Search the Community

Basically what the title says. I've had the card for a few years, and I'm not too knowledgeable about card degradation, but I feel like that's not what's happening here (inb4 need to replace). Only one single part of the card temp is actually spiking that high, while the other two sensors are reading stuff normally around 70 and 75 (sometimes a bit higher). It used to be that I'd be able to play high-end games at max graphics while streaming and it wouldn't overheat. But now, if I'm just chilling and not streaming, it'll overheat, sometimes even if it isn't on max graphics. I'm worried something is wrong with the card, but I've no ide why. The important specs for my computer are included in the screenshot, which shows my temps. This happened earlier today and I immediately closed the game I was playing, which was tetris effect (not in vr mode) After shutting it down, it immediately went down to these levels These were actually done with my fans all blowing at 100%, and it was worse than earlier when I didn't have them that high. I'm confused and worried. My case shouldn't be a problem, and I've been dusting semi-regularly (last dusting was a month ago). Oh, and my case is the Corsair D500 Obsidian, which I'm assuming has decent airflow, and even if not, these temps still happen with both side panels off. It should be worth mentioning, that there was a defining point that this all started happening, it didn't gradually occur. Just one day the temps started spiking like crazy. Also, overall gpu temps have just been much higher than normal since this point, not just the random one part getting super hot. Wat do?

Hello everyone! I just realized that my gpu (asus rx6600) is pci-e 8x gen 4 by checking gpu-z. But it showing that my gpu is running at 8x gen3. With that in mind I tried to set pci-e speed to gen 4 in motherboard (gigabyte z590m) bios but can't do it. There are only gen 1/2/3 options (see attachments). Maybe I'm looking at a wrong place. Tried googling but cant find anything. Please help Pc specs Gigabyte z590m i5 10400f Asus rx6600 Kingston fury 3200mhz 4x8gb ram Fsp atx-500pnr pro 500w psu

My PC is right next to my window and yesterday it was raining pretty bad while I was at work. My PC wasn’t wet but there was moisture on the glass panel so there was probably some moisture inside as well. I let it dry out and padded some of the components dry. This morning I went to turn it on for the first time and heard a pretty loud pop- sounded kind of like a single firecracker going off. Usually it turns on right away but this time there was a slight delay for all the LEDs to turn on, which is shortly after when the pop happened. I didn’t see smoke or smell anything unusual, and the PC was able to turn on and work- the CPU temps seemed to be normal. I was able to turn it off and on again. My initial thoughts were a blown capacitor or a transistor blowing in the psu from the moisture. Or it could have been something going on with my fans’ LEDs. Does anyone know what it could be, or how I can test/ do diagnostics? My PC is very new I just built it 3 months ago. CPU: RYZEN 7 7700x GPU: RTX 4070ti Motherboard: MSI B650 Tomahawk PSU: 850 W Corsair RAM: 32 GB DDR5 Corsair Mem: 1 TB Samsung M2.2

I reposted this from the graphics cards section cus I really need this to get fixed, here’s the copied version: Hi guys I got a RX 580 4GB eGPU and my graphics card is acting up lately. It refuses to install drivers, and the way it refuses is a blue screen (DRIVER_POWER_STATE_FAILURE). After the blue screen the laptop refuses to log in, and it’s stuck in (Please wait for the Local Session Manager). I tried DDU it multiple times and it just show the same results. Can someone help me fix it? Also while it’s “installing” the drivers I noticed that it’s stuck in 92% (around 00:12 remaining). At that time the device manager picked up the Radeon RX 580 Series. And waiting for another 5 minutes and Mr. Blue Screen is here.

