AsRock Fatality 990FX (CPU Throttling)
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=1711
Printed Date: 05 May 2024 at 7:13am
Software Version: Web Wiz Forums 12.04 - http://www.webwizforums.com
Topic: AsRock Fatality 990FX (CPU Throttling)
Posted By: Atomic
Subject: AsRock Fatality 990FX (CPU Throttling)
Date Posted: 11 Jan 2016 at 11:06pm
|
Hi Guys, (I'm from Australia but hopefully I can still post here and maybe someone can help me). I have built a new system around your AMD 990 FX Fatality Chipset. The motherboard looks good with all the features you have on there. However my issue has been with the AMD FX-6300 (95 Watt) CPU I have noticed using programs like CPU-Z and Windows 10 "Task Manager". Under heavy load using 3D Mark or FurMark to test stability of the system. The multiplier on the CPU fluctuates from between 7 and 22.5 times on the CPU bus speed multiplier. I understand that with the AMD cool n quiet and other BIOS features the mother board throttles and tries to conserve the power. However when the CPU is under full load the speed of the CPU changes from 1,000 Mhz to 3,700 Mhz. (This shouldn't happen under stress load...it should stay the same). Also in further testing I have also disabled the AMD Cool and Quiet, AMD Power Saving modes etc....and retested and the same issue still appears. Is there a hidden screen in the BIOS where I change the Load Line Control or to disable APM? On another note I have 2 x 8 Gb sticks of Corsair Vengence Pro 2400 RAM which I can't get to boot at the nominated speed. The memory needs to be down clocked to around 1,000Mhz...which I think is a joke. It says your motherboard supports DDR3 2450 OC RAM even. I am also Interested in Over-clocking the AMD CPU. I bought the Fatality motherboard because I believed this was the best one for AMD chips on the market. What is the best way for me to push a safe and stable over-clock without causing instability? I would also like to know if there is a new BIOS which will be released for this motherboard soon? (Will their be any performance or over-clocking enhancements for this new upcoming BIOS revision?) |
Replies:
Posted By: WKjun
Date Posted: 12 Jan 2016 at 6:34pm
Hello!  If Cool'n'Quiet is active, it saves energy by lowering the multiplier / clock frequency if not under load. Most processors nowadays additionally have a Turbo Mode. It raises the clock frequencies of some cores above the basic frequency for a certain time, if certain conditions take place, like power consumption, temperature and load. The FX-6300 has a base clock of 3.5 GHz, while Turbo is 4.1 GHz. Idle clock (cool'n'quiet) is somewhere between 1.0 and 1.5 GHz. Keep in mind, that there may be several steps between them. Turbo may get active for just one or two cores. If fully loaded (i.e. prime95) at all 6 cores, the frequency cannot exceed 3.5 GHz and further might drop to around 3.0 GHz for all cores, to stay within power limits and/or 60° C. This safety function can (but shouldn't) be deactivated by disabling APM (AMD Power Management). I think, there is no practical use for that, because no application would utilize or need 6 cores fully loaded. The system works best / as intended if:
is ENABLED! (AUTO sometimes does not enable it) What do you mean by: "However when the CPU is under full load the speed of the CPU changes from 1,000 Mhz to 3,700 Mhz. (This shouldn't happen under stress load...it should stay the same)." As written above, loading all cores fully is good for stresstesting it, but has not practical use and causes the core frequencies to drop. I don't know of any hidden screens. There are some on Gigabyte boards, though. Usually no maker gives hints of upcoming BIOS updates, sorry. Have you looked at the product site for updates? Officially the CPU supports DDR3-1866. I personally never had problems with DDR3-2133, at least not with AMD Radeon and Crucial/Micron RAM. To go obove 2133, or theoretically 2200 MHz, you have to OC the CPU-NB (CPU-Northbridge, the integrated memory controller) to at least that frequency. 2400 in your case. Getting that stable is tricky and not worth the trouble. Even a prime-stable CPU-NB OC can cause strange effects in Windows or games. Either way, the performance of FX CPUs does not scale well with memory clock. If you have 1866 or 2133, go with it. The only thing you may optimize are memory timings. Yours should be CL11-13-13-31, if I am correct. Auto mode detects it correctly. So, using them at lower frequencies instead of 2400, the timings can be detected/set lower. Good modules have memory profiles programmed in (SPD) and may be accessed through XMP / AMP option in BIOS. If just the 2400 profile shows up, but you want to use 2133, then consult CPU-Z or AMD Overdrive for additional RAM timing tables for 2133, so you can set the values manually. I suspect them to be CL10-12-12-28 or so. However, there is never a guarantee that certain modules work with your MB and CPU. Consult the QVL (Qualified Vendors List) for this board, to see if your modules were tested. I generally don't recommend modules with such oversized and useless heatsinks that prevents you from using most bigger CPU coolers... As for that Fatality branding, come on, it's just a marketing gimmick... although there is no reason to think this was not a good choice. But I don't have it and can't tell. If you are an absolute beginner in the OC field, this is a completely different matter and exceeds my time, sorry. But feel free to look around the Web for some the many FX OC guides!  Ask again, if something is unclear.
