Search the Community

I have read on reddit that there is a way to configure Jedec ram profiles by writing to the SPD on the ram modules. https://www.reddit.com/r/AMDLaptops/comments/omj6mu/guideproviding_advices_for_tightening_ram_timings/ I have been thinking of making some changes that improve the performance and latency on my NON-XMP laptop a bit, but i was wondering if anyone knows which timings are harmful to mess around with when doing this. Which ram timings can cause serious harm and which timings would you recommend me to change to get the best performance and latency out of a jedec profile? I want to take the safe route and only mess with the timings that dont brick my sticks that easily or cause issues that easily. Current ram kit: hx432s20ibk2/32 3200 cl20 Laptop: Lenovo Legion 7 16achg6

Hi i tried to overclock my PC. Original Specs : CPU : Ryzen 5 3600 @ 3.6 Ghz. Memory : Cruical Ballistix DDR4 2666 Mhz CL 16. ( Probably Micron E Die. Pls Correct me if i am wrong ) GPU : Zotac 1650 Super -> 1530 Mhz Core, 6000 Mhz Memory. Motherboard : Gigabyte B550m DS3H. Cooler : Deepcool Gammax 400 Pro -> Single Tower, Dual 120mm Fans Air cooler. Overclocked Specs : CPU : Ryzen 5 3600 -> 4.35 Ghz All Cores @1.26v VCore and 1.14v SoC. Idles at 45°C and reaches 83°C when stressed to 75W TDP. Memory : 3200 Mhz CL 16-18-18-18-40 1T @ 1.35v GPU : 1725 Mhz Core, 7500 Mhz Memory. Question : Are my voltages safe? Can anybody suggest how to improve on it? Can voltages on RAM be increased more? If yes, pls suggest SoC voltages for it. Thanks in Advance!!

I have an MSi X570 A Pro motherboard and I tried overclocking my Corsair Vengeance LPX 3200Mhz RAM timings with the Ryzen calculator. And so after every single attempt failed with even the smallest changes I decided to use the preset overclocks and shoot for 3600 Mhz as I was rather annoyed. First try I tried using the medium tight timings, didn't work. So second try I went for the lowest preset timings, meaning the highest numbers for the 3600Mhz preset, and then I went out of the room for a minute expecting it to go through it's now regular, off, on, sudden off, then on, then black screen, then waiting for a bit, and finally "Memory overclock failed. Press F1 for bios" type of thing... but no. It just stayed in the turned on but not posting stage. So I turned it off, tried again, same thing. *ok* ... So I turned off the PSU, held the power button to clear the CMOS, started the computer and now the power light isn't on, the motherboard fan is maxed out, and the CPU debug light is on, and all of that keeps going until I shut it off... so, did I screw myself over super hard because I wanted to squeeze just a little more performance?

Hey, It's been 20+ years since I last build and used a desktop so rusty AF. Regardless I'd like to get every bit of performance out of the CPU. i9-11900K Gigabyte Vision D Z590 Rev 0.0 Gigabyte RTX 3070 Vision G. Skill F4-3600C14Q-32GTZNA 1.45v Corsair H150i Capellix 360mm AIO Corsair RM850x PSU That's about covers the important tidbits, but now what? Anyone with this board and/or CPU that did some OC? Going to need a step by step guide. So hopefully there are people that actually want to help as I want to give it my all but have to start all over since it's been too long. No memory and GPU oc, only the CPU. Regards,

Hello my english bad sorry how can i overclock my rams? XPG Gammix D30 i add taiphoon burner screenshot please help me