Budget (including currency): $600 - $700 USD Country: United States Games: Pavlov, Onward, VRChat, several other VR games with a Quest 2, Squad, Arma 3, Destiny 2, Cyberpunk, Jedi Survivor Other details: Currently using a laptop with a 3080 and 10750H as my daily driver (Aorus 15P I think). My old desktop has as GTX 1660 Super and 3600x. The games I typically play aren't extremely demanding other than some of the VR titles I play, although I would like something that will give me a consistently smooth experience for both flatscreen and VR games. I'll be pairing this new GPU with an i7-13700K with 32gb DDR5 (I'll be getting the Microcenter Bundle). So far I've considered the following GPUs: RTX 3070 RTX 4070 RX 6900xt RTX 3080ti I've looked at several benchmarks comparing each of these cards against one another and I can't come to a decision on what would be the best value for my money at 1440p. I strongly considered the 6900xt because of how much better it performs against the 4070 at the same price, but I feel like I'd be missing out on DLSS 3 and better ray tracing. I've also used Nvidia cards for years and I'm kind of scared to go red team since I use shadow play a lot, and I know some games tend to favor green team more. I also considered the 30-series GPUs because of how cheap they are used nowadays and how easy they are to get, especially since most of the games I play aren't all AAA titles. TL;DR: I'm unsure what GPU is a good value in 2023 and I need people to tell me how to spend my money.

Just like the title says. Budget (including currency): Preferably under $2000 Country: United States Games, programs or workloads that it will be used for: Modern gaming, recording/streaming, some video editing The parts list is attached, if any parts can be swapped out for better value or performance please let me know. Thank you!

I am looking upgrade my GPU from Vega 11 but I have some limiting factors, Mainly being my PSU and my CPU. I don't want to upgrade anything besides my GPU, What can I get with $200 Budget? (Only New cards) Specs: CPU: Ryzen 5 3400g 4C 8T 3.7Ghz GPU: Vega 11 Motherboard: ASRock A320M HDV-R4.0 (Has PCI-E 3.0) PSU 400 Watt Powerspec NON 80+ RAM: Cosair Vengeance LPX 16GB DDR4 3200 Mhz SSD: 970 Evo Plus 500GB M.2

While I was removing my graphics card to install some NVME storage, I could not help but notice the graphics card coming loose well before I thought it should, and after removing it noticed the locking mechanism was just gone. I don’t recall breaking it or it coming off at all. Just afterwords it being not present. Mysterious vanishing hardware issue aside, I'm now left with a missing lock. I know functionally it should be no issue at all, as long as I don't handle it too roughly and if anything makes it a bit easier. But modern graphics cards are heavy, I don’t like the idea that its missing. What are my options for finding a replacement? Would micro center help? The motherboard is a MSI Z59-A pro.

Hi everyone, I just got a new PSU so I can look at buying a half-decent GPU now. An attractive option I came across was this GTX 1050 on local FB Marketplace. It is approximately the same price as the local going rate for 750Ti's, but it's condition is making me raise some questions. Do you think this would be a reliable purchase?

Hi all, I am trying to repair my Dell Inspiron 5759 laptop and am trying to find an AMD Radeon R5 M335 GPU. Unfortunately, it is an older model and not sold by Dell. Any ideas where I can pick one up? Kind Regards, WeTuLo

Hey, so I got a Gigabyte Aorus RTX 3080 ti gaming box egpu connected to my intel nuc 12 pro, and when I play Fortnite, I get terrible performance. In Fortnite, I'm playing at the highest settings with directx12 at 2k resolution, but I get low 40s and high 30s fps, and it's a 3080 ti it can do those settings. I checked my CPU, and I believe it says it's only at 23 CPU power. Does anyone have any tips on how to fix this terrible performance loss??? here are my setup specs Intel Nuc 12 pro: 12th gen intel core i5 with 12 cores and 16 threads. Iris xe graphics. 32gb of ddr4 ram. 1tb SSD. thunderbolt 4 ports. Gigabyte Aorus RTX 3080 ti gaming box egpu. PSU wattage 550 watts 80 plus gold. thunderbolt 3. GPU water cooling

