970 Performance FX-9000 & RAM incompatibility
Printed From: ASRock.com
Category: Technical Support
Forum Name: AMD Motherboards
Forum Description: Question about ASRock AMD motherboards
URL: https://forum.asrock.com/forum_posts.asp?TID=660
Printed Date: 05 Oct 2024 at 6:51am
Software Version: Web Wiz Forums 12.04 - http://www.webwizforums.com
Topic: 970 Performance FX-9000 & RAM incompatibility
Posted By: WKjun
Subject: 970 Performance FX-9000 & RAM incompatibility
Date Posted: 17 Aug 2015 at 7:02am
|
Hello! I'm used to the FX-9000 series and use Noctua NH-D14/15 coolers without problems on Sabertooth 990FX R2.0 motherboards. I always use AMD Radeon RAM 2133 and 2400(@2133 AMP profile, CMD 1T). Sadly I have two problems with my 970 Performance board. First, at default(!) settings the FX-9590 straines the board so hard that mine began to burn! I have a picture of the chip (not uploaded until needed) which sits between the ATX12V plug and the screw of the CPU socket, which got so hot that the system shut down and damaged the CPU as well. To make it clear, CPU cooling itself was flawlessly applied and the CPU never reached 60° under any circumstances! Second, AMD Radeon RAM seems to have a compatibility issue although the RAM controller is inside the CPU, works on the Sabertooth, and is listed in the QVL. The history: I bought the 970 Performance in March. After some time it got more and more unstable. Fist just freezes at idle and sometimes at GTA V and Dota 2. Then sudden bluescreens and even more often game crashes. Then, in July, it shutted off and no way to turn it on again (cut power for some time, etc.). Then I've discovered a burned chip besides the CPU! I switched it through RMA / the store where I bought it. Then, after receiving a new board (it states Made in Vietnam now instead of China, strange), I discovered that the CPU has been damaged as well (just running fans without getting a POST screen or any further reaction) and returned that one too. After RMA: I have updated to the latest BIOS, even though it states just M2 compatibility improvement. Then I have looked deeper into the BIOS settings and found that the "CPU Voltage Offset" was at "+50mV" by default! This is insane and useless while using a FX-9590 which has already a very high voltage to reach 5 GHz turbo! I'm not going to overclock it and now I find a power wasting setting and maybe the reason why my first board began to burn. To check the uselessness I have used prime95 for a short period of time and HWmonitor to see what temperature the VRM gets. Besides the hopelessly overstrained heatsink even at default settings, here are the results: +0mV (manual): VRM 85-90° | CPU ~55° +50mV (default): 90-95° | CPU <=60° +100mV (manual): >100° (would have gone beyond that if not interrupted at that point) | CPU >60° +150mV (not tested...) At +50mV and reaching 90-95° VRM for some seconds, the board begins to make noise around the VRM, like sizzeling! It stops doing that after quitting prime95. At +0mV this never happend and there were no stability issues at prime or Windows itself. So why set +50mV by default?! This board is obviously wrong configured and not suitable to be used with a FX-9000 processor in my opinion! Mind that my Define R4 case is open on both sides - all case fans running though - I can imagine what happend inside the closed case. The RAM incompatiblity is the next thing. QVL states, it is supported (CPU and RAM both AMD brand anyway) and it works on a Sabertooth that does not have it on QVL... The system is only stable when I set fully Auto without AMP and that results in 800 MHz... if I set 1866 or 2133 MHz, prime is not stable for longer than some minutes. Switching to the black slots instead of the red ones, didn't help. Further, SPD readout inside the BIOS is wrong for some values (tRFC 1 instead of ~300). So what I am going to do? I am not willing to experiment with undervolting to get a trustworthy system and buy some other RAM kits that might work!? My store would hardly let me switch to a Sabertooth instead and return it in used condition... Any help is appreciated, hopefully a BIOS update that fixes these issues!
