Jump to content

I think we could use a video about GPU undervolting.

 

I've seen increasing discussion about it recently because the Diablo II Resurrected and New World betas were really hard on GPU temperatures. Some GPUs died. Some people were too scared to run the games because if their GPU dies they can't afford to replace it.

 

I've seen this conversation dozens of times:

Bob> why is my gpu 10c hotter in this game than other games?

Dave> mine too
Steve> your cooling sux, git gud

 

I saw a video of someone limiting their gpu voltage in New World and they didn't even lose any performance! Some reading later reveals that undervolting properly (whatever that means) can quite often also result in better performance.

 

Why would a GPU draw power but not give you more performance in return. Consider all the GPUs in the world, if they're all drawing more power than they need then that's a significant environmental impact and financial cost.

 

People who tried to lower their GPU temperature by other means - lowering settings, resolution, no AA - ended up being CPU-limited instead, and being CPU limited in New World will result in your CPU having all the same thermal issues instead. So maybe undervolting the GPU is the solution that keeps you GPU-limited while also preventing your GPU from drawing excessive power and overheating.

 

There's definitely something to this line of thinking that needs more investigation by some nerdy YouTubers.

Link to comment
https://linustechtips.com/topic/1373682-gpu-undervolting/
Share on other sites

Link to post
Share on other sites

the gpus are released on the exponential part of the curve of performance per watt, it's not for efficiency, hence undervolting exists, there are already videos.

 

As for new world blowing up cards, that was on nvidia's cards ovp not working correctly, gotta have a fps limit in place just in case, lesson learned.

5950x 1.33v 5.05 4.5 88C 195w ll R20 12k ll drp4 ll x570 dark hero ll gskill 4x8gb 3666 14-14-14-32-320-24-2T (zen trfc)  1.45v 45C 1.15v soc ll 6950xt gaming x trio 325w 60C ll samsung 970 500gb nvme os ll sandisk 4tb ssd ll 6x nf12/14 ippc fans ll tt gt10 case ll evga g2 1300w ll w10 pro ll 34GN850B ll AW3423DW

 

9900k 1.36v 5.1avx 4.9ring 85C 195w (daily) 1.02v 4.3ghz 80w 50C R20 temps score=5500 ll D15 ll Z390 taichi ult 1.60 bios ll gskill 4x8gb 14-14-14-30-280-20 ddr3666bdie 1.45v 45C 1.22sa/1.18 io  ll EVGA 30 non90 tie ftw3 1920//10000 0.85v 300w 71C ll  6x nf14 ippc 2000rpm ll 500gb nvme 970 evo ll l sandisk 4tb sata ssd +4tb exssd backup ll 2x 500gb samsung 970 evo raid 0 llCorsair graphite 780T ll EVGA P2 1200w ll w10p ll NEC PA241w ll pa32ucg-k

 

prebuilt 5800 stock ll 2x8gb ddr4 cl17 3466 ll oem 3080 0.85v 1890//10000 290w 74C ll 27gl850b ll pa272w ll w11

 

Link to comment
https://linustechtips.com/topic/1373682-gpu-undervolting/#findComment-14993554
Share on other sites

Link to post
Share on other sites

I suppose everyone has an OC video to call themselves a tech youtuber, assume same for undervolting?

 

But no it doesnt need further investigation. Just because less people outside the tech world know about it doesn't mean enthusiasts don't. Just videos that carry the message to the masses will do.

 

18 minutes ago, throwaway3224 said:

if they're all drawing more power than they need then that's a significant environmental impact and financial cost.

That's called margin to account for difference in silicon capability. Heard of Nvidia 30 series mass crashing? That's Nvidia pushing too close to the boundary of stability. Testing each GPU individually for a GPU specific profile is an even bigger use of money than wasting power.