Hey first time poster just don't know where else to go, My 3090 FE has suddenly started crashing and I can't identify why. After around 10 minutes of gaming both my screens go black and the gpu's fans peg to 100%, I looked around online and saw that the most likely cause of this problem is memory overheating as my gpu temps in task manager were reading normal (around 64 celcius when gaming at 1440p) so I launched HWmonitor and watched my memory temps, They got up to around 100 celcius and then the crash happened. Looked around online and everywhere said the thermal pads needed replacing and the chip needed re-pasting so I watched a bunch of tutorials and got all the recomended pads and thermal paste, took it apart took the old pads off and cleaned the memory with isopropyl alcohol and applied the new pads. Did the same with the chip and closed it all up, booted up and temps were looking way better. I was now idling at around 28-29 celcius on the chip whereas idle before was around 40 celcius and the memory was down to idling at around 40 celcius and before it was idling around 60 so I thought great everything worked and I should be fine, Launched a game and was playing for about 10-15 minutes before the same crash happened again and after turning my pc off and letting it sit for a while and then turning it back on and letting it idle for a while my temps were exactly the same as before I repasted and replaced the thermal pads so i'm really at a loss as to how to fix this or if i just didn't actually make a difference repasting and re-padding the gpu. Any help would be greatly appreciated as I truly have no idea what to do and I live in the uk so there's not really any good pc stores I can take it to for them to look at or fix for me. Sorry for long post, Thanks.

Hi all, I am trying to build a computer and I am getting the error code D6 when turning it on for the first time. Here is the build: ROG Strix Z690-E Gaming WiFi Intel i5-13600K Trident Z5 RGB DDR5 (2x16GB) Gigabyte Gaming OC 12GB 4070Ti Seasonic Prime TX-1000 WD_Black SN850X 1TB NVMe SSD Noctua NH-D15 NZXT H5 Flow When I try posting the machine after building it, the graphics card lights up and the fans begin to spin, then maybe 5 seconds later the fans stop spinning and the lights turn off. The motherboard then finally lands on error code D6. (Note: the only things plugged in are: the power cable, DisplayPort cable, usb keyboard, usb wireless mouse dongle) Things I have tried so far: Unplugging the GPU and using the iGPU as well as unplugging and reseating the GPU. Result: error D6 Using the Seasonic included 12VHPWR as well as the Nvidia branded 12VHPWR to pcie psu cables. Result: error D6 Reseating the RAM as well as using only one stick of RAM. Result: error D6 Reflashing the bios using BIOS Flaskback as well as clearing the CMOS using the Clear CMOS button. Result: error D6 I don’t have a separate motherboard that works to test the GPU, but because I get the same error message when I try to use the integrated graphics on the cpu, I suspect it is not the graphics card fault but instead the motherboard. Is there any other troubleshooting measures I can try to take before RMAing the motherboard? Thank you in advance EDIT: The issue has been resolved but I do not know exactly what fixed it so Ill just say what happened in case someone has a similar problem. After trying different things to no avail, I tried turning off the hybrid mode (no fan) on the power supply and suddenly the iGPU started working. I went through the Windows 11 install process and updated the nvidia drivers (Windows was detecting the gpu but when I tried to switch to it, it did not work). Then I tried moving the HDMI cable plugged into the motherboard down to the GPU (I had primarily been using DisplayPort on the GPU) and it worked. I then tried unplugging the DisplayPort cable and plugging it back in and it finally worked. Im not sure where along the way the GPU started working but hopefully this helps

Hi, been having issues with certain games (Usually more intensive games). The two games i've specifically had issues with are just cause 3 and bullets per minute What ends up happening is the game freezes, music keeps playing, and my GPU usage lowers significantly, from around 90-95% to around 30%. It doesnt seem to happen in less intensive games, I've tried borderlands 2 and it runs fine. Oddly bullets per minute WAS running fine the other day but as of yesterday it's started doing this. Otherwise the games run perfectly fine, well over 60 FPS, no lag or anything just this stuttering issue. I have tried: Uninstalling Graphics drivers using DDU and reinstalling Changing settings in Nvidia control panel, also resetting back to defaults Installing on different drives, Both my WD 2TB hardrive and my Sandisk 1TB SSD Various settings in both games, highest and lowest. No effect Disabling all audio devices that aren't the main one Turning off Ingame stuff like XBOX game bar and Nvidias In game overlay Computer Specs: Windows 10 Pro Sandisk 1 Terabyte Sata Boot drive Evga 500 W1, 80+ White 500W, Power Supply 100-W1-0500-K3 B450 TOMAHAWK MAX Bios version 7C02v3G1 (Latest) Ryzen 5 2600 No overclocking Geforce RTX 3060 No overclocking 2x8 GB Corsair CMZ16GX3M2A1600C9 DDR4 3000 MHz RAM XMP profile 1 Thanks for the help in advance and feel free to ask if i've missed something 2023-04-21 22-51-48.mp4