|
Posted By: Atomic
Date Posted: 12 Jan 2016 at 6:41pm
|
Hi WKjun, CPU Cooling: The heat sink I have installed onto the AMD CPU is an after
market cooler from Deep Cool. This is a sealed water cooling unit. It is making good contact on the CPU and is tightly held on. So I don?™t see this being a problem. I also have several case fans inside which are pushing fresh
air inside to cool the components. AMD CPU FX-6300 (Vishera 95 Watt): Does NOT turbo boost to
from idle 3.5Ghz to 4.1Ghz as it should be: I reset the UEFI BIOS back to factory default when I run a
benchmark like FURMARK or 3D MARK. The CPU hovers between 3.5 and 3.8GHz and is constantly
fluctuating even when the CPU is put under load. Now my understanding is when the CPU is under load then the
CPU will boost to 4.1Ghz and stay there until the stress testing is finished. Then when idle or if you quit the program then it will
throttle down to 3.5Ghz...none of this is happening. The multiplier is switching from RAM Corsair 2400 Vengence Pro DDR3: After having reset the BIOS back to FACTORY DEFAULT. The RAM automatically detects it?™s speed and runs it at
maximum. I have found the motherboard automatically runs it at only
1300Mhz and not the 2400Mhz that it should be. If I try to run the RAM manually to 2400 then the system
doesn?™t boot and fails to give me a screen. If I run the RAM at a slower speed like 2000 or 1866 then it
boots and seems to be OK. Also to let you know that when I select the XMP 1 for the
RAM modules the motherboard doesn?™t boot after saving the changes. Overclocking: Now I have tried to manually do some overclocking on the
motherboard as is provided in the UEFI BIOS: I disabled Cool and Quiet, AMD Power Saving, Spread
Spectrum, Thermal Throttling etc. Then I boot into windows and look at a program called CPU-Z
for example and I can see the multiplier changing between 17 to 19 times
multiplier for the Core CPU Clock of 200Mhz. I?™ve noticed also that the CPU Core Voltage is alternating
between 2.5v and 3.0v. Looks like the North Bridge chipset on the motherboard isn?™t
providing a stable current to the CPU which is causing the CPU frequency to not
remain %100 stable. Even after I turn off all the power saving features and
anything else that might change the CPU clock speed or multiplier. The CPU performs erratically. I have bought a Corsair 1000W PSU so power wouldn?™t be the
issue here. I am at a loss and I don?™t know what to do. I feel it?™s the motherboard NOT providing enough stable
current to the CPU and the RAM not being recognised properly which is causing
issues and major headaches for me. I am honestly feeling very frustrated after all this and I
don?™t know what else to do. I look forward to your reply. |
Posted By: WKjun
Date Posted: 12 Jan 2016 at 8:13pm
|
Hi Atomic! Okay, now I understand your frequency troubles. To me it sounds like either a BIOS problem (bug or not the latest one) or a monitoring issue (CPU-Z), if your CPU cooling performance is good enough. I know, water cooler sounds good, but I've gone through a misery with my one some time ago. It was supposed to be the best one and was 15° warmer than my, and I think the best air cooler, Noctua NH-D14(/D15). The next disadvantage is, that the Chipset and VRMs (voltage regulator modules near the CPU socket) are not being cooled passively, like through a regular top-blow cooler. Case fans by itself are not sufficient cooling for hot mainboard VRMs! Your 95W TDP CPU is not that critical, but I don't know of the realization of the cooling solution build into this board and its longevity. Bottom line, I scrapped my water cooler. But yours may be better