Overclock.net › Forums › AMD › AMD CPUs › AMD Bulldozer and Piledriver Overclocking Guide - Asus Motherboard
New Posts  All Forums:Forum Nav:

AMD Bulldozer and Piledriver Overclocking Guide - Asus Motherboard - Page 68

Poll Results: Was this helpful?

 
  • 96% (404)
    Yes
  • 3% (13)
    No
417 Total Votes  
post #671 of 4087
Thread Starter 
Quote:
Originally Posted by asdf2 View Post

Hi! First, sorry for my english.
So, I've got a problem with my mobo. If I try to set higher 'FSB' (host clock) for example 220 then my memory and NB and HT clocks get higher. That's okay i set them lowe (memories on 14xx, NB 2200, HT2640, sorry i dont remember the numbers.. ) Then I save it reset, and after it just dont beep this machine... And nothing on screen. Then reboot (long press power button, to power off) it says 'last boot was failed...' so i just enter the UEFI and there i can see: memory: 14xx, NB: 2400, HT 2860(!!)

Si i think it just dont save my NB and HT clocks, and my mobo still run them at 11x220 and 13*220. What can i do? (im running 1503 bios)

And my mobo is: M5A99FX PRO R2.0

I've never seen it where the motherboard would save some settings and not others.

Are you using Manual Overclock mode?
Is your CPU being overclocked (if it's unstable and fails to boot, it might set the HT and CPU/NB back to the stock multiplier upon reboot)
post #672 of 4087
@ComputerRestore, et al,

Thank you for a wonderful guide.
I've been a professional geek for almost 20 years, and I've never OC'ed a thing. I recently retired an old dual-Opteron mobo (ca. 2006, Opteron 140 or 160), and started testing some code on it. It was painfully slow, and since it was just a dev machine, I switched that CPU/mobo out for consumer parts, but something more modern. I ended with an ASUS M5A99X EVO R2.0 and an FX-6350. While I'm bummed that I didn't get the 8350, I'm quite happy with the rig. My project compile went from overnight-and-then-some to ~2 hours.

These "Black Edition" FX processors looked like they were "easy" to OC, and since I bought a decent heat-sink to go with it (Noctua NH-L12), and because the case has lots of room (Old CM Stacker STC-T01), I decided to try overclocking for the very first time. So, thanks again for the OP and everyone else who chimed in with their experiences.

Long story short, I was able to get 4.4 GHz on air, passing Prime95 Blend (75% memory) for 12 hours and also passing my own torture test.

I was bummed I couldn't get more, because I'm only hitting about 55-deg at load (Prime95, Blend, 75%). But, despite 5 days of fiddling CPU, VDDA, and any other setting I thought applicable, I just couldn't get 4.5 stable for more than 3 hours of P95 or my personal test suite. I have a notion that the rig could handle it, but it needs more finesse than I have time to develop at the moment.

My personal test suite consists of building an entire GNU/Linux "distribution" from scratch. It's based off LFS (the Linux From Scratch project), with my own bits for cloud-deployment (i.e., using Xen as a VM hypervisor). My script (again, heavily leveraging the work the wonderful folks at LFS provided) builds an entire GNU/Linux system--including bootstrapping binutils, the compiler, then the rest of the system, and finally the kernel. I run all the available regression-test suites (which, on the old dual-Opteron system would take more than 15 hours). That particular workload is mostly integer math (not much compiling happening on the FPU), and I use '-j 6', which allows 'make' to push independent compiles to all 6 "cores". I'm not sure that this is optimal for the BD/PD architecture, since the "cores" share so much in the module, but it's probably fairly stressful. Building the compiler, glibc, and the kernel is a pretty arduous workload. In fact, at 4.5, even with 3+ hours of P95 passed, the compile suite failed after about 90 minutes.

So, I wanted a 12-hour stable P95 build, and my own test suite to build (with all the regression tests passing).

This is Day-6, and I've finally achieved that, thanks to this board and the various posters.

Here's where I ended up:

Freqs
----
22.0 / 200 / 100 (multi, bus, PCIe)
2200 MHz - CPU/NB Freq
2600 MHz - HT Link Speed

Voltages
----
1.38125 - CPU
1.25 - CPU/NB
2.55625 - VDDA
1.4 - DRAM
1.1 - NB
1.2 - NB HT
1.8 - NB 1.8
1.1 - SB

DIGI+ Power Control
----
High - CPU LLC
High - CPU/NB LLC
130% - CPU Current
130% - CPU/NB Current
Optimized - CPU Power Phase Control
Auto - CPU Voltage Freq
Enabled - VRM Spread Spectrum
T.Probe - CPU Power Duty Control
Auto - CPU Power Response Control
Auto - CPU/NB Power Response Control
130 - CPU Power Thermal Control
130% - DRAM Current
300 - DRAM Voltage Freq
Optimized - DRAM Power Phase Control

The settings in BOLD differ from the OP's "Recommended Settings". Here's my story (I'm sure the veterans know all this, but in case some nubsicle like myself is trying this, I'll offer my rookie insights.):

Initially, I stuck with the Recommended Settings. And, even at 4.6 GHz (multi-only, at 23.0), I was able to get, at stock voltages, a 10-min pass of Prime95 Small FFT. But, once I started the Blend tests, the system failed quickly. Almost always (and in the past 6 days I've done the SmFFT test many dozens of times), it was a single core failing, and always in the 3rd module. I assumed it was the 3rd because I used AMD OverDrive to monitor the system (the UI made it immediately obvious when P95 stopped working) and it was always showing either Core 5 or Core 6 failing.

I started playing with voltages, and here's where things started going sideways.

First, I noticed (I became more aware over time of things to look for) that my DRAM was not showing the right settings. I first saw this in my Memtest+ (which I *always* do before building a machine). It passed 2 passes, and I figured that my version (4.10) was old, and just wasn't seeing the right SPD settings. I learned, after much surfing, that SPD is sort of like a "configuration" written to the RAM, and may not reflect its actual capabilities.