I've switched back to water cooling after a long hiatus. And I've run into an issue that I haven't experienced before. My GPU is drawing less power i.e. Watts than before, which has impacted the FPS in gaming. This is something I haven't experienced before. The GPU is a XFX THICC III 5700XT Ultra. I did switch out the old thermal pads for thermal grizzly, but I don't think that could cause any issues, as the PC powers on and there are no black screens / hangs / crashes. Google search on this issue only brings up how to lower power consumption, which doesn't directly deal with what I am experiencing. I'm using Hardware Monitor, MSI Afterburner and AMD Radeon to monitor the power consumption, which is driving me nuts. Even a full Furmark / Prime95 Test is not pushing past 60 watts of power consumption. Before, this use to be around the 200-255 watt mark. Specs: Ryzen 5 5600x CPU 32GB Corsair Ram 3600MHZ Ram XFX THICC III 5700XT Ultra (Owned for almost 5 years now) Gigabyte B550 Aorus Elite Motherboard Power: Bequiet Straight Power 11 Gold 750 watt. What I've done so far to try and fix it: 1. Swapping out the previous VGA / PCIE. Before, it was one 8pin to two separate 6+2. Now its 2x 8pins to 6+2. 2. DDU Fresh Drivers / Rollbacks 3. Default Performance settings for the Card 4. Bios Update 5. Turned off / on Re-size Bar and AMD Smart Access Memory I'm just short of draining my water build and opening the Graphics Card to have a look if there is anything unusual. But I'm not sure that's the issue. Everything seems to running well, even as I type this. The issues are in multiple games I've tested over the last 2-3 Days. None of my games, with the lowest settings set, are able to crack 30fps minimum. I'm honestly boggled. Short of draining, reapplying paste and giving it another go, I might have to fork out a few thousand for a new GPU, (Australian Prices for GPUs are shocking) which I'm trying to avoid at all costs. Looking to see if the community has bumped into this issue. I feel like I've attacked multiple issue solving angles here. Not sure. Provide your input and I'll be happy to answer any questions / queries. I'll be checking this throughout the day too. Thanks for the Read!

Please excuse if this is accidentally a duplicate thread, I have done some research but not found a specific answer so far. I see a lot of talk about whether to use a 2, 3 or 4 PCIe connectors to a 16-Pin adapter but nothing about the PCIe adapter vs. a 16-Pin M-M. I was wondering when using a PSU with a 16-Pin connector built in, is there an advantage to using the M-M cable included over a PCIe adapter? Or is there an advantage to using a (3 or 4)PCIe to 16-Pin adapter? If there is, is it a safety advantage, performance etc.? Obviously the 16-Pin M-M would have a cable management advantage but I was wondering if there was others or if they otherwise work identically or is a specific option desired. To create an example to work with, using Cooler Master GX GOLD 1250 ATX 3.0 and the Asus ROG Strix GeForce RTX™ 4090 OC Edition 24GB GDDR6X Appreciate your time, thank you.

I'm very new to Overclocking lets just put that put there lol. However, I have heard the term "Silicon Lottery" used a thousand times. I brought out my old GTX 1060 as I was going to give it to a friend for free who was just getting into PC gaming . When I installed it into his PC he was a bit underwhelmed with his performance in modern titles so I decided to overclock the old thing to see if it could give him some better performance. Over the course of the next couple of hours, I was shocked to see just how far I could push this card! Please let me know is this is a normal experience to have but I will include the overclocking stats below. CORE VOLTAGE: 100% (Included just for completion sake) CORE CLOCK: +240 MHZ MEMORY CLOCK: + 1000 MHZ POWER LIMIT: + 116% TEMP LIMIT: + 92 Degrees Once again the main aspect I am looking at is the core clocks and the memory clocks. Also this thing runs at around 75-80 degrees under sustained load in heaven benchmark for quite some time (mind you I live in Canada and its freezing up here in the winter)

EVGA RTX 2060 6gb Model EVGA208 REV 0.2 Manufactured after 18th August 2005 The card I own is it the same as the the KR versions? Which family of models does it come from Gaming, SC, and or XC?