now! However, as long as you don't know your temperatures, it may or may not be the cause. Please try HWMonitor ( http://www.cpuid.com/downloads/hwmonitor/hwmonitor_1.28.zip" rel="nofollow - http://www.cpuid.com/downloads/hwmonitor/hwmonitor_1.28.zip ) to doublecheck the frequencies seen at CPU-Z and have a look at your CPU temperature ("core temp" favored to "package temp", names may vary). You have a comfortable view there that shows the highest and lowest frequency measured per core. If no app is running in background, turbo can be triggered easily by running prime95 @ 1 or 2 threads for some seconds while monitoring with HWMonitor. If the CPU temperature is too high, it may cause a turbo frequency lower than expected or none at all. Lots of boards have some necessary features deactivated at default. Enable the settings I've suggested in my last post to ensure Turbo can run reliably. To make one thing clear about the Turbo mode: It just works within these certain parameters and its occurrence is rather underwhelming. There is not much to be gained by using it. Especially if 3 or more of your cores are being utilized, it may almost never be triggered. Furthermore, Windows loadbalances tasks to all cores. So, if you do not lock an app to one core (i.e. Task Manager) then the app jumps through the cores. Turbo may be triggered and disabled fast than CPU-Z can monitor in that case. Therefore I recommend HWMonitor instead. About the XMP profile, it is just as I thought. You get one profile read out and that's the 2400 one which does not work, because either the CPU-NB does not get OC'ed correctly or the board is just not compatible to your modules and speed in combination. To be honest, I had a painstaking journey with the 970 Performance (also "Fatality") board, defect-wise and RAM-setting-wise, and gave it back. There is a detailed thread in this forum of mine about that. Setting up the RAM was just horrible. However, stay at 2133 and lower timings manually to my guessed values above. Remember to raise DRAM Voltage from 1.5 to 1.65V, as the 2400 profile would do. Then run prime for some hours in blend mode (2 threads are sufficient) to ensure system stability. If you like, read through my thread of misery with my board, partly setting up RAM correctly. http://forum.asrock.com/forum_posts.asp?TID=660&title=970-performance-fx9590-overheating-ram-incomp" rel="nofollow - http://forum.asrock.com/forum_posts.asp?TID=660&title=970-performance-fx9590-overheating-ram-incomp If you really have the latest official ASRock BIOS installed, you may write an official ASRock Support Ticket. But don't expect an answer within a month or so...  As for the OC: If the LLC (Load Line Calibration) value is set incorrectly (manually or automatically), it causes erratic voltage and temperature fluctuations. You can never trust automatic settings, but leave settings alone (on Auto) if you don't understand them. That's my wisdom for you. Be warned, that wrong settings may end in a dead system and worse, loss of warranty! I know, you have to start somewhere if you want to learn about OC'ing, but it is risky. By the way, current / its stability is not provided by the chipset but by the VRMs and power supply. Further, it is impossible that you have 2.5-3.0V lying on your CPU Core. Anything above 1.45V is critical! 1.6V can be named deadly, so please be cautious if you have not the skills or tools to do OC right. No offense, I just want your system to live on! :)
|
Posted By: wardog
Date Posted: 12 Jan 2016 at 9:17pm
|