------------- PC1: FX-9590@def|290 |16GB@2133|Sabertooth 990FX R2.0 PC2: FX-8320@4.5|290 |16GB@2133| " PC3: FX-9590@def|280X|16GB@2133| " PC4: FX-9370@def|280X|16GB@2133| " PC5: FX-6300@4.6|7950|16GB@1866|990FX-UD3 |
Replies:
Posted By: parsec
Date Posted: 17 Aug 2015 at 2:28pm
If you look at the CPU Support list for the 970 Performance, you'll see this warning for all of the FX-8000 and FX-9000 series processors: For cooling the CPU and its surrounding components, please install a CPU cooler with a top-down blowing design. That means a CPU cooler that has a fan pushing air down onto the board, so air flows over the VRM heat sink and VRMs. A Noctua CPU cooler of that type is this model: http://www.noctua.at/inc/imageviewer.php?item=111&pnr=0" rel="nofollow - http://www.noctua.at/inc/imageviewer.php?item=111&pnr=0 Your Noctua cooler is a great CPU cooler, I use one myself, but it is not a "top-down blowing" design. It would be very hard to find a top-down blowing type CPU cooler that cooled a CPU as good as your Noctua NH-D14/15 does. You know what the problem is, over heating of the VRM chips. Also, the use of an eight phase VRM design must have extra good cooling of the VRM heat sink to control any over heating. A 12 phase VRM design will operate cooler than an eight phase design, but still needs air moving across the VRM heat sink when using FX-9000 series processors. The sizzling you heard was probably the heat sink pad used between the VRM chips and the heat sink, or the thermal paste if that was used. Your board and almost every AMD board needs CPU VRM cooling when using a FX-9000 series CPU. Randomly mounted case fans are generally not enough. The top, rear fan mounting location found on most PC cases should have a fan blowing air into the PC case, close to the VRMs. The case must have fans on the top pulling air out of the case. Somehow mounting a fan to push air over the VRM heat sink is required IMO. Video cards that exhaust air into the PC case will make cooling the VRMs and the CPU more difficult. About the CPU Voltage Offset of 50mv, was that using a built in CPU over clock option, or just the default BIOS settings for the FX-9590? Either way, that CPU Voltage Offset might have been chosen so any and all FX-9590 processors would be able to work at 5GHz, since all processors are not the same. About the RAM compatibility situation, the Memory Support list for your board has this warning: When overclocking, some AMD CPU models may not support DDR3 1600 MHz or higher frequency DIMMs. Did the Sabertooth board run your memory above 1600? One thing I think we should remember (admit?) is the FX-9000 series processors need more power from a mother board to run at 5GHz than anyone realized. I think mother board manufactures were surprised by this, and discovered that some of their boards they thought would work fine with the FX-9000 series processors, were not good enough to supply to them all the power they need. I've heard that some AMD boards are getting BIOS updates that won't let them boot if an FX-9000 series processor is used in the board! That is how bad the situation is with these processors has become. AMD's "most everything is compatible" designs have reached a hard limit with FX-9000 series processors, and for some low end board, that includes the FX-8000 series processors too IMO. I don't know if your board will get a BIOS update to fix the issues you had with the memory, or to change the default CPU Voltage Offset default value. ------------- http://valid.x86.fr/48rujh" rel="nofollow"> 
|
WKjun wrote:Posted By: WKjun
Date Posted: 17 Aug 2015 at 7:05pm
|
Hello & thanks in advance that your participation! I admit the Noctua NH-D14/15 ist not a top-down cooler per se. But its unique design combines the advantages of both, top-down and tower coolers. As you know, the middle fan's protrusion offers "Excellent component cooling"  : http://noctua.at/main.php?show=productview&products_id=34&lng=en" rel="nofollow - http://noctua.at/main.php?show=productview&products_id=34&lng=en Therefore I think and have the experience that this model cools the CPU and surrounding components better than any other air cooler. Which air cooler would suffice, if not this one? My Define case is not that small either and I have no problems running six other FX-CPUs even overclocked stable. At this point I have to say that if the VRM design and its cooling solution is that tenuous, ASRock should never have approved FX-9000 support! Even 125W chips in that list have the top-blow warning, how could it handle 220W? Even as an aware consumer I find it infeasible to let me try and find out ways to have a durable system without blowing something out - on my other board it works that way, plug & play. Furthermore the hand picked FX-9000 models are able to operate at reasonable voltage levels and might not even use up the 220W TDP margin. Overclocking a FX-8000 (to that level) would lead to much more heat and power needs. Even if these chips have been underestimated power-wise, two years after its release, mainboard manufacturers have to be aware of that! I would have admitted that for models available back then, but not for a 2015 release! The voltage offset is by default +50mV! And there is definetly no reason to overvolt the CPU additionally by a mainboard manufacturer to reach 4.7 GHz base clock and 5 GHz turbo clock. Turbo works automatically (on just one core), as you might know, if certain conditions apply and APM (Application Power Management) looks after that. So no (CPU) overheating could occure, if the chip is fully loaded or too hot. By raising voltages (and temperature) it would even hinder reaching turbo and any CPU has the right voltage information built-in! The Sabertooth has a better, cooler running VRM design and a better cooling solution mounted (same config leads to <75° for the VRMs). But importantly its BIOS has more options, like Phase Change for VRM - an extreme setting for OC - where a warning is visible to actively cool the VRMs at non-default(!) settings and such an option is missing in the 970 P. BIOS. Maybe this setting is misplaced for the FX-9590 without knowledge about its function and creates unnecessary heat and strain?! About the RAM speed I can tell you this: The FX-9590 is not overclocked in a common sense. OC is something a user would apply. 1866 MHz is officially supported by AMD on any of these CPUs, as long as you have 2 modules. 4 modules *could* lead to instability at full speed, so 1600 MHz is advisable. I just use 2 modules and the same modules work at 2133 MHz on my Sabertooth. And even 4 modules work at 2133 MHz there...