CPU: i7-2600K 4751MHz 1.44V (software) --> 1.47V at the back of the socket Motherboard: Asrock Z77 Extreme4 (BCLK: 103.3MHz) CPU Cooler: Noctua NH-D15 RAM: Adata XPG 2x8GB DDR3 (XMP: 2133MHz 10-11-11-30 CR2, custom: 2203MHz 10-11-10-26 CR1 tRFC:230 tREFI:14000) GPU: Asus GTX 1070 Dual (Super Jetstream vbios, +70(2025-2088MHz)/+400(8.8Gbps)) SSD: Samsung 840 Pro 256GB (main boot drive), Transcend SSD370 128GB PSU: Seasonic X-660 80+ Gold Case: Antec P110 Silent, 5 intakes 1 exhaust Monitor: AOC G2460PF 1080p 144Hz (150Hz max w/ DP, 121Hz max w/ HDMI) TN panel Keyboard: Logitech G610 Orion (Cherry MX Blue) with SteelSeries Apex M260 keycaps Mouse: BenQ Zowie FK1

 

Model: HP Omen 17 17-an110ca CPU: i7-8750H (0.125V core & cache, 50mV SA undervolt) GPU: GTX 1060 6GB Mobile (+80/+450, 1650MHz~1750MHz 0.78V~0.85V) RAM: 8+8GB DDR4-2400 18-17-17-39 2T Storage: HP EX920 1TB PCIe x4 M.2 SSD + Crucial MX500 1TB 2.5" SATA SSD, 128GB Toshiba PCIe x2 M.2 SSD (KBG30ZMV128G) gone cooking externally, 1TB Seagate 7200RPM 2.5" HDD (ST1000LM049-2GH172) left outside Monitor: 1080p 126Hz IPS G-sync

 

Desktop benching:

Cinebench R15 Single thread:168 Multi-thread: 833 

SuperPi (v1.5 from Techpowerup, PI value output) 16K: 0.100s 1M: 8.255s 32M: 7m 45.93s

Link to comment
https://linustechtips.com/topic/1373682-gpu-undervolting/#findComment-14993575
Share on other sites

Link to post
Share on other sites

30 minutes ago, throwaway3224 said:

I think we could use a video about GPU undervolting.

 

I've seen increasing discussion about it recently because the Diablo II Resurrected and New World betas were really hard on GPU temperatures. Some GPUs died. Some people were too scared to run the games because if their GPU dies they can't afford to replace it.

 

I've seen this conversation dozens of times:

Bob> why is my gpu 10c hotter in this game than other games?

Dave> mine too
Steve> your cooling sux, git gud

 

I saw a video of someone limiting their gpu voltage in New World and they didn't even lose any performance! Some reading later reveals that undervolting properly (whatever that means) can quite often also result in better performance.

 

Why would a GPU draw power but not give you more performance in return. Consider all the GPUs in the world, if they're all drawing more power than they need then that's a significant environmental impact and financial cost.

 

People who tried to lower their GPU temperature by other means - lowering settings, resolution, no AA - ended up being CPU-limited instead, and being CPU limited in New World will result in your CPU having all the same thermal issues instead. So maybe undervolting the GPU is the solution that keeps you GPU-limited while also preventing your GPU from drawing excessive power and overheating.

 

There's definitely something to this line of thinking that needs more investigation by some nerdy YouTubers.

First of all, I don't know about Diablo II but New World is beta and unoptimized - also it was confirmed that the EVGA cards that broke had a fault. It was not New World's fault, New World just triggered this fault in the graphics cards. Yes, I have also seen others report that their graphics card have died in New World, all sort of models, AMD and Nvidia cards but nothing was ever confirmed as far as I know - only the EVGA cards was confirmed, and by EVGA themselves. 

 

Some games just utilize the graphics cards much harder than most other games - two examples would be Days Gone and Metro Exodus Enhanced Edition. These two makes the GPU pull way more wattage than most other games, causing increase heat output aswell, and this is something some people report as a fault when it isn't. It's just like some benchmarks tanks hardware way more than other benchmarks. Also, yes, some might not really see the increase heat output, because they have a monster cooling setup or very, very good airflow. There are so many factors to why some report higher temps than other when having the same graphics card. 