I looked to the BIOS. The settings were off there, too. It showed 1333 MHz, and my sticks had 1600 MHz printed on the box, and these were "name-brand" sticks. Not the revered G.Skills, but Corsair (though the lower-quality XMS3 sticks). Since I had 32 GB of this stuff (4x8), I figured that running it at speed would be nice. In case anyone else has this memory, it's: Corsair XMS3 CMX16GX3M2A1600C11.

So, in the Ai Tweaker, I upped the Auto 1333 to Manual 1600. Boy was that a...mistake.

After that "fix", I wasn't able to get P95 Blend working for more than 2 minutes at a time. Seemed pretty obvious that Blend used a ton of memory, and I either had bad sticks (which seemed unlikely, given the multiple Memtest+ passes) or bad settings (duh). So, I took more steps, but sadly for me, that was BEFORE I read about SPD. I set the DRAM Timings to 9-9-9-24-T1, manually. I thought to myself: "How clever, man, you can totally rock this."

Oh was I wrong. Windows sometimes didn't boot, and sometimes BSODed even before I got to start P95.

Turns out, I should learn to read memory model numbers. The "...C11" should have been a clue. After some more research on the Interwebs, I realized that the timings which were actually verified for that memory at 1600 MHz were 11-11-11-30. So, I plugged that in. I thought: "Good job. You've got it now." Wrong again. More research. Thanks to other blog posts (heck, it may have been another OCN thread) I realized the correct DRAM timing was actually 11-11-11-30-T2.

That worked. Now, Blend was running, but never more than about 1 hour at a time before failure.

So, I was back to voltage-grinding (think MMORPG levels here...).

I kept going up, until I hit the thermal ceiling (also took a while to realize the the PACKAGE temp was actually Core Temp--should have read OP's post more carefully, and that it would be lower than the SOCKET temp (labeled "CPU"). Anyway, I wasn't having much luck with the multi-only method. So, I tried to turn down the multiplier and bumped the FSB. Still no joy.

Then, going back and reading the original post more carefully (that post is dense, I tells ya), I started over. In doing so, I realized that I had never bumped the CPU/NB Manual Voltage setting. Realizing that it was related to the IMC--and that my Small FFTs were perfectly solid and that my Blends were failing hard--I then bumped the CPU/NB voltage to 1.25. This seemed to help, but I still wasn't able to get across the 3-hour mark in Blend.

Same with VDDA. I hadn't adjusted that up from what Auto was (don't have clear notes about that setting). Went back to read the original post. Same story; I hadn't been careful enough...

Also, at some point around Day-3, I installed OCCT. ZOMGwonderful. That told a lot of the story right away. I set it to watch Vcore. Now, after achieving a stable setup, I have no idea if what I did was "right", or even "on track". I can only observe the result, and I only have a vague intuition about correlation. Before correctly setting VDDA, which by Day-4 I had not touched (again, not careful enough reading the original post), I saw two different things with Vcore:

1) On Prime95 Small FFT, I saw a straight line. Virtually no deviation once load was applied, at least as far as the sampling aliased the measurements.

2) On Prime95 Blend, I saw a...saw. The voltage was super-erratic.

On Blend, I would notice an immediate spike. Followed by an erratic shark's mouth of voltage "teeth". Now, for those who are even casually versed in signals knows that it's pretty hard to tell a spike from a droop. A short spike might actually be 2 long droops, etc. So, what I'm describing might not actually be the case, but I'm going on my gut--since I have no theory, only observations. Point is, I was seeing lots of spikes, a lot of the time. For example, if I had CPU at 1.38125 (@ 4.5 GHz), idle Vcore was 1.38. When Blend started, there were spikes up to 1.392. Then, maybe there were stretches at 1.392, and then other stretches at 1.38, and in between those plateaus there were just these awful-looking spikes and droops, occurring at ~1sec intervals. At that point, it's hard to tell one from the other.

The key word now, of course, is "droop". And so started an investigation into Vdroop (which I thought was hacker slang). I ended up in a mess of technical documents, far out of my depth. So, here's where I had to turn knobs "in the dark", and made some "educated" guesses. Vdroop was a real thing. I developed a vague notion of what LLC was doing. And, I thought that might be the cause of the spikes. I got worried that spikes were causing the instability (pushing the CPU too hot with LLC st to "Ultra-High"). And, since VDDA helps with stability, I did two things (not at the same time, but I think I've been going on for a quite a while now, given how small my scroll thumb has gotten): I upped VDDA to 2.5. That helped. I was getting past 3 hours. I also changed the Load Line Calibration settings. I went from CPU LLC of "Ultra High" to just "High". I also eventually landed on VDDA at 2.55625, because that seemed to offer more stability than just upping the CPU Manual Voltage setting and didn't seem to push the temps as hard.

In the end, I still see frequent voltage spikes, in between stretches of voltage quiescence, even in my final stable settings. Vcore swung from 1.356 to 1.368. But, they look like droops now (which I expected, from a lower LLC setting). And, that sorta makes sense to me....

Even still, the Vcore swings look wrong. I feel like, even at load, Vcore should be stable. And, that's not to say it isn't ever stable. I might get 20 minutes of a flat line, but that's followed by 5 minutes of swings. But, maybe that's how voltage always looks when running Blend...IDK...maybe memory accesses cause droops/spikes. I'd very much welcome a knowledgeable opinion here on the Vcore swings.

I lowered the DRAM voltage down to 1.4 eventually, and that didn't have any observable effect on the saw-tooth voltage, though OP relayed that lower DRAM voltage might help reduce the stress on the IMC, which I figured was good given that Blend--and never Small FFT--was the problem.