|
Posted By: WKjun
Date Posted: 20 Aug 2015 at 5:12pm
|
Yesterday I've tried another RAM kit: Crucial Ballistix DDR3-1866 (2x8GB) with XMP. Independet of activating the XMP profile, it just works at 1333 MHz or below. If I select 1600 or above, the system hangs with a black screen. I have to turn it off and on again, wait some time until the BIOS recovers with default settings. This is annoying, because both, the AMD Radeon and Crucial Ballistix Kits, run flawlessly on my Sabertooth boards. And why shouldn't it work, the memory controller sits within the CPU! I could live with the boisterous voltage settings and undervolt the system, but buying another RAM kit is out of the question. What could I have done wrong? Should I open an official support ticket?
|
Posted By: Xaltar
Date Posted: 20 Aug 2015 at 11:51pm
| What voltage are the RAM kits rated for? It could be they are 1.65v kits and are only able to run at 1333 at the default 1.5v the board uses. If they are 1.65v kits then you will have to either set RAM voltage to auto or manually set it to 1.6 or 1.65 depending on their rating. |
Posted By: parsec
Date Posted: 21 Aug 2015 at 12:12am
|
I don't know exactly why your memory will not run above 1333 on the ASRock board, except for the warning in the Memory Support list: When overclocking, some AMD CPU models may not support DDR3 1600 MHz or higher frequency DIMMs. Yes I know you said your memory worked on the SaberTooth board. There are BIOS settings that are not available to users that might be making the difference. Were the memory controller voltages the same as the SaberTooth on the 970 Performance? Or memory voltage? The VRM option on the SaberTooth is an Asus only option, not a standard option that is common to all AMD 970 boards. I use a Noctua DH-14 socket 2011 version, and it's not that great for VRM cooling. Different boards are... different. So is their price. ------------- http://valid.x86.fr/48rujh" rel="nofollow">
|
Posted By: WKjun
Date Posted: 21 Aug 2015 at 4:54pm
|
Hello! This Crucial kit is rated for 1.5V at 1866 MHz (while the AMD kit needs 1.65V for 2133/2400 MHz) and runs at that setting on the other mainboard. Your words, parsec, about the memory controller's voltage gives me the idea of a possibly wrong voltage setting for the CPU-NB! If it hasn't enough juice it would react like this. I'm going to check this out soon! Thanks for your input!  PS: Which top-blowing cooler would ASRock (or you) recommend? Have you relocated the fan in the middle of your NH-D14 as near as possible down to the socket? In theory this would hinder airflow between the backside fan and middle fan to a small degree (never benched it) but it gives the VRM a better airflow from downside. |
Posted By: WKjun
Date Posted: 24 Aug 2015 at 5:43pm
|
I was mislead (again) by several factors: .) DIMM slots' channel linkage seems to be the other way round than usual .) Switching "Overclock Mode" from Auto to Manual kills RAM auto detection and stability at high frequencies .) System is unstable in Turbo mode This is what happended: I've switched back to the original two 8GB Radeon DDR3-2400 modules. BIOS is at default values (C6 setting is Disabled, therefore Turbo inactive!). RAM is automatically set to 1866 MHz (CL13) and 1.65V. If I select XMP, it runs at 2400 MHz (11-12-12) - not benched, but POST is alright If I select AMP it runs at 2133 MHz (11-12-12; some sub-timings a bit faster) - Prime-stable If I select AMP and tweak timings to 10-11-11 (taken over from the 2133 MHz Kit) - Prime-stable too! Another strange thing is the DIMM slot linkage and "corresponing" Command Rate timing. Usually channel A (red slots, nearer to CPU) should be filled first, prior channel B (black slots). If channel B is filled first, it reacts like 4 filled slots, which usually leads to a bit slower Command Rate of 2T. In my case, when using channel A, the Command Rate was 2T automatically and no POST if 1T was selected manually. In the black slots, which are supposed to be channel B, 1T was set automatically and works stably! So far so good! The next phenomenon is the "Overclock Mode". If