Atomic, what you're seeing Turbo MHz wise is just how the 6300 is designed to function. There are two "steps' it will employ while in Turbo mode. Level 2 is 3800MHz on all cores. Level 1 is 4100MHz on only two cores. Prime loads all cores so 3800MHz is correct for running all 6 cores. http://www.cpu-world.com/CPUs/Bulldozer/AMD-FX-Series%20FX-6300.html" rel="nofollow - http://www.cpu-world.com/CPUs/Bulldozer/AMD-FX-Series%20FX-6300.html
This how most new memory is designed to function. How happy would you be were you to simply put the 2400MHz memory in place and have your computer not boot because the board and or CPU can not run at 2400MHz. Say Hallelujah! to JEDEC Standards that it runs the first time albeit at 1300MHz. It''s you the end user that must enter the BIOS and make changes to your memory settings to get it running at or near the 2400MHz, CPU dependent. I would suggest you search the internet for a 990FX overclocking guide. The questions you ask here have serious consequences if you are to just go off willy nilly changing BIOS options. As WKjun hinted at, I won't beat around the bush in telling you you need to understand more than you do now before your overclocking expedition begins otherwise you're going to destroy what you have in front of you. And with all due respect I say that. |
Atomic wrote:Posted By: Atomic
Date Posted: 13 Jan 2016 at 12:38am
|
Hey WKjun and Wardog, Thanks so much for taking the time to read my post and putting in a very meaty response in reply. In addition to my responses above, I have been doing a little more digging regarding reviews for my motherboard the AMD Fatality 990 FX. Other people have also mentioned that the AsRock VRM's are having issues. Both from an insufficient cooling of the VRM chips perspective and also the potential BUG in the BIOS which needs to be fixed by AsRock. Also someone mentioned an issue with big VDroop, about 0.8v. (I'm not sure exactly how this would affect over-clocking stability, so if anyone can provide insight that would be helpful). ??? The main reason I bought this AsRock motherboard was to over-clock it with the provided AsRock tools. I set it manually in the BIOS to the 4.33GHz range or 250MHz on the CPU Core Frequency. Then I disable the Cool and Quiet, Thermal Throttling, Spread Spectrum and Turbo boost. I want to get the CPU to run at 100% speed 24/7 without downclocking. I check the BIOS settings and it says "disabled" for all those options. BUT when I go into Windows I see the multiplier still fluctuating from anything from 1.0GHz all the way up to 4.3GHz. (I don't want this, I want to get the frequency to remain at 4.3GHz all the time). As when I am gaming I can see very clearly when it goes from 4.3GHz down to 1.0GHz, I am getting very noticeable lag and skipping in the game at that very moment. For the sake of argument, even when I reset the BIOS to factory default. Nothing is changed, I just save the default settings and load Windows 10. I have NEVER once seen the CPU Core Speed go up to the 4.1 Turbo boost for the FX-6300 CPU. Even if I try to run FurMark, Prime 95, Crysis 3, GTA 5, Counter Strike Global Offensive and many other games or Benchmarks. (I would imagine at least 2 of the Cores would be stressed very heavily and cause the CPU to Turbo clock, at least several times during the play test). I used Speed Fan, HWMonitor, AMD Overdrive but NONE of these monitoring applications show a CPU Core Frequency reading of 4.1 GHz. On default AsRock BIOS settings. I read also that the VRM's that Asus use are much better then the competition as they are a more advanced design. More efficient and precise apparently. Supposably the ASUS VRM's 6+2 are better then other competitors 8+2 VRM Phase. Is this true? Is Asus a better motherboard?