Alright, so I am SUPER confused on this, hehe get it - But in actuality, I can't figure out if my GPU is bad, My MOBO is Bad, if it's Undervolted (which it doesn't seem to be) if it's not OC'd right, or if It's just the game being... well, itself - I've done a LOT of Trouble Shooting, and I am at my wits end. I have a RTX 2080 Super, factory OC'd of course, from Alienware, (Yeah, yeah, prebuilts are overpriced an bad) and it happens a lot in applications and games, primarily games, but I noticed there's a DISTINCT flicker, since day 1, I don't think I have any Bottlenecks that would cause it to bug out like it is, Currently, I have a i9-10850K w/ an Corsair H60 Waterblock (OC Profile 2 for Alienware and Thermal Grizzly for Paste), with HyperX Fury RAM @ (That should run @) 3200MHz XMP Profile 2 (it's actually running @ 3192 MHz tCL 20 which it was 15 when I first bought it), 2 Samsung 850 Evo's SATA and a Western Digital 1TB @ 7200 RPM 64MB Cache (WDC WD10EZEX-75WN4A1) and as for my Main Drive, that would be a Samsung 980 NVMe 1TB, PSU is from Dell, @ 1000W, The MOBO is from Alienware, specific IPCML-SH Model, Z490 Chipset. The trouble shooting steps I have took has been. • Restarting the PC • Reinstalling NVIDA Drivers (default install) • Reinstalling NVIDIA Drivers (clean reinstall) • Reinstalling NVIDIA Drivers manually after using DDU in Safe Mode • Reinstalling/Reseating the card • Reseating the card multiple times • Trying different power cables • Trying the 2nd PCIe 16x slot w/ different cables • Reinstalling Windows 100% Fresh w/ New Drive • Configuring NVIDIA Settings + Game Settings I have not tried inserting it into a second Desktop yet, that would be an older model Alienware, an Alienware R6, and that's because I noticed if I UNDER Clock it using MSI Afterburner (Cause Alienware Command Center and it's OC Abilities are a joke and have broke almost 99.9% of the time I have EVER used it) - it would run just fine, I have to Underclock it by almost -300 which would set Core Clock to 1500MHz, WAY under base clock @ 1650MHz, and then it seems to stabilize, it happens in more than 1 game, and even on my Web Browser at times if I Alt-Tab while running anything GPU Intense, I have heard that it is an issue with specific games not being able to handle the hardware... but it does it outside of game too at times.

Anyone have a guide to overclocking/downclocking a ryzen 7 1700 non x that i could use?

I just want to know if overclocking ram with different speed will work. For example my mobo has 4 ram slot, slot 1 and 3 is using a corsair 16Gb 3200Mhz ram and the other 2 and 4 use a 16Gb 2666Mhz ram sticks. So what happen if i overclock it, will it work or what will happend to it? is it possible to do this?

Hi guys. I'm currently modding one of my old cards (r7 370X) and there is some weird power plain and power distribution. I don't want to overload my PCI power (I found moded bios where are all limits disabled) and when I measured PCI power it was pulling 110W. So I want to solder wire (+12V) from 8 pin to the PCI power. I don't think it should short anything, but to be sure I'm asking if it would be problem. Because I don't want to short my PSU rails. In the past I just cut the lines from pci power and soldered wire from 6 pin to pci on my 1650S. I don't want to do this here. So what's your opinion