So, at the end of the last Prime95 Blend (75%) test, I stopped at 12 hours stable (no cores failed), and the max temps reported by OCCT were 55-deg C (package/core) and 59-deg C (socket/cpu). That seemed healthy enough to me. And, I passed my own compilation torture test in the middle of the afternoon (ambient was ~30-deg C).

I have some images, too, of my air-cooled setup (it's pretty...ugly, but, IMO, clever), and of the OCCT voltage spikes, if anyone happens to have made it past this wall-of-text-crittage, and would like to see those.

Sorry about the ramblage...

TL;DR - Thanks for creating this post. 4.4 GHz (13% OC) on air is a fine result. For me, stability is a big issue--I used it to test the a system I eventually duplicate in production, so stable >>> crazy-fast. Grats, too, for all those who have successfully OC'ed from this post. Perhaps, armed with more time and experience, I'll be able to get better results in the future. And, for other intrepid explorers, READ THE ORIGINAL POST CAREFULLY. Duh. And, may the source be with you.
post #673 of 4087
Quote:
Originally Posted by Destrto View Post

Thanks for the extra details about the Sata connectors on the Mobo. If I'm understanding you correctly, I should move my RAID drives over to the Blue Ports?

As for Prime95, I have the latest build, and atleast for my particular system, regardless of what I tried, stock, overclocked or what else, it would not run without giving some error. My CPU does not seem to be as power hungry as yours, I'm running 4.5Ghz at 1.368V. Prime 95 will, however, run a good number of hours without issues, but eventually it does. Usually after the 7-8 hour mark.

I appreciate the information you've given. It definitely gives myself and possibly others, a better idea of what other causes for instability may be.

I am sorry for my terrible syntax. I am not native english...

What i meant is that raid goes to the white ports (sb950), and all other sata devices must go
to the blue ports in order not to have conflicts and performance issues. Also depending on
the number of your raid devices (usually two or four as raid 5 with 3 devices is not recommended)
you should start populating first the bottom white ports (sata3_1 sata3_2) and move upwards.

As for 7 hours prime95 without errors means that your oc is fairly stable, but something minor
should be adjusted. Most propably you won't ever notice because in real use your not gonna
stress your cpu that much, and if that error is actually ever reproduced in gaming or other use
it might be a hickup, a small framedrop or something that wont affect your experience.
So if your not having issues in everyday use go on and enjoy your oc!

On a side note. Have you tried stock clocks but without default settings in bios?
Meaning that you set the settings as described by the guide and leave only
cpu multiplier, cpu voltage and cpu/nb voltage at default.
I would be very surprised if this setup produced prime95 errors.
Pilerider Mk8
(18 items)
 
  
CPUMotherboardGraphicsRAM
FX 8350 @4.4ghz M5A99FX PRO R2.0 GTX 670 @1162Mhz max 4x4GB Kingston HyperX Genesis 1600Mhz 9-9-9-24-... 
Hard DriveHard DriveOptical DriveCooling
2xSamsung 830 128GB Raid0 Western Digital Caviar green 1TB Sata3 Lg optical dvd rw +r Corsair H80i 
OSMonitorKeyboardPower
Winf 7 pro Samsung S27B370H 60hz 1080p Microsoft X4 Nexus 1000w 
CaseMouseMouse PadAudio
Sharkoon  Razar DeathAdder 3.5g Roccat Taito Creative Labs Recon 3D 
OtherOther
M-AUDIO AV-40 Aerocool Touch 1000 
  hide details  
Reply
Pilerider Mk8
(18 items)
 
  
CPUMotherboardGraphicsRAM
FX 8350 @4.4ghz M5A99FX PRO R2.0 GTX 670 @1162Mhz max 4x4GB Kingston HyperX Genesis 1600Mhz 9-9-9-24-... 
Hard DriveHard DriveOptical DriveCooling
2xSamsung 830 128GB Raid0 Western Digital Caviar green 1TB Sata3 Lg optical dvd rw +r Corsair H80i 
OSMonitorKeyboardPower
Winf 7 pro Samsung S27B370H 60hz 1080p Microsoft X4 Nexus 1000w 
CaseMouseMouse PadAudio
Sharkoon  Razar DeathAdder 3.5g Roccat Taito Creative Labs Recon 3D 
OtherOther
M-AUDIO AV-40 Aerocool Touch 1000 
  hide details  
Reply
post #674 of 4087
Thread Starter 
Quote:
Originally Posted by Qrux View Post


Warning: Spoiler! (Click to show)
Quote:
@ComputerRestore, et al,

Thank you for a wonderful guide. I've been a professional geek for almost 20 years, and I've never OC'ed a thing. I recently retired an old dual-Opteron mobo (ca. 2006, Opteron 140 or 160), and started testing some code on it. It was painfully slow, and since it was just a dev machine, I switched that CPU/mobo out for consumer parts, but something more modern. I ended with an ASUS M5A99X EVO R2.0 and an FX-6350. While I'm bummed that I didn't get the 8350, I'm quite happy with the rig. My project compile went from overnight-and-then-some to ~2 hours.

These "Black Edition" FX processors looked like they were "easy" to OC, and since I bought a decent heat-sink to go with it (Noctua NH-L12), and because the case has lots of room (Old CM Stacker STC-T01), I decided to try overclocking for the very first time. So, thanks again for the OP and everyone else who chimed in with their experiences.

Long story short, I was able to get 4.4 GHz on air, passing Prime95 Blend (75% memory) for 12 hours and also passing my own torture test.

I was bummed I couldn't get more, because I'm only hitting about 55-deg at load (Prime95, Blend, 75%). But, despite 5 days of fiddling CPU, VDDA, and any other setting I thought applicable, I just couldn't get 4.5 stable for more than 3 hours of P95 or my personal test suite. I have a notion that the rig could handle it, but it needs more finesse than I have time to develop at the moment.