 

Yes, undervolting CAN give you more performance than stock. A good example would be that when running stock / using a stock voltage curve, your GPU might pull 1.100V at 1900MHz, but at this point it has hit powerlimit - standard 100%. By removing the overall voltage that GPU can pull (undervolt), lets say we cap it at 0.900mV, this might save you 40watts over the 1.100V, and these last 40watts will let the GPU turbo to 1950MHz - therefor the undervolt just increased performance. Now the tricky part is to not lower the voltage too much because then the GPU becomes unstable. 

Yes, this could be further uptimized by the graphics card manufacturers but with a power limit in place, the GPU will "never" pull more power than said, therefor it is not more enviromentally unfriendly - if you could call it that. You pay for a 350W GPU, and it pulls 350W at max (kinda). Also the auto voltage, that feeds the GPU with more voltage than needed, just means stability is more certain. So it's also a way for the GPU manufacturers to make sure that the graphics cards run stable when costumers get them. 

 

I played the New World open beta and my GPU was never fully utilized and was never hot, but my CPU on the other hand, it was hot. 

Just my thoughts on your questions 

GAMER 

SPEC: HYTE Y60 White/Black + Custom ColdZero ventilation sidepanel – AMD Ryzen 7 9800X3D + ASUS ROG RYUJIN III 360 – G.SKILL Trident Z5 Neo RGB DDR5-6000 64GB CL30-40-40-96 – ASUS RTX ROG ASTRAL 5080 OC – ASUS ROG STRIX B850-E GAMING Wi-Fi – Samsung EVO Plus 1TB, Samsung EVO Plus 1TB – Corsair HX1200i
GEAR: Samsung Odyssey Neo G9 G95NC 57" – ASUS ROG Harpe Ace Aim Lab Edition Wireless – ASUS ROG Claymore II Wireless –
ASUS ROG Sheath BLK LTD – FiiO K7 DAC/AMP + Beyerdynamic DT 770 PRO X (Limited Editon), Edifier S351DB, Steelseries Arctis Nova Pro Wireless


Home LAB
SPEC: Corsair 4000D Airflow TG – AMD Ryzen Threadripper 1950X – Corsair Vengeance LPX 128GB (CMK32GX4M2D3000C16 x4) – ASUS GeForce GT710 1GB – ASUS ROG Zenith Extreme – ROG AREION 10Gbps – Toshiba XG6 1TB, WD 256GB, Kioxia 256GB – Corsair HX750
GEAR: Philips 346B1C 34" – Logitech MX Keys – Logitech M705 – Jabra Evolve2 75

 

Racing SIM

GEAR: Sim-Lab GT1 EVO Sim Racing Cockpit + Integrated Vario Triple mount – Svive Racing D1 Seat – MSI 325CQRXF 32" x3 – Simagic Alpha Mini + Simagic GT4 (Dual Clutch) – CSL Elite Pedals V2 – Logitech K400 Plus

Link to comment
https://linustechtips.com/topic/1373682-gpu-undervolting/#findComment-14993578
Share on other sites

Link to post
Share on other sites

  • 4 weeks later...
On 9/16/2021 at 2:27 PM, Tan3l6 said:

I prefer lowering the power limit to 80% at Afterburner - much lower temps always, performance hit not so great.

I just tried the same.

Managed to have my MSI 3060 gaming x run at 61 degrees C with a fan speed set at 35% while gaming Warhammer 2.

Havent benchemarked it yet, but the noise was much more tolerable.

mITX is awesome! I regret nothing (apart from when picking parts or have to do maintainance *cough*cough*)

Link to comment
https://linustechtips.com/topic/1373682-gpu-undervolting/#findComment-15041600
Share on other sites

Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

×