Hi fellow OC enthusiasts, I am currently fiddling around with overclocking my hardware for the first time, so I am a bit of a newbie. I generally tend to inform myself as much as possible and I am quite careful so I don't take unnecessary risks. However, now I apparently have run into something where I can't find an answer that I understand. Software I use: MSI Afterburner CPU-Z GPU-Z Valley Benchmark HWInfo64 I have OC'd both my CPU (AMD Ryzen 5 1600x) and my GPU (Gainward GTX 1070) and I'm also running a Memory Try It! preset for my RAM. For more information regarding my specs, please see my signature. If you need more info regarding my overclocks, please let me know. My CPU is running at a stable 3900 MHz with a voltage of 1.375, tested with Ryzen Master and made permanent through my BIOS. My GPU has so far been running at a stable +130 MHz Core Clock and +620 MHz Memory Clock. Now I was curious about what else I can do if I adjust the GPU voltage. From what I have read there is a hard limit set by Nvidia, but there still seems to be some potential that I could use. I've found a guide that recommended to first set a fix voltage. I can post a link if that's okay, but basically what he says is I have locked my GPU to a constant voltage of 1000 mV (which is lower than what it would usually reach). I have then set Temp and Power limits to their max values (92c and 112%). When increasing the memory clock to 650 MHz I am beginning to see "stars" when running Valley, so I am sticking to 620 MHz. However, I can boost the core clock far higher than before it seems. Temps stick to a maximum of 79c even with +160 MHz core clock, however Valley benchmark scores don't improve beyond the margin of error. At a +140 MHz I had a Min FPS of 33 instead of the normal 22, so I assume that this is a core clock at which games perform a little more stable, I am yet to test in an actual game though. Any other boost has given me 22 Min FPS again. I haven't tested higher boosts yet as I am unsure if they currently have any effect at all and if they might cause issues when unlocking the voltage again. What made me curious is the fact that despite a higher core clock there seems to be no improvement in performance, no increase in temperature and what makes me wonder the most is that GPU-Z now only shows "Idle" as the PerfCap Reason when it had shown "Pwr" and/or "vRel" before. If it would show these reasons again that'd make sense to me, but "Idle"?. This makes me think that locking my GPU voltage had no effect or at least doesn't let the GPU reach a higher level of performance now. Since the whole voltage-thing still is quite blurry to me anyway, I am now looking for advice. Again, I wasn't able to find anything that I actually understand when doing my research. There seems to be people benefiting from undervolting as it decreases the temps and therefore lets the core boost to a higher clock, then again overvolting seems to be better for others as it increases the amount of power the core can draw in oder to boost its frequencies. But what would I have to do to find the sweet spot for my GPU? I am not afraid to test a few things out, I just don't know where or how to start. Thanks for your time.

So, i was thinking.... Can get around 4.2ghz after doing OC on a cheap Xeon E5 2470 V2 for around 50$ from AliExpress or Alibaba? IS IT POSSIBLE? Or should i just go with RYZEN 3 3300x instead? Stuffs I'll use the PC for- *Photo editing *Video editing *Gaming *Screen recording

Hello. I decided to overclock my r7 5800x, but when I set 4400MHz and 1.25V the temperatures for IDLE are normal (40-50 celsius degrees) but when i run ap Cinebench r23 the temperatures are very high (about 100-105 celsius degrees). I have nzxt kraken x73 and aorus b550 master

Will overclocking my 590 make my gpu life span less longer? Or only if I touch like the voltage? I am overclocking it with msi afterburner and MSi kombustor.

Ok so I had 3 desktops with 6 vega 64 nitro+ GPUs (2 in each dekstop case) I used them for mining, I downgraded the bios to vega 56 nitro+ and configured custom timings on the HMB memory and used different powertables. The cards were cramped inside the cases with temps (at full load) around 60C for the GPU and around 70C for the HBM memory But at 1100Mhz (GPU) and 1045Mhz (HBM) they were mining at around 54.8Mh/s on ethereum (a couple could go 1060Mhz which resulted in 56.7Mh/s) The cards run just fine 100% stable 24/7 What's the issue then? Well I decided to buy an asrock h110 pro BTC motherboard and make an open mining rig with extra fans etc. The temperatures of the cards are much lower now (I also changed the paste they had with Kryonaut Extreme I think it is the best after liquid metal) Now I have like 30-40ishC on the GPUs and 50-60ishC on the HBM at full load The issue is that for some reasons the same exactly cards that used to work (same OS same drivers same mining software,same memory timings same powertables) at 1045 and 1060Mhz 100% stable 24/7 at higher temps on the memory now refuse to go a step above 1020Mhz! It is not the fact that I took them apart and changed the paste because I tested them for a couple of hours (on the desktops) and they run at the same settings with lower temps stable. I can get a card or two run above 1020Mhz but its always a lottery Its like something happens if I set the frequency of all of them (e.g GPU1 may hang, but on reboot GPU1 may work but GPU4 now hangs) and I get BSOD "thread_stuck_in_device_driver" or something like that. I think it has something to do with lag/interference with the driver since I use 6 GPUs (even adrenaline gets laggy if I click on GPU boxes to check their settings like a 5 to 10 second delay when clicking e.g GPU2 from GPU1 etc) I use the GUI of this tool https://github.com/Eliovp/amdmemorytweak To set the OC profiles. Any ideas on what to try? P.S EnableULPS and EnableCrossfireAutolink registry keys are disabled for all the GPUs.