My personal test suite consists of building an entire GNU/Linux "distribution" from scratch. It's based off LFS (the Linux From Scratch project), with my own bits for cloud-deployment (i.e., using Xen as a VM hypervisor). My script (again, heavily leveraging the work the wonderful folks at LFS provided) builds an entire GNU/Linux system--including bootstrapping binutils, the compiler, then the rest of the system, and finally the kernel. I run all the available regression-test suites (which, on the old dual-Opteron system would take more than 15 hours). That particular workload is mostly integer math (not much compiling happening on the FPU), and I use '-j 6', which allows 'make' to push independent compiles to all 6 "cores". I'm not sure that this is optimal for the BD/PD architecture, since the "cores" share so much in the module, but it's probably fairly stressful. Building the compiler, glibc, and the kernel is a pretty arduous workload. In fact, at 4.5, even with 3+ hours of P95 passed, the compile suite failed after about 90 minutes.

So, I wanted a 12-hour stable P95 build, and my own test suite to build (with all the regression tests passing).

This is Day-6, and I've finally achieved that, thanks to this board and the various posters.

Here's where I ended up:

Freqs
----
22.0 / 200 / 100 (multi, bus, PCIe)
2200 MHz - CPU/NB Freq
2600 MHz - HT Link Speed

Voltages
----
1.38125 - CPU
1.25 - CPU/NB
2.55625 - VDDA
1.4 - DRAM
1.1 - NB
1.2 - NB HT
1.8 - NB 1.8
1.1 - SB

DIGI+ Power Control
----
High - CPU LLC
High - CPU/NB LLC
130% - CPU Current
130% - CPU/NB Current
Optimized - CPU Power Phase Control
Auto - CPU Voltage Freq
Enabled - VRM Spread Spectrum
T.Probe - CPU Power Duty Control
Auto - CPU Power Response Control
Auto - CPU/NB Power Response Control
130 - CPU Power Thermal Control
130% - DRAM Current
300 - DRAM Voltage Freq
Optimized - DRAM Power Phase Control

The settings in BOLD differ from the OP's "Recommended Settings". Here's my story (I'm sure the veterans know all this, but in case some nubsicle like myself is trying this, I'll offer my rookie insights.):

Initially, I stuck with the Recommended Settings. And, even at 4.6 GHz (multi-only, at 23.0), I was able to get, at stock voltages, a 10-min pass of Prime95 Small FFT. But, once I started the Blend tests, the system failed quickly. Almost always (and in the past 6 days I've done the SmFFT test many dozens of times), it was a single core failing, and always in the 3rd module. I assumed it was the 3rd because I used AMD OverDrive to monitor the system (the UI made it immediately obvious when P95 stopped working) and it was always showing either Core 5 or Core 6 failing.

I started playing with voltages, and here's where things started going sideways.

First, I noticed (I became more aware over time of things to look for) that my DRAM was not showing the right settings. I first saw this in my Memtest+ (which I *always* do before building a machine). It passed 2 passes, and I figured that my version (4.10) was old, and just wasn't seeing the right SPD settings. I learned, after much surfing, that SPD is sort of like a "configuration" written to the RAM, and may not reflect its actual capabilities.

I looked to the BIOS. The settings were off there, too. It showed 1333 MHz, and my sticks had 1600 MHz printed on the box, and these were "name-brand" sticks. Not the revered G.Skills, but Corsair (though the lower-quality XMS3 sticks). Since I had 32 GB of this stuff (4x8), I figured that running it at speed would be nice. In case anyone else has this memory, it's: Corsair XMS3 CMX16GX3M2A1600C11.

So, in the Ai Tweaker, I upped the Auto 1333 to Manual 1600. Boy was that a...mistake.

After that "fix", I wasn't able to get P95 Blend working for more than 2 minutes at a time. Seemed pretty obvious that Blend used a ton of memory, and I either had bad sticks (which seemed unlikely, given the multiple Memtest+ passes) or bad settings (duh). So, I took more steps, but sadly for me, that was BEFORE I read about SPD. I set the DRAM Timings to 9-9-9-24-T1, manually. I thought to myself: "How clever, man, you can totally rock this."

Oh was I wrong. Windows sometimes didn't boot, and sometimes BSODed even before I got to start P95.

Turns out, I should learn to read memory model numbers. The "...C11" should have been a clue. After some more research on the Interwebs, I realized that the timings which were actually verified for that memory at 1600 MHz were 11-11-11-30. So, I plugged that in. I thought: "Good job. You've got it now." Wrong again. More research. Thanks to other blog posts (heck, it may have been another OCN thread) I realized the correct DRAM timing was actually 11-11-11-30-T2.

That worked. Now, Blend was running, but never more than about 1 hour at a time before failure.

So, I was back to voltage-grinding (think MMORPG levels here...).

I kept going up, until I hit the thermal ceiling (also took a while to realize the the PACKAGE temp was actually Core Temp--should have read OP's post more carefully, and that it would be lower than the SOCKET temp (labeled "CPU"). Anyway, I wasn't having much luck with the multi-only method. So, I tried to turn down the multiplier and bumped the FSB. Still no joy.

Then, going back and reading the original post more carefully (that post is dense, I tells ya), I started over. In doing so, I realized that I had never bumped the CPU/NB Manual Voltage setting. Realizing that it was related to the IMC--and that my Small FFTs were perfectly solid and that my Blends were failing hard--I then bumped the CPU/NB voltage to 1.25. This seemed to help, but I still wasn't able to get across the 3-hour mark in Blend.

Same with VDDA. I hadn't adjusted that up from what Auto was (don't have clear notes about that setting). Went back to read the original post. Same story; I hadn't been careful enough...