|
Posted By: wardog
Date Posted: 13 Jan 2016 at 1:02am
|
What Fatal1ty 990FX board DO you have? There are a few. As to whether or not the VRMs can't handle the FX-6300, let me assure they can and do. Without issue on any of the 990FX boards ASRock produces. I've had my FX-8150 hog on them all and what you're attempting at/in a NON-overclocked setting(ie: beyond any stock Turbo freq) with an FX-6300 should not present any problem at all. Breeze right along it should. Please, list out all your components with make and model number, including the PSU and kindly include the PSU's age. That way we here can know what you're dealing with parts wise. ie: Help us help you. Also, w/o re-reading the thread, what BIOS revision is your board running now? |
Posted By: wardog
Date Posted: 13 Jan 2016 at 2:20am
|
@ Atomic Knowing what you don't yet I cannot offer any more help till I have what components your system is made up of. And then I'll only point you in the general direction to 4.3GHz. But w/o knowing your system I'll be damned to say so and later have you pointing fingers at me for trashing YOUR computer. I trust you undersstand from where I come when I say that. And please, do not set the Bus Speed to 250. That's NOT the CORRECT way and in fact may harm your motherboard and or CPU in doing so. Reading what someone blah blah did with their system on the internet will most likely not be functional on your system. Every system is different unless you have the EXACT everything that that particular poster has, and even the it may not work on yours. |
Posted By: PetrolHead
Date Posted: 13 Jan 2016 at 11:01am
|
^Seconded. Atomic, in your first post you asked "What is the best way for me to push a safe and stable over-clock without causing instability?". The answer: Slow and steady. Setting your CPU to run @ roughly 4.3 GHz on all cores is not that. First of all, you CPU seizes being a 95 W TDP CPU. It may cause your socket and VRM section to run way too hot due to your choice of CPU cooling. Second, it will in all likelihood not be stable on stock voltages (you didn't mention changing voltages). When overclocking you should slowly work your way up in both clock frequency and voltage (and possibly CPU-NB/HT frequency and voltage, if the system seems to demand it), checking your stability and temperatures under stress testing after each change you make (unless you know what you're doing). If you're unlucky in the silicon lottery, your CPU might not even be able to do 4.3 GHz on all six cores on any reasonable Vcore, so that's definitely not the point where you should start. ------------- Ryzen 5 1500X, ASRock AB350M Pro4, 2x8 GB G.Skill Trident Z 3466CL16, Sapphire Pulse RX Vega56 8G HBM2, Corsair RM550x, Samsung 960 EVO SSD (NVMe) 250GB, Samsung 850 EVO SSD 500 GB, Windows 10 64-bit |
Posted By: Xaltar
Date Posted: 13 Jan 2016 at 2:40pm
It looks like all our resident AMD gurus are chiming in here, feel free to ask questions if you are unsure about anything. Overclocking is a great way to eek more performance from your system but if done wrong it can be catastrophic. You are in good hands 
|
Posted By: Atomic
Date Posted: 15 Jan 2016 at 12:43pm
|
Hey everyone, I'm not looking to over clock my system haphazardly and so I am seeking expert Council from those in the know. So I totally understand the concern from everyone involved here. As mentioned already in my earlier posts I want to know the best/safest way to over clock my system to get as much performance out of the AMD CPU as I can. I want to have the AMD CPU to run at the fastest frequency I can safely achieve at 100% 24/7. Which is where my question about the CPU throttling question came from. Even when I disabled AMD Power Saving, Cool and Quiet, Throttling the CPU speed was still throttling. But anyway, here is a list of the specifications of my build. (All the components except for the "HDD 2TB Western Digital" are brand new and purchased on the 21'st of December 2015). CPU: AMD FX-6300 CPU Cooler: Deep Cool Maelstrom 120 (I have mounted (x2) 120mm fans of either said of the radiator which is sucking hot air outside of the case). Motherboard: AsRock Fatality Killer 990 FX (BIOS Revision loaded is "P1.05" created on 08/04/2015) RAM: Corsair 8Gb 2400 MHz modules (x2) = 16 Gb total Graphics Card: Gigabyte G1 980 GTX 4GB Power Supply: Corsair RM1000 (1000 Watt) Case: Corsair Carbide Series "SPEC 02" (OS Boot Drive) SSD: OCZ 240 Gb (Drive for Programs, Games, Downloads etc) HDD: 2 TB Western Digital Black Series For the Monitor I have a 24 Inch 1920 by 1080 (Ben Q - BL2410)
|
Posted By: wardog
Date Posted: 15 Jan 2016 at 3:41pm