So, I'll try to explain as clearly as I can, also it's my first time writing here so please be kind. (Also, English isn't exactly my first language) Okay so, my monitor shows "No signal detected" after the windows logo, this only happens if I have my graphics drivers installed and it works normally if I boot into safe mode and delete the drivers. It worked perfectly before, also I use an adaptor (VGA to HDMI) so couple days ago I tried fixing this by removing the graphics card and reinstalling it and it started working again but yeah I had to like move the cable a bit. Then after restarting it once it stopped working again. I thought it was an adaptor issue so I bought a new adaptor and nothing, still having the same issue. Any help? Ah I have tried couple of things so far, reinstalled windows too. So at this point I am starting to think that it's a cable or a port issue, my monitor does support DVI so maybe I can try buying that? Anyway I am here to see if I can try doing something else.

Considering to upgrade from my 2060 Super ROG. Mostly running games, anything else is lower priority. Country is Norway. I tried to flash in a custom vBIOS but some things was weird so I flashed original vBIOS back. Got me some noise maybe from VRM I didn't have before, but also some noise from one of the fans. Sometimes it sounds like it's snoring lol. I've always been kinda dissapointed with the performance of this card and now I got a 165Hz Freesync monitor so I would like a bit more oomph inside the case. Sometimes Windows have had it's Explorer shell gone weird forcing me to press the reset button but I fixed that by turning off power saving on PCIe. Still I suspect it is the graphics card. With this situation in regards to availability and price of cards, It's gonna cost a pretty penny. I'm not sure if it's a relatively worthwhile upgrade though. I can get a XFX Qick 6700 XT for around a 1000€. XFX Radeon RX 6700 XT Speedster QICK 319 Would something like this be a decent upgrade for my rig? Also would it be fine to run in gen 3 x8 or should I first upgrade this cheapo B450 mobo? Ryzen 5 3600x stock. 32 GB ran at 3000 MT/s CL16 (dont seem to be able to get it stable at rated 3200). B450-F Gaming with 2x M.2 so PCIe slot for GPU is currently running at gen 3 8x as mentioned. Corsair RM 650x. Thanks for reading.