Also, at some point around Day-3, I installed OCCT. ZOMGwonderful. That told a lot of the story right away. I set it to watch Vcore. Now, after achieving a stable setup, I have no idea if what I did was "right", or even "on track". I can only observe the result, and I only have a vague intuition about correlation. Before correctly setting VDDA, which by Day-4 I had not touched (again, not careful enough reading the original post), I saw two different things with Vcore:

1) On Prime95 Small FFT, I saw a straight line. Virtually no deviation once load was applied, at least as far as the sampling aliased the measurements.

2) On Prime95 Blend, I saw a...saw. The voltage was super-erratic.

On Blend, I would notice an immediate spike. Followed by an erratic shark's mouth of voltage "teeth". Now, for those who are even casually versed in signals knows that it's pretty hard to tell a spike from a droop. A short spike might actually be 2 long droops, etc. So, what I'm describing might not actually be the case, but I'm going on my gut--since I have no theory, only observations. Point is, I was seeing lots of spikes, a lot of the time. For example, if I had CPU at 1.38125 (@ 4.5 GHz), idle Vcore was 1.38. When Blend started, there were spikes up to 1.392. Then, maybe there were stretches at 1.392, and then other stretches at 1.38, and in between those plateaus there were just these awful-looking spikes and droops, occurring at ~1sec intervals. At that point, it's hard to tell one from the other.

The key word now, of course, is "droop". And so started an investigation into Vdroop (which I thought was hacker slang). I ended up in a mess of technical documents, far out of my depth. So, here's where I had to turn knobs "in the dark", and made some "educated" guesses. Vdroop was a real thing. I developed a vague notion of what LLC was doing. And, I thought that might be the cause of the spikes. I got worried that spikes were causing the instability (pushing the CPU too hot with LLC st to "Ultra-High"). And, since VDDA helps with stability, I did two things (not at the same time, but I think I've been going on for a quite a while now, given how small my scroll thumb has gotten): I upped VDDA to 2.5. That helped. I was getting past 3 hours. I also changed the Load Line Calibration settings. I went from CPU LLC of "Ultra High" to just "High". I also eventually landed on VDDA at 2.55625, because that seemed to offer more stability than just upping the CPU Manual Voltage setting and didn't seem to push the temps as hard.

In the end, I still see frequent voltage spikes, in between stretches of voltage quiescence, even in my final stable settings. Vcore swung from 1.356 to 1.368. But, they look like droops now (which I expected, from a lower LLC setting). And, that sorta makes sense to me....

Even still, the Vcore swings look wrong. I feel like, even at load, Vcore should be stable. And, that's not to say it isn't ever stable. I might get 20 minutes of a flat line, but that's followed by 5 minutes of swings. But, maybe that's how voltage always looks when running Blend...IDK...maybe memory accesses cause droops/spikes. I'd very much welcome a knowledgeable opinion here on the Vcore swings.

I lowered the DRAM voltage down to 1.4 eventually, and that didn't have any observable effect on the saw-tooth voltage, though OP relayed that lower DRAM voltage might help reduce the stress on the IMC, which I figured was good given that Blend--and never Small FFT--was the problem.

So, at the end of the last Prime95 Blend (75%) test, I stopped at 12 hours stable (no cores failed), and the max temps reported by OCCT were 55-deg C (package/core) and 59-deg C (socket/cpu). That seemed healthy enough to me. And, I passed my own compilation torture test in the middle of the afternoon (ambient was ~30-deg C).

I have some images, too, of my air-cooled setup (it's pretty...ugly, but, IMO, clever), and of the OCCT voltage spikes, if anyone happens to have made it past this wall-of-text-crittage, and would like to see those.

Sorry about the ramblage...

TL;DR - Thanks for creating this post. 4.4 GHz (13% OC) on air is a fine result. For me, stability is a big issue--I used it to test the a system I eventually duplicate in production, so stable >>> crazy-fast. Grats, too, for all those who have successfully OC'ed from this post. Perhaps, armed with more time and experience, I'll be able to get better results in the future. And, for other intrepid explorers, READ THE ORIGINAL POST CAREFULLY. Duh. And, may the source be with you.

Thanks for posting your Overclock experience. It was both entertaining and I learned a bit more about how these beasts work.

If you could provide a bit more insight into your setting choices I would really appreciate it.

DRAM Current - 130% - In your case, did increasing this setting help with stability/memory usage etc?
Quote:
Since I had 32 GB of this stuff (4x8), I figured that running it at speed would be nice.

Your rig differs from mine, as I was only using 2 X 4GB Modules of Samsung Green which runs at 1.3v. I think that's an area of my guide I need to improve on, as the Recommended Settings are based on pushing a tricyle - where as someone like yourself who is pushing a dump truck may have issues using those settings.

What did you end up with for Ram Speed/Settings? (Did you stick with the 1600Mhz 11-11-11-30 2T @ 1.4v)

Did you have to drop your DRAM voltage to 1.4v to maintain stability with the 32GB?
Edited by ComputerRestore - 6/1/13 at 8:05am
post #675 of 4087
Warning: Spoiler! (Click to show)
Quote:
Originally Posted by DeusXXX View Post

I am sorry for my terrible syntax. I am not native english...

What i meant is that raid goes to the white ports (sb950), and all other sata devices must go
to the blue ports in order not to have conflicts and performance issues. Also depending on
the number of your raid devices (usually two or four as raid 5 with 3 devices is not recommended)
you should start populating first the bottom white ports (sata3_1 sata3_2) and move upwards.