Hello, I recently got a 3070 and I want to undervolt/oc it but i do not know how.. i watch all the tutorials, i still dont understand , and i dont want to do it my self and make mistakes.. if someone is willing to come on discord and help me to do it that would be great, i dont have any friends that know that can help me so here I am, thank you very much, and sorry for the silly question.

Recently i upgraded on low cost PC, by buying xeon 5687 (from i7-920), 16gb RAM(from 8gb) and a cpu's cooler (from intel's stock). After few attemps set OC cpu to 4.2ghz, i decided to increase voltage (apparently increasing voltage can improve stability, which at this cpu any OC make sound stuttering/near collapse). At first i set on 1.5V and tried to run OCCT, but after 2 min i decided to against it, so i reseted pc and decrase to 1.4V (1.35 is maximum for VRM according to intel ark). Overall i abandoned OC since Cpu's tcase was exceeded on OCCT in 3 min, but then a issue happened randomly within 1H there is freeze/hang which stay until hard reset (mouse,keyboard,headphone lost connection to pc). Somehow after restart i didn't get any freeze yet (unless its turned off and left for hour(s)). Now the questions are : -How much "100%" probability its caused by cpu because of being damaged by 0.15 V above? -Is there any program to monitor and save data, to doublecheck if this caused by cpu or maybe RAM? specs: -motherboard : dx58so -ram :pc3l-12800u-11-11-a1 1.35V (old one was KHX1600C9D3K2_4G 1.5V) -cpu : intel xeon 5687 (old one i7-920) -ssd : idk, it's from samsung 250gb from lenovo g50-30 laptop. -cooler : coolermaster hyper t200 (old one Intel E29477-002) -PSU : tagan tg600-u33ii PS. sorry for any grammar mistakes, somehow it does check correctness in different language.

Hello, I have: Gtx 970 Rog Strix i5-6600k Asus Z170 Pro Gaming 600W Psu I want to OC but not sure what settings to up on msi afterburner. can anyone help?

So i was going to buy the ASUS RTX 3090 strix OC but i noticed people saying that if you watercool it its going to loose its warranty vs the EVGA 3090 FTW3 Ultra which doesn't void warranty if you watercool it. So i found the EVGA 3090 FTW3 Ultra quite interesting instead and was wondering if anybody knows if there's a big difference between the 2 cards comparing to overclocking and such? I'm going to overclock the card I'm getting so would you just pull the trigger at the STRIX and not care for warranty or would you get the EVGA instead? Is the 2 cards more or less the same when overclocking or is there a big difference between the EVGA and STRIX? Hope to get some tips on which to buy Thanks!

Hello everyone As you can see in one of my last posts I hav oc'd my whole system without major problems. The problem now is whilst an hour of furmark does work without problems, Rocket League on the other hand fails after 5 minutes. I don't get why. Setting down Memory Clock did not help, ramping the Voltage to the max (apparently 1213 mV) does not help either. But as soon as I change to the ASUS standard OC profile everything works fine. Why? Can anyone explain this to me? Any Idea how to fix it? Temps are typically at 74°C after an hour of stresstesting GPU and CPU ath the same time. Thank you beforehand Greetings, Ale

As my specs below suggest, I have a 5800X, running on an Asrock B550 Pro4. I had clocked in a modest all core OC of about 4200 MHz the day I got the CPU, with voltage set at 1.15V, it was running fine with no hick ups. I was also running the stock XMP profile on the GSKILL Sniper Xs (3600 MHz, 1.35V, 19-20-20-20), no problems there either. Today, I decided to clean my PC. On turning it back on again and going into the BIOS, I find that all my OCs have been reset. No problem, I think, must've unseated my CMOS battery by error or something, so I check to see that it is seated properly, restart again and go into my BIOS again to set the original OCs which I had set. Now this is where the trouble starts. No matter what I do, it won't accept my original OCs, even the XMP profile. In fact no OCs are stable for some reason and the PC just won't post if I dial in something. This is, of course, quite startling, since I have never seen anything like this before in my experience with overclocking stuff. Is the MOBO completely shafted, or something? I'm not sure. Any help is appreciated, cheers.