I change it from Auto to Manual (leaving reference clock at 200), RAM is automatically set to 800 MHz and values above 1600 MHz are unstable or not bootable, as described initially. So I have to leave it at Auto. Although it is not my intention to OC this system, it degrades future possibilities. [EDIT] With +50mV Offset (default value) Prime SEEMS to run for a longer time, but freezes in the end (10 min. or so). I'm going to try more Vcore instead/additionally to Offset Voltage now... [EDIT 2] It's no use. Even if I use +100mV Offset as well as 1.525V Vcore instead of 1.500V it keeps freezing. Then I've tried to reset BIOS to defaults and just activate C6 (& Turbo). No difference. RAM was at 1866 MHz. I didn't bother to try 800 MHz... So voltage is not the issue as previously suspected. Without Turbo (& C6) and +0mV Offset, Prime ran 7 house continously. |
Posted By: WKjun
Date Posted: 28 Sep 2015 at 8:52pm
I've reported the issues (seperated into two cases) to ASRock ( http://event.asrock.com/tsd.asp" rel="nofollow - http://event.asrock.com/tsd.asp - German language) three weeks ago! No reaction, except for two confirmation mails! How long does it usually take?! Or is (German) support non-existent? 
|
Posted By: WKjun
Date Posted: 07 Jan 2016 at 7:57pm
|
To close this matter unsuccessful: ASRock actually did answer to one of my tickets some weeks after my last post here, offering me to send it in, but by then my goal was to get rid of it. So I've returned the mainboard to my retailer for the second time. Waiting about 7 weeks for a decision of its wholesaler, I've got my money back.
------------- PC1: FX-9590@def|290 |16GB@2133|Sabertooth 990FX R2.0 PC2: FX-8320@4.5|290 |16GB@2133| " PC3: FX-9590@def|280X|16GB@2133| " PC4: FX-9370@def|280X|16GB@2133| " PC5: FX-6300@4.6|7950|16GB@1866|990FX-UD3 |
Posted By: WKjun
Date Posted: 19 Jan 2016 at 8:20pm
Feeling like a guinea pig, I'm afraid I have to persist on the removal of the FX-9000 support tag for this board, for the sake of all future customers having the same trouble with it and come into defective boards! What a waste of time...
 ------------- PC1: FX-9590@def|290 |16GB@2133|Sabertooth 990FX R2.0 PC2: FX-8320@4.5|290 |16GB@2133| " PC3: FX-9590@def|280X|16GB@2133| " PC4: FX-9370@def|280X|16GB@2133| " PC5: FX-6300@4.6|7950|16GB@1866|990FX-UD3 |
wardog wrote:
Posted By: Nonorg
Date Posted: 29 May 2016 at 9:11pm
|
Just my 2 cents to anyone still trying with this board: - I can't speak for the FX-9590, but the FX-9370 it's definitely not a 220W processor - except at BIOS defaults it will likely be. - The +50mv offset is being misinterpreted. The only thing it does is to mitigate the (insane) voltage drop this board experiences when switching from idle to full load. The only reason OP experienced higher temperatures when increasing the offset is that the effective voltage provided to the CPU was increasing (since vdroop was reduced). - Hence, the true problem is the 1.535 default vcore that leads to the 220W rating. This ensures the FX-9370 will work stably (provided heat is taken care of), but it is way overkill. Therefore, the solution for those set on using the FX-9370 on the 970 Performance is: undervolt that CPU! Your mileage may vary, but in my experience you can drastically reduce vcore without experiencing stability issues, and highly improving the thermal environment. In order to do that, you need to remember that vcore or cpu offset themselves don't drive the temperature under load: what really matters is the actual voltage delivered to the CPU by your vocore/offset combination. In my experience, temperatures are a bit better if the effective vcore is stable rather than bouncing around (which, unfortunately, is very common with this board). Hence, I recommend lowering the vcore and trying which combination of vcore+offset works better for you (as a general rule, you need to combine higher offsets with lower vcores to achieve the same effective vcore during load). Another thing to take into account: I found that RAM speed is also an important determinant of VRM/CPU socket temperature and vcore stability (I guess it's due to the IMC). Now, my results so far: - With +0mv offset, I could get the FX-9370 to work with 1.4125 vcore. Effective vcore was of the order of 1.32v, and not very stable. 