As for 7 hours prime95 without errors means that your oc is fairly stable, but something minor
should be adjusted. Most propably you won't ever notice because in real use your not gonna
stress your cpu that much, and if that error is actually ever reproduced in gaming or other use
it might be a hickup, a small framedrop or something that wont affect your experience.
So if your not having issues in everyday use go on and enjoy your oc!
On a side note. Have you tried stock clocks but without default settings in bios?
Meaning that you set the settings as described by the guide and leave only
cpu multiplier, cpu voltage and cpu/nb voltage at default.
I would be very surprised if this setup produced prime95 errors.
I did actually test Prime95 with manually setting my clocks to stock numbers, and I still received crashes and failures. Prime95 has just never liked my system, and although I know a lot of people swear by it, I'm personally not a fan. Intel Burn Test does the same job (Looking up the technical details explains they both perform the same standard test) and I've gotten reliable results from it on repeated occasions. However, I am also aware that reliability between IBT and Prime95 is user specific, so I stay unbiased towards or against any one program and simply state "this is what worked for me".

Thank you for clarifying about the Sata ports. I do in fact have the Raid drives occupying the bottom most White Slots on the board, and my 2 storage drives on the next 2 above those. Everything is running smoothly for me currently. Games play swimmingly, 1080p videos play without hiccups, and overall I'm satisfied.
Destrto3
(17 items)
 
  
CPUMotherboardGraphicsRAM
AMD FX-8120 M5A99FX PRO R2.0 Radeon R9 270X Kingston HyperX Blu 8GB (2 x 4GB) 240-Pin DDR3 ... 
RAMHard DriveHard DriveHard Drive
Kingston HyperX Blu 8GB (2 x 4GB) 240-Pin DDR3 ... SanDisk Extreme 240 Western Digital WDC WD10 EARS 1TB Seagate ST2000DM 2TB 
Optical DriveCoolingOSMonitor
TSSTcorp CDDVDW SH-222BB EK Supreme LTx (AMD) Windows 8.1 Sceptre 32" LCD Flat Panel TV 
KeyboardPowerCaseMouse
Cooler Master Devastator ThermalTake TR2 RX 850W Corsair Obsidian 750D Cooler Master Devastator 
Audio
OnBoard 
  hide details  
Reply
Destrto3
(17 items)
 
  
CPUMotherboardGraphicsRAM
AMD FX-8120 M5A99FX PRO R2.0 Radeon R9 270X Kingston HyperX Blu 8GB (2 x 4GB) 240-Pin DDR3 ... 
RAMHard DriveHard DriveHard Drive
Kingston HyperX Blu 8GB (2 x 4GB) 240-Pin DDR3 ... SanDisk Extreme 240 Western Digital WDC WD10 EARS 1TB Seagate ST2000DM 2TB 
Optical DriveCoolingOSMonitor
TSSTcorp CDDVDW SH-222BB EK Supreme LTx (AMD) Windows 8.1 Sceptre 32" LCD Flat Panel TV 
KeyboardPowerCaseMouse
Cooler Master Devastator ThermalTake TR2 RX 850W Corsair Obsidian 750D Cooler Master Devastator 
Audio
OnBoard 
  hide details  
Reply
post #676 of 4087
I am not a fan of Prime95 and also tend to prefer IBT to stress the CPU but what Prime95 can do that IBT can not perform as well is testing the memory and the imc for stability which is equally important for overall stability. I will usually keep things at stock except the cpu and work on my CPU overclock. about the time I hit a wall I start my memory overclock in frequency and latency using memtest. And finally I start my IMC stress testing and overclocking if needed using Prime95 blend testing. If I can perform these test at my max overclocks I can pretty much be assured that I can be stable at most anything else.
post #677 of 4087
Quote:
Originally Posted by aas88keyz View Post

I am not a fan of Prime95 and also tend to prefer IBT to stress the CPU but what Prime95 can do that IBT can not perform as well is testing the memory and the imc for stability which is equally important for overall stability. I will usually keep things at stock except the cpu and work on my CPU overclock. about the time I hit a wall I start my memory overclock in frequency and latency using memtest. And finally I start my IMC stress testing and overclocking if needed using Prime95 blend testing. If I can perform these test at my max overclocks I can pretty much be assured that I can be stable at most anything else.

I've found that IBT is able to perform a decent memory stress with the higher test settings, which utilize more RAM, or you can set your own value. Granted, the tests then take longer with the more memory you choose to use. but that's just my personal experience.
Destrto3
(17 items)
 
  
CPUMotherboardGraphicsRAM
AMD FX-8120 M5A99FX PRO R2.0 Radeon R9 270X Kingston HyperX Blu 8GB (2 x 4GB) 240-Pin DDR3 ... 
RAMHard DriveHard DriveHard Drive
Kingston HyperX Blu 8GB (2 x 4GB) 240-Pin DDR3 ... SanDisk Extreme 240 Western Digital WDC WD10 EARS 1TB Seagate ST2000DM 2TB 
Optical DriveCoolingOSMonitor
TSSTcorp CDDVDW SH-222BB EK Supreme LTx (AMD) Windows 8.1 Sceptre 32" LCD Flat Panel TV 
KeyboardPowerCaseMouse
Cooler Master Devastator ThermalTake TR2 RX 850W Corsair Obsidian 750D Cooler Master Devastator 
Audio
OnBoard 
  hide details  
Reply
Destrto3
(17 items)
 
  
CPUMotherboardGraphicsRAM
AMD FX-8120 M5A99FX PRO R2.0 Radeon R9 270X Kingston HyperX Blu 8GB (2 x 4GB) 240-Pin DDR3 ... 
RAMHard DriveHard DriveHard Drive
Kingston HyperX Blu 8GB (2 x 4GB) 240-Pin DDR3 ... SanDisk Extreme 240 Western Digital WDC WD10 EARS 1TB Seagate ST2000DM 2TB 
Optical DriveCoolingOSMonitor
TSSTcorp CDDVDW SH-222BB EK Supreme LTx (AMD) Windows 8.1 Sceptre 32" LCD Flat Panel TV 
KeyboardPowerCaseMouse
Cooler Master Devastator ThermalTake TR2 RX 850W Corsair Obsidian 750D Cooler Master Devastator 
Audio
OnBoard 
  hide details  
Reply
post #678 of 4087
Quote:
Originally Posted by Destrto View Post

Quote:
Originally Posted by aas88keyz View Post

I am not a fan of Prime95 and also tend to prefer IBT to stress the CPU but what Prime95 can do that IBT can not perform as well is testing the memory and the imc for stability which is equally important for overall stability. I will usually keep things at stock except the cpu and work on my CPU overclock. about the time I hit a wall I start my memory overclock in frequency and latency using memtest. And finally I start my IMC stress testing and overclocking if needed using Prime95 blend testing. If I can perform these test at my max overclocks I can pretty much be assured that I can be stable at most anything else.