First thing you must do is safely and securely position a fan to blow over the VRM sink, the big black one that says Killer on it. Water cooling removes any air flow over this that would otherwise be there when using a normal top-down OEM style air cooler. This reason, and this alone, is most likely the cause of your throttling. The VRM's must have air blowing over them to cool them. For when they do heat up what you describe is a likely occurrence. Throttling. While I cannot say with 100% certainty this IS your problem, from here where I sit with my knowledge of AMD boards/cpus this is most likely your issue. And your only one I'll add. Nice system. Post back when you get that fan jerry-rigged securely. For w/o that fan there is no point in moving forward. We'll proceed afterward. |
Posted By: Atomic
Date Posted: 18 Jan 2016 at 11:46pm
|
Hey Wardog, If I am honest, that is a pretty big oversight by AsRock. To design a motherboard titled as gaming and for this particular product not being able to handle a pre-set over clock included in the motherboard as one of it's features. Just because the heat sink which covers the VRM's (The one with the Killer logo), is so poorly designed it is critically overheating and causing the system to loose stability. There really should be better thermal dissipation on the VRM heatsink to allow for water cooling units to be installed without the need to have a dedicated fan to be jerry rigged to blow air directly over the top. Am I wrong in saying that Asus or Gigabyte DO NOT have these same problems? I can't figure out how to jerry rig the fan into the space right next to the VRM heatsink as the radiator is nearby. So I am thinking either of these options: 1) If only I could get an RMA or refund for the water Deep Cool cooling unit and get something else that will work better for this particular motherboard. But how could I do that? Is it the CPU coolers fault? 2) Get a refund for the AsRock motherboard if AsROCK agree to raise an RMA for me and give authorisation for a swap by the place of purchase I got it from. Can that be done? 3) Have I made a mistake by going with AMD instead of INTEL for a gaming system? If not, then why? :-(
|
Posted By: WKjun
Date Posted: 19 Jan 2016 at 1:40am
|
Atomic, VRM & chipset cooling is an oversight of many MB makers. BUT if your CPU throttles, it triggered its lifeline to keep its temp. down and save your chip! It is advisable NOT to turn off APM, which fortunately doesn't in your case. This is CPU-wise. MB-/VRM-wise there is no such safeguard and you may damage your board permanently, at least age your components faster, each time you overheat them. Please accept the fact, that there is no guarantee of how much (or at all) you can OC any component and definetly the manufacturer has nothing to do with it, as OC'ing voids your warranty in most cases! Think of that: It's like a car that redlines between 7000 and 8000 rpm. Just because its there, the maker does not suggest, support or is liable if you over-rev the engine. Overcoming the limiter is at your peril! The next thing is, if you use a water-cooler, wo don't talk about enough air, but about air at all! There is no fan, passively cooling the surrounding components. These heatsinks are built to have passive cooling all the time. A distant case fan does not provide that. As a consolation, I ran into the same mess some time ago, having bought two(!) of them, performing 15° wores than my air cooler... I doubt whether AMD nor ASRock suggested a liquid cooler. They are some kind of press favourites for OC'ing, but you have learned that OC'ing is not in the mind of manufacturers. ;) So much for that. Back to your issue. Does it throttle even at default values? I didn't find that info yet. Do you know it is the VRM temp. for sure?! What does HWMonitor say in idle and under load - are there any temp. values raising above 80° (60° for CPU, "package" or "core" temp.)? If so, cooling has to be better, regardless.
Well, no, there is nothing wrong with your CPU, isn't it!? PS: If a (AMD) CPU is fully or nearly full loaded, it throttles, because temp. or power consumption is too high (if APM or another safeguard is active). This is normal behaviour. Practically a 50% load should keep clocks stable at max. (no turbo!).
------------- 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: PetrolHead
Date Posted: 19 Jan 2016 at 2:59pm
Those preset overclock settings are not really guaranteed to work, since each CPU and combination of components is basically unique. The system those settings were determined on may he able to handle them, but your particular CPU may, for example, need more voltage to stay stable. The same goes for cooling; if your cooling is not arranged in the exact same manner that was used for obtaining those settings, you may need to put in some extra effort.
Yes. Some of their boards may have stronger VRM sections, but water cooling combined with an overclock pretty much always necessitates the use of a spot fan - depending on the CPU and OC, of course.