As many reviewers mentioned in their reviews of the new NVidia RTX 4070 (NON-Ti) yesterday, each generation of new graphics cards has seen a rise in pricing and I have an idea as to why that is, and unfortunately if this theory pans out, things will only get worse from here... You see, Moore's law has been steadily dying off for quite some time now. Its not completely dead yet, but the advancements in computing horsepower per dollar have significantly slowed down. Yes, the last few generations of Graphics Cards and CPUs have shown rather large jumps in performance, but they have also seen quite large jumps in pricing. There is a reason for this that is unavoidable, allow me to explain: You see, Moore's law is slowing down because we are approaching the absolute physical limits of what traditional silicon transistors can actually do. Years ago when TSMC began work on the 5nm process node now found in current generation hardware, an article came about that stated there could be major delays with manufacturing functional 5nm cores because of some serious engineering challenges. The article stated that TSMC engineers had even spoken about the possibility of 5nm not being feasible at all. This was because they were concerned that the transistors were so small and so dense that it would create a "nuclear core" thermal run-away scenario where cores are producing so much heat so fast with this density that a standard IHS and standard cooling solution just wouldn't be able to remove the heat as fast as the cores create the heat. This would create a situation in which exotic cooling would be required just to cool consumer desktop PC cores. To be honest, in a way, their prediction and concerns are starting to show some truth to them. We have Ryzen 7000 and Intel 13th Gen that both run so hot at full load that they essentially throttle by default, "targeting" their TJMax and "boosting until they hit the limit". Obviously we enthusiasts aren't dumb enough to fall for the marketing. We know well that the cores have simply gotten too hot and maximum performance now has to be throttled to suit each end-users cooling solution. Thats what it really is. Even a 360mm AIO CANNOT properly cool an Intel i9-13900K without some sort of modification or advantage such as direct-die cooling. Properly cool meaning keeping the CPU under 90-95°C during all-core 100% load workloads at default settings. Besides the thermal issues with continuing to shrink transistors ever more, there is the sheer amounts of incredibly expensive technology and engineering involved to even get transistors this small, and the yields aren't great either. Some people across internet forums have begun to speculate that for 3nm and smaller transistors to actually function correctly and not run into nuclear run-away thermal issues, we will be forced to step away from traditional silicon transistors and switch to a more exotic technology - something like carbon nanotubes for example. And I believe they may just be correct in that assumption seeing how many issues there are already - 600 watt GPUs and Flagship CPUs so hot, modifications are required just to keep them under TJMax. So seeing as were are knocking on the door of what is physically possible with standard silicon no matter how advanced the manufacturing process gets, it only makes sense that prices would rise as we reach the very edge and the R&D + engineering required to make further improvements just gets exponentially more expensive and more difficult to pursue. It also explains how the last few generations of hardware just keep getting more power hungry and run hotter. Now a lot of you might point out the fact that the absolute monster that is the RTX 4090 doesn't actually run that hot. Well guys, why do you think this generation of flagship GPUs have coolers the size of actual bricks you use to build houses? Meanwhile, CPU coolers can't really get any bigger because of case compatibility. Imagine Noctua trying to make a 250mm tall version of the D-15, that would be absolutely rediculous. How about an 800mm (200mm x 4) radiator for liquid cooling? It just wouldn't make any sense. So in conclusion, I feel that MSRPs will continue to spiral out of control until the majority of people simply stop upgrading. Then some time can be taken to focus on pure efficentcy and lowering prices. I don't even want to think about how expensive the first carbon nanotube-based CPUs will be. Can you imagine the equivalent of a modern 8-Core typical CPU that costs over $1000? A Ryzen 7700X that costs as much as a flagship GPU? That is the day that PC advancement for the general populace essentially comes to an end for a while. So what do you guys think about the situation on pricing and what do you think will happen in the future? Makes for some pretty interesting thoughts, that much is for sure...

Hi, this is my first time posting here, and i am in need of a few opinions. So basically, i currently have the ASUS ROG Strix GTX 1080 8GB, and my friend is selling his Zotac RTX 3060 12GB. I was thinking about buying it. I would like to know if its worth switching to the RTX 3060. I saw a few benchmarks on youtube comparing it and it seems there is a slight FPS difference. I mainly play F1 22, CSGO, Warzone, and occasionally RUST with a 27" 1080p monitor. Your opinions would be very much appreciated, should i switch to the RTX 3060 or stay on the GTX 1080?

Budget (including currency): Roughly $3800 USD in total. Country: Australia Games, programs or workloads that it will be used for: KVM, Compiling, Programming, Video Editing, Gaming, VR, 3D Modeling, etc. Other details (existing parts lists, whether any peripherals are needed, what you're upgrading from, when you're going to buy, what resolution and refresh rate you want to play at, etc): Wanting to run a KVM virtual machine with PCI passthrough via OVMF. I plan on running Arch Linux as the host operating system, and running Windows 11 as the guest system. Only wanting to run the AM5 platform, and I plan on using a Ryzen 9 7900. (non X) My problem is finding a motherboard that can support 2x GPUs. Preferably 2x RTX 4070 Ti's. Every motherboard that I have come across has a full PCIe 5.0 x16 slot that can run in x16 mode, but finding one that can run another graphics card at x16 seems impossible. Maybe I'm missing something. Just can't seem to find a motherboard that can run both GPUs at x16 speeds/lanes. Most seem to throttle down to x4 speeds/lanes. The PCIe version doesn't bother me. I don't care if it's PCIe 4.0. I just need to find a decent motherboard that can support 2x GPUs that run at full speed, but won't break the bank. (I don't exactly want to spend $1000 on a motherboard.) I mostly know all the other parts I plan on using in my system, but am just struggling to find a motherboard that can support my requirements. Thanks for any recommendations/help!

Should I get the XFX Speedster QICK RX 6650XT or the Sapphire Pulse RX 6650XT? Which one will be better for longevity?

I am getting a 4080 for my birthday however..... i cant really figure out with board partner i should go with. Any ideas?