Hi, I recently bought a kit of 2 corsair vengeance rgb pro at 3600 cl16 (Micron E-die) and overclock them, I want to know if it's a good overclock. The xmp profile is 3600mhz cl16-19-19-36 1.35v and i managed to run them stable (i think) at 4266 cl17-22-22-42 1.45v, I left the other timings in auto as well as the other voltages. I use memtestpro86 to verify if it's stable runing it for 45 minutes. Is this overclock good or the tRCD, tRP and tRAS values are too high? Motherboard: Z490 gaming edge wifi CPU: i5 10400f

Hello everyone After some thought I found that this could be the right place to ask. I yet never had problems with my performance until I bought my new Monitor. I upgraded from 16:9 20" 1080p to 32:9 49" 1440p. I expected my GPU to limit me but not that hard. This way I remembered why I bought an OC GPU and CPU and started to tinker around. By now I found some stable values but I am not fully sure If it is okay to use this as an everyday profile. My settings: GPU: Asus Strix GTX970 It never gets over 74°C even after an hour of stresstesting. I think these are pretty safe? CPU: Intel I7 4790K (normally at 4GHz, Boost 4.4GHZ) I have set the values in the BIOS but the AI Suite picked them up. (All cores are the same) If you wonder about the Values shown at the bottom, speedstepping and turbo is still enabled as I don't think it has to heat the Room unnecessary and may shorten the lifespan less. The Voltage under load shows about 1.296V with short spikes up to 1.301V. Whilst stresstesting the temperatures are varying between 70°C and 85°C. Averaging at around 74°C. 85°C Are often just while fans are ramping up afterwards 80°C seems to be the maximum hit after an hour. Do the temps always vary so much during stresstesting or is my cooling setup a bit sketchy? I think my 800W powersupply should be able to deliver without problems and it also did not seem to limit whilst I stresstested GPU and CPU for an hour simultaniously. What do you think about these settings? Is this acceptable (for the next years)? I think I will replace the GPU with a 3070 or a 4060 as soon as they are available for a relatable price. If there is anything left I forgot, tell me! :) Thank you all very much beforehand and enjoy the day! Greetings from Switzerland, Ale.

So i have been looking around the internet for some OC results on the asus strix 3090 OC and the normal asus strix 3090 and couldnt find anything Im looking to upgrade my PC and was thinking of buying a asus strix 3090 OC but a friend of mine told me that you can get better overclocking results out of a normal asus strix 3090 which seems odd to me Does anybody here have some overclocking results from a normal asus strix 3090 vs a OC version? Isnt the OC version a kind of "binned" version of the normal version and therefor should be able to get better overclocking results out of the OC version vs the normal version?

Hello I need help for my new build i got a few days ago my ram speed isn't at its advertised speed, I ENABLE XMP and its sets one part of my RAM frequency to 3200 but the other part still says 2133. My build is a APU build so my memory speed is very important i have dual channel G.Skill ram. I have provided a video of me trying to get 3200mhz please help me if you can and here is my build: https://pcpartpicker.com/user/txhh/saved/DMr2Vn tiktokvideo_1613571666.mp4

When I use ryzenmaster, if I link all cores to the same frequency and run a benchmark, ryzenmaster will show every core at the speed I set it to. But when I try to clock one core to a slightly higher speed, all of the other cores drop back to near stock speed. I've tried using creator mode, and profile 1, how can I set each individual core clock without the other clocks going back down to stock? <---- What I set the clocks To (They stay at all 4625mhz btw) <-----What they go back to