1.4v would now be stable. - With +50mv offset, I could go to 1.3875 vcore. That would translate to a quite stable 1.33-1.34v delivered to the CPU at load. My temps for VRAM (measured with IR gun and a temp sensor attached to the back of the MB) were lowest with this setting, and CPU socket sensor would barely approach 55C. - It seems I could go lower, although delivered voltage would become less stable, and no temperature improvement was observed. However, at this point I realized RAM was defaulting to 800Mhz. So, time fix that. - I was using G.skill sticks rated at 1866 MHz, so I aimed for 1866. Temperatures went higher and delivered voltage less stable as I did this, and it would become unstable before 1866. However, I didn't need to adust vcore: increasing CPU NB (i.e, the IMC) voltage one notch to 1.15v was enough to get everything to run stable. Now the effective voltage dances around more, the VRM temperature (back of the MB) is in the low 70s, and the CPU socket temp reaches 66C under load. - As a side note, core temps barely reached 30C (~22C ambient) at 800MHz, and never passed 41C at 1866MHz. For reference, I'm using OCCT to measure actual voltage delivery, Corsair Link software to track temperature sensors, and an IR temperature sensor to calibrate/validate the outcome of the non-built in sensors. I'm not surprised by these results, since no one overclocking an FX-8XXX ever needed 1.53v to reach 4.4GHz (and 8x4.4GHz is the maximum power consumption and temperature, not 1x4.7GHz, which is the FX-9370 turbo mode).
|
Posted By: WKjun
Date Posted: 31 May 2016 at 9:10pm
|
Thank you for your in-depth contribution - I was looking forward to that! Do you mean the FX-9370 is not a "220W processor" because you have managed a successful undervolt? Actually its TDP is 220W, which means it does not necessarily need 220W, but peaks there eventually. I bet it uses even more than 220W if used on a 970 Performance at BIOS defaults, because voltage is so high (1.535V), like overclocked. Apart from this, AM3+ FX CPUs are known for rather ineffectual undervolting results and high overall power consumption. Congratulations you've managed and hopefully it lasts! To keep within TDP, AMD uses its LLC (Load Line Calibration) technology to lower voltage the more load it gets. That is why it is possible to keep it within TDP specs and/or 60° C. This should be the voltage drop you have experienced. So no Offset should be needed. Was it unstable that way? It definetly shouldn't be. Btw. this board lead me more than once to wrong conclusions. Another AMD thing is APM (AMD Power Management). You will notice that at full load (like Prime95) the clock rates begin to fluctuate between say 4400 and 3500 MHz. This allows for even lower voltages at full load, while maintaining a certain performance level. If LLC and APM is active, there should be no need of extra cooling components like VRM - other boards prove that. It's an imposition anyway if users have to experiment with undervolting to get it right. What about users that are not capable as you are? A decent case like Fratal Design should be fine without extra VRM cooling to manage FX-9000 chips - mine do. But the real issue here is, as you have discoverd too, wrong default BIOS settings that lead to improper voltages used by the CPU! I don't remember seeing LLC in there... is Offset ASrock's interpretation of LLC? It should be there and it should be set to Regular, not Medium, High, Extreme or Auto to keep everything cool! However, I still can't see the difference between your and my understanding of Offset Voltage. The drop is desired by AMD at full load and if not implemented right, you get a too hot CPU and way too hot VRM. ------------- PC1: FX-9590@def|290 |16GB@2133|Sabertooth 990FX R2.0 PC2: FX-8320@4.5|290 |16GB@2133| " PC3: FX-9590@def|280X|16GB@2133| " PC4: FX-9370@def|280X|16GB@2133| " PC5: FX-6300@4.6|7950|16GB@1866|990FX-UD3 |