the board as below. Only the last revision says anything about supporting the AM3+ processor. So your results are going to vary anyway because of the 8xx chipset and then on what is the Revision of your motherboard.
Rev. 2.0
Rev. 2.1
Rev. 2.2
Rev. 3.0
Rev. 3.1 < the only one with any mention of some sort of support for AM3+ processor.

Click to expand...

YEP, its rev 3.1 which supports AM3+, its in my signature, sorry I did not mention that on my First post. I do not know what else can I disable in my BIOS so that it always runs at max clock speed I set for it.

Even the AOD work around would not work for APM conrol. See post #5. I have no idea you will have any such luck.

Overclocking Issues with GA-870A-USB3 v3.1

RGone said:

Even the AOD work around would not work for APM conrol. See post #5. I have no idea you will have any such luck.

Overclocking Issues with GA-870A-USB3 v3.1

Click to expand...

Thank you,
Yes, I just read that post and even custom BIOS, he got did not work well for him, I do not know what can I do??? I might just have to go back to my old x4 B55,
Before that, I will give Gigabyte a call to hear what they say.

thanks again.

This is all I can see on BIOS screens:

I worked with people with FX processors for nearly one year before I bought one. Having worked for a motherboard company for 6.5 years I knew that IF what I had seen in the forums was anything like "real" then a cheap board would not do the job.

I sat at my keyboard looking at the EGG and had already had 6 or 7 people buy the original 990FX Sabertooth because it was my dream motherboard. I saw where they were almost to a man happy with the Sabertooth. Anything man makes will have issues so the majority being very glad of their purchase, I was ready to hit the Sabertooth buy button at the Egg. At the very last moment I took the Sabertooth out of the cart and inserted the CHV board for about $60.00 more and I could not really be more pleased. Oh it has some things I will likely never use for the extra $60.00 but it is a tough board and overclocks just awesomely. AND it was the correct chipset for the FX processors.

After 6.5 years of using every cheap to top of the line motherboard DFI manufactured, I knew that there was a difference in money spent. I even knew much of what the money went towards. So once I made the split second decision to buy the Asus CHV, I have never felt disappointed. That is a pretty extreme statement. Usually with anything man-made there is something disappointing. But after working with these users and their FX processors I knew pretty much what to be believing an FX would do. Clock pretty good and run HOT as the proverbial h*ll. And they do.

But the first CHV bios for FX had APM enable/disable in it. Gigabyte had to release Rev 1.1 cheaper UD3 boards and put components on them to have APM enable/disable. My $60.00 bought me "ready" to go and it was very pleasing. Most evvery hiccup that users come in here with are not my problem. Period. Well designed, up to date parts and pieces can make the journey so much more enjoyable.

You won't really lose a great deal in going back to Deneb core processor. You really do have to overclock the pee ouf the of the FX to get a great deal of benefit. Once a certain level of overclock is reached the heat is Ginormous. Heavy duty water cooling is a must. My video eddittting that has software that can use all the cores of my FX-8350 is even faster than my FX-8120 was at the same speed. So PD is better than BD anyway. So not using FX-6100 in my mind is no great loss if you have a good Deneb. PD, vishera based FX processors are really better. Maybe by the time you decide to upgrade to the real deal mobo for FX style processors they will have a slight update to the Vishera based FX and you can really get some performance.
RGone...

Thanks RGonester,

I bought this Motherboard about 2.5 yrs ago and by then, it was one of the good MBs for Phenom IIs. Technology is advanced in such a fast pace that any parts or electronics are getting obsolete within a year a two.

I really do not feel like buying new MB right now and I got this FX 6100 free and thought to play with it. I do not do gaming, sometimes my 4 yr old cutie sits on this computer and play angrybirds or alike games that's it and all.
But quite often, I do record Kids shows on TV on this computer using media center and has to do video converting. I thought 6 cores might do better job for this task than the 4 cores I had.

About the issue, Just talked with a guy in Gigabyte USA tech department, sounds like he understood the problem well but had no solution right away (obvious). He had no idea if the issue has been raised by some other customers or not?
He has promised to get back to me within a day or two if he can give me any updated unofficial BIOS / beta that might help to resolve the issue I am having.

Hope, to hear from him soon.

thanks

One thing I do not understand with this CPU and MB is :-

I am running Folding@Home client now and CPU-Z shows constant 3.8 GHz (x19) though, voltage is at 1.296 constant and core temp is max 33 C.

why is it no throttling on F@H and only on P95????

why is it not throttling on F@H and only on P95? = I might make guesses but up until now you are the only one I have seen that has said that about F@H and P95. 18 months into FX processors and I just cannot remember anyone else making that statement.