Not really. Generally speaking your GPU is the most important component in your rig when it comes to gaming. A stock FX-6300 should be plenty for most games and GPUs, although some games (like Crysis 3) could benefit from an FX-8xxx CPU or a more powerful Intel CPU, and if you had an SLI setup you might actually see the CPU bottleneck the GPUs at 1080p. Intel's CPUs do have an advantage in performance (and power usage) at the moment, but that doesn't directly make AMD a bad choice for gaming. They provide decent bang for buck, especially when ovwrclocked. We just need to find you the best settings for your system. ------------- Ryzen 5 1500X, ASRock AB350M Pro4, 2x8 GB G.Skill Trident Z 3466CL16, Sapphire Pulse RX Vega56 8G HBM2, Corsair RM550x, Samsung 960 EVO SSD (NVMe) 250GB, Samsung 850 EVO SSD 500 GB, Windows 10 64-bit |
Posted By: wardog
Date Posted: 19 Jan 2016 at 5:07pm
|
Atomic, I can't add anything to that which WKjun and PetrolHead have already replied. Every manufacturer that I know, when pressed, will say that when water-cooling is used an additional source of air flow MUST be accounted to blow over the VRM sink/section. I'm sorry, but that's common knowledge, OC'ing or running at stock, that must be adhered to. There's no getting around it. The VRMs need cooled. That's not by design, but by necessity. |
Posted By: PetrolHead
Date Posted: 20 Jan 2016 at 11:27am
|
Atomic, while you figure out how to keep your VRM cool, you could maybe try to find out what you can get out of your CPU at stock voltages. Increasing the clock speed alone has a smaller effect on power draw than increasing voltage, which should make life easier for the VRM. Just take your time with it and you should see warning signs before your rig is in any real danger of getting damaged: Note: You may want to check you have Turbo Core, Cool'n'Quiet etc. disabled. Since we are just finding the maximum stock voltage clock speed, I personally don't think having those enabled matters much. It's when you go for maximum OC that you need maximum control over your system parameters, but for now all you really need is CPU-Z, HWMonitor and Prime95. 1. Start with stock settings and run Prime95 blend on all cores for about 20-30 minutes with HWMonitor running as well. This will give you baseline temperatures and show you how much Vcore and the clock speeds fluctuate (if at all) when loaded. 2. Increase the CPU clock speed multiplier by 0.5 (or the clock speed by 100 MHz), check with CPU-Z that the setting has stuck and run Prime95 again on all cores for 20-30 minutes while monitoring everything on HWMonitor. If temperatures stay below 60 C (CPU/package) and 70 C (socket/CPUTIN most likely) and no workers are dropped, repeat this step. 3a. At some point one or more workers will stop before 20-30 minutes has passed. This means the system is unstable and will (usually) need more Vcore to function at this clock speed. Lower the clock speed by 100 Mhz and run Prime95 blend on all cores for two hours. If workers are dropped, you need to lower the clock speed again. If no workers are dropped, you have reached what some people consider reasonable stability. Personally I would suggest running Prime95 for at least 4 hours, preferably 8. 3b. If the temperatures reach 60 C/70 C, you've pretty much reached your limits with the current cooling. If Prime95 is still running and the temperatures are running, stop the test immediately. I'm not expecting heat to be an issue for the CPU itself, but your socket temps may be high due that water cooling system. Also, if you notice clock speeds dropping or cores not being under 100% load while running Prime95, that's likely your VRM throttling and you need to stop the test and back off with the OC. Note that there will be short periods of time between tests that the CPU may not be under 100% load, so before you panic, make sure that the core(s) in question are actually in the middle of a test. All of this at your own risk, naturally. ;) It's hard to say what your maximum OC at stock Vcore (1.35 V, I think) will be. The 3.8 GHz p state has a Vcore of 1.4125 V seems high considering the increase in clock speed is under 10%. I wouldn't be surprised if you could hit that at stock voltage, especially if the water cooler is keeping the CPU cool. ------------- Ryzen 5 1500X, ASRock AB350M Pro4, 2x8 GB G.Skill Trident Z 3466CL16, Sapphire Pulse RX Vega56 8G HBM2, Corsair RM550x, Samsung 960 EVO SSD (NVMe) 250GB, Samsung 850 EVO SSD 500 GB, Windows 10 64-bit |