I've found that IBT is able to perform a decent memory stress with the higher test settings, which utilize more RAM, or you can set your own value. Granted, the tests then take longer with the more memory you choose to use. but that's just my personal experience.

Yes, there are settings in IBT to increase memory to put more stress on the cpu because the computations are more complex. Though, when centered on the memory stability itself, memtest will use a variety of tests and trials that will find failures in memory faster and will tell you exactly what those failures are. There are but a few programs out there that will find errors in memory timings than with memtest. But the cpu/nb can't be tested as well with IBT or memtest. There are better programs for CPU/NB including but not only Prime95 blend test. I use Prime95 blend, AMD overdrive is another and lately I have been using AIDA64 that has proven to help me find the instabilities of the CPU/NB that every day use programs would be shown to have in real time scenarios. But I have learned in my experience that all in all there is not necessarily one program that test and gives answers to overall stability. If someone only tests on IBT and the stability test fails where do they start? Most people would assume CPU and adjust voltage and settings "appropriately" But how will someone know for sure it is the CPU and not memory. IBT can't answer that. Same for testing the CPU/NB through IBT. Does not give a clear answer that it is the CPU/NB that is unstable. Still adjusting setting for CPU voltage and settings. I have learned that turning knobs on one hardware component a lot of times masks the stability with more voltage or high/lower frequencies. Some people concentrating on CPU stabilities could eventually put more unnecessary stress and heat on that one component. But if someone takes time and works on each component, one at a time, they could find the cooler more stable balance.
post #679 of 4087
Quote:
Originally Posted by aas88keyz View Post

Yes, there are settings in IBT to increase memory to put more stress on the cpu because the computations are more complex. Though, when centered on the memory stability itself, memtest will use a variety of tests and trials that will find failures in memory faster and will tell you exactly what those failures are. There are but a few programs out there that will find errors in memory timings than with memtest. But the cpu/nb can't be tested as well with IBT or memtest. There are better programs for CPU/NB including but not only Prime95 blend test. I use Prime95 blend, AMD overdrive is another and lately I have been using AIDA64 that has proven to help me find the instabilities of the CPU/NB that every day use programs would be shown to have in real time scenarios. But I have learned in my experience that all in all there is not necessarily one program that test and gives answers to overall stability. If someone only tests on IBT and the stability test fails where do they start? Most people would assume CPU and adjust voltage and settings "appropriately" But how will someone know for sure it is the CPU and not memory. IBT can't answer that. Same for testing the CPU/NB through IBT. Does not give a clear answer that it is the CPU/NB that is unstable. Still adjusting setting for CPU voltage and settings. I have learned that turning knobs on one hardware component a lot of times masks the stability with more voltage or high/lower frequencies. Some people concentrating on CPU stabilities could eventually put more unnecessary stress and heat on that one component. But if someone takes time and works on each component, one at a time, they could find the cooler more stable balance.

That is very true. I might have to check out these other programs and see what results I get.

About AIDA64, is there a way to see the specific details of any error it gives? one stability test gave a hardware error after about an hour, but I was not able to find any info on which hardware device exactly.
Edited by Destrto - 6/2/13 at 3:45pm
Destrto3
(17 items)
 
  
CPUMotherboardGraphicsRAM
AMD FX-8120 M5A99FX PRO R2.0 Radeon R9 270X Kingston HyperX Blu 8GB (2 x 4GB) 240-Pin DDR3 ... 
RAMHard DriveHard DriveHard Drive
Kingston HyperX Blu 8GB (2 x 4GB) 240-Pin DDR3 ... SanDisk Extreme 240 Western Digital WDC WD10 EARS 1TB Seagate ST2000DM 2TB 
Optical DriveCoolingOSMonitor
TSSTcorp CDDVDW SH-222BB EK Supreme LTx (AMD) Windows 8.1 Sceptre 32" LCD Flat Panel TV 
KeyboardPowerCaseMouse
Cooler Master Devastator ThermalTake TR2 RX 850W Corsair Obsidian 750D Cooler Master Devastator 
Audio
OnBoard 
  hide details  
Reply
Destrto3
(17 items)
 
  
CPUMotherboardGraphicsRAM
AMD FX-8120 M5A99FX PRO R2.0 Radeon R9 270X Kingston HyperX Blu 8GB (2 x 4GB) 240-Pin DDR3 ... 
RAMHard DriveHard DriveHard Drive
Kingston HyperX Blu 8GB (2 x 4GB) 240-Pin DDR3 ... SanDisk Extreme 240 Western Digital WDC WD10 EARS 1TB Seagate ST2000DM 2TB 
Optical DriveCoolingOSMonitor
TSSTcorp CDDVDW SH-222BB EK Supreme LTx (AMD) Windows 8.1 Sceptre 32" LCD Flat Panel TV 
KeyboardPowerCaseMouse
Cooler Master Devastator ThermalTake TR2 RX 850W Corsair Obsidian 750D Cooler Master Devastator 
Audio
OnBoard 
  hide details  
Reply
post #680 of 4087
@ComputerRestore - I'm definitely not claiming expertise. Most of my choices were trial-and-error, and I can only confirm that my results worked for my particular case. Additionally, I have no idea about "binning", or whether I have "flaky" parts or "solid" parts...(Also, I don't know how to quote...)