From the fall of 2011 to maybe April or even May of 2012, there were plenty of throttling going on but I just do not remember that particular circumstance coming to the fore. Maybe it is something that is tied to that motherboard and chipset alone. Nothing I own could begin to simulate for looking for answer.

Without any chance to simulate and never having seen it in here with those months of early throttling, I am just not going to even guess why you see what you see. Odd for sure.
RGone...

From my experience with F@H and prime on my 8350 I have found that my cpu temps are higher during prime then on F@H about 4-5c. My core/package temps are about the same. Why I have no idea but that's what I saw after running F@H for 4-5 days 24/7. Maybe your cpu temps are causing the issue not the core/package temps? Also my llc will bump my cpu v core to 1.476 from 1.464 where F@H it will stay at 1.464 all day long. I'll post 2 screen shots one of prime running for 15 min the other after 4 days of prime both pics at at a 4.7 overclock.

Drake that is interesting.
Maybe your cpu temps are causing the issue not the core/package temps? = At the release of the Bulldozer processors the cpus were throttling from high Cpu/Socket temps. Well actually we did not think the socket temps were that high but the throttling came on anyway and that is when the AOD work around came on the scene in the early days and then the mobo companies put APM enable or disable on the 9xx chipset boards and the socket temp throttles all seemed to disappear.

Also my llc will bump my cpu v core to 1.476 from 1.464 where F@H it will stay at 1.464 all day long. = That would seem to indicate F@H is not loading the cpu down to the same extent that P95 does. Yet "csimon" has that thread that seems to say for him at least it is the other way around. Crazy stuff on some of these systems.
RGone...

Rgone if you remember the comment by csimon and ssjwizard id what made me run F@H to see if my 4.7 24/7 OC was indeed stable, which it was. What was interesting to me was the fact that I noticed that my temps weren't as high running F@H, I also know that my ambient temp were the same. I checked my cpu usage and it was at 100% on both prime and F@H. So I looked at other variables in the HWmonitor which made me notice that CPU V core being raised, if you also notice the cpu/nb, cpu and Vin8 are also slightly higher. It gets me thinking the certain portions of prime may stress the processor and or Dram harder.

I am at work right now and typing on cell phone. One thing i know for sure is cpu/socket temp on either F@H or P95 were below 45. I did not check all other parameters. I will check all other veriables and try to record them all tomorrow and of course will post here.
Thanks

Mandrake4565 said:

Rgone if you remember the comment by csimon and ssjwizard id what made me run F@H to see if my 4.7 24/7 OC was indeed stable, which it was. What was interesting to me was the fact that I noticed that my temps weren't as high running F@H, I also know that my ambient temp were the same. I checked my cpu usage and it was at 100% on both prime and F@H. So I looked at other variables in the HWmonitor which made me notice that CPU V core being raised, if you also notice the cpu/nb, cpu and Vin8 are also slightly higher. It gets me thinking the certain portions of prime may stress the processor and or Dram harder.

Click to expand...

Yes, I remember you moving to do F@H for a few days. Five I think it was. The result was you seemed quite stable. But we did not go over the voltage stuff or I don't remember it now.

What LLC are you using? The way I understand it CPU LLC responds to Voltage and not current load. So theoretically the voltage dropped due to load and LLC responded with an increase. Certainly the rise in temps shows load increase.

I cannot tell if we are chasing a rabbit or not. "csimon" had P95 stable but failed F@H if I remember. You had P95 stable and F@H stable. So what is the difference? Any difference worth worrying about? We got people going thru the forum now dissing P95 Blend for stressing. Does not matter to me really. I am going to test with P95 anyway and Lin-X. There are stressers not even as well known as those two are known. I have run a few of them. Six of one and Half a dozen of the other as far as I could tell. Some did not seem to fully load or as you saw the temps were not the same over the different stressers and that seems to also indicate a difference in load. But how in the heck are you to 'see' what the real differnce is when the cpu shows 100% load in either situation? Does a 100% load have a difference?

I only know that I could walk thru 50 runs of Lin-X at a noticeabley lower Vcore than I had to use for running P95 Blend without some issue showing up.

A lot of the weird stuff we see today, I just don't remember some of this odd crap back in other cpus and chipsets. Sorry I cannot think of more yet.
RGone...