To get to your questions...I chose DRAM Current of 130% because I inferred--though this isn't based on any evidence (because I'm not a get-out-the-multimeter-and-start-measuring-pins-on-the-mobo kind of guy)--that just like CPU Current, upping the current through the memory doesn't actually cause it to use more current; but simply allows more current to flow.

Most of the posts in this thread talk about 2 sticks. I think DEUSXXX's post that you linked to in the guide was the only one that stuck out for me as someone using 4 sticks (though, that may just be bad analog memory on my part). Since I'm already busy making assumption (and asses), I just sorta thought, based on shaky high school physics first principles, that more sticks probably means "more juice needed". Meaning, since I was using a fixed voltage setting for the DRAM, I should probably allow more current. If I had to guess, (making more asses by the minute), I'd say that the 130% DRAM Current Capability is probably helping more than my DRAM voltage setting. Astute high-schoolers will realize at this point that I failed to factor in "resistance" in my shoddy guesswork. And they'd be right, because I just have no idea if these setting apply to a single slot, or to the entire DRAM subsystem (assuming that's a thing), or how RAM interacts with the rest of the system electronically.

The reason I suspected memory in the first place (again, empirically, and not based on a sound theoretical understanding of how these things work) is that on the same OC profile, Prime95 Small FFT always worked (little-to-no memory accesses)--with very stable voltages and temps--but Blend (75% mem, which for me meant typing in '24576') was causing a lot of voltage fluctuations. So, I figured that since my sticks weren't at all warm (I've a fan blowing right onto them, so they stay pretty chilly--at least their heatsinks do) I could allow more current. Of course, I have no way of knowing if that change actually affected anything...

I had also read somewhere that the the M5A99X (and similar boards--I assumed they were referring to chipset) didn't really deal well with 4 sticks. I assumed that meant that it didn't allow high mem overclocks. I wanted them at 1600 (despite the SPD config), but I didn't push them past what the box said they could do.

I ended up at manually setting them (4 sticks of 8 GB) to 1600 MHz at 11-11-11-30-2T, with 1.4 V. At a later date, I may try to "tighten up the timings", but I suspect (based on prior failures at 9-9-9-24-1T) that I don't have much leeway here. Even with those settings, I still get Vcore fluctuations using Blend. But, the tests are passing, so I'm living with it...

As for the DRAM voltage...I chose 1.4 V as a "subjective compromise" between what DEUSXXX said about lowering DRAM voltage to reduce stress on the IMC "to improve stability" (which is probably being stressed pretty hard during the Blend-75% test) and going the other way (upping voltages) to...well, frankly, I don't know what a higher voltage potential for RAM does, other than assuming it was for analogous reasons to upping CPU voltage--which I would guess also improves stability. Odd that it could go both ways with DRAM.

The short answer is, I chose 1.4 V because of the potential "reduced IMC stress". Though, now that you've asked, I might try increasing the voltage--at least up to 1.65. Does anyone know if a DRAM voltage setting that's too low can cause CPU voltage ("Vcore", according to OCCT) to swing? I honestly don't know if upping it will "improve" things (i.e., allow a higher overclock for when I'm ready to try that, if ever), or even if lowering it to 1.4 is actually helping.

So, to the extent that I can add any data to this thread, it would be that for my memory, the 4x8GB sticks, my timings allow my setup to work. Now, having said that, the Corsair site lists this memory as "Dual Channel": Corsair XMS3 CMX16GX3M2A1600C11. I'm not sure what this means, and I'm worried that it might mean that 4 sticks of this stuff were never meant to be used together. I base that off the specs from a better Corsair product: Corsair Vengence Extreme, which explicitly state "Dual / Quad Channel". It makes me suspect that there might be some difference. (Info under "Tech Specs".)

@DEUSXXX - I'm pretty confused by your conclusion that running DRAM at 1.5 V (instead of 1.65 V) causes less power to go through the IMC. I'm definitely not saying you're wrong; I'm just trying to figure out 1) how you measured that, and 2) if this means that DRAM voltage is somehow "pushes" through to the CPU/NB (since you speculated that lower DRAM voltages might stabilize the IMC).

I'm just having a hard time seeing why DRAM voltage would affect the IMC. From my admittedly sketchy guesses, it would seem that simply using 4 sticks--as you were and as I current am--is what's increasing the current to the IMC (more stuff to control, more current flowing). When you measured the power through the IMC, was after changing settings? Or dropping down to 2 sticks?

And, you end with this:
Quote:
Most people using 4 ram nodules reduce their ram speed or cpu/nb speed to have acceptable cpu/nb voltage whereas reducing ram voltage (when possible of course) could give better yields/stability.
What did you mean when you said "to have acceptable CPU/NB voltage"? Do you mean that increasing RAM speed requires a concomitant increase to CPU/NB voltage, so it was necessary to drop RAM speed to keep CPU/NB voltages from hurting the OC (by increasing the thermal load)...?

As for my own problem...To the extent my "problem" is these Vcore swings, I'm not sure which part of the chain (CPU/NB/IMC/DRAM) should be adjusted--or if the Vcore swings are even an issue (the system produces correct results, but maybe I'm frying my CPU/RAM?).

So, maybe going down to a DRAM Voltage of 1.4 V was unnecessary...I can see spending the next long-weekend doing some more tests. mad.gif I'm still interested to see if it's possible to somehow get my Vcore to be flat under Prime95 Blend load. It takes so long to figure out which knobs to turn and now to turn them...No conclusions here, yet.

Which is all to say...I realize I'm probably confusing many issues here, so thank you all for all the work you've put into this thread to help a nubsicle like me get even a modest overclock!
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: AMD CPUs
Overclock.net › Forums › AMD › AMD CPUs › AMD Bulldozer and Piledriver Overclocking Guide - Asus Motherboard