Search the Community

I have a GTX 980ti and I am trying to run a program that need to use openCL. Is there any way to activate it on my PC? I have the latest Nvidia drivers and have had no issues with my pc up until this point. Any help would be muck appreciated, Thanks

I have a R7 240, I wanted to run BOINC, but I can't find any OpenCL runtime support for this graphics card. My OS is Ubuntu 20.10 with Linux Kernel 5.11.0 rc4. System log show this radeon 0000:01:00.0: failed VCE resume (-110). on start up. I've installed Windows7 on this machine, and it works, it shows OpenCL 1.2. I searched through AMD's official suport site, and none of them supports Ubuntu20.10. What's more, neither the drivers supports Ubuntu 20.04 but NOT for R7, nor it only supports Ubuntu 18.04. I've tried to install the drivers for 18.04, and dpkg said that it only supports ubuntu18, and stopped the installation. I've also tried to install the driver for 20.04 and it just shows compile errors and dies. How can I solve this problem??

Hello and thank you for reading this, I Just got myself a 3D printer and was trying to work on some 3D Models on my old Laptop. I tried to use Meshmixer, Lychee Slicer and Photon Workshop but I'm always getting error messages telling me that my GPU doesn't support OpenCL2 (Meshmixer is asking for OpenCL3). My Laptops specs are: Intel i5-460M 2.53GHz AMD Mobility Radeon HD 5850 8GB of basic Ram It is a 17" Acer Laptop I bought back in 2013. All the other basic components are standard Acer OEM parts and I'm running Windows 10. So let me get into some detail. As I described, I couldn't get the necessary software running on my Radeon HD 5850. I disabled my GPU and had the software run on my CPUs integrated graphics which apparently supports OpenCL2. I checked my GPU with CapsViewer before that but couldn't find any support information about OpenCL for my Card. I checked all my driver's and they are up to date as far as I could see. Since it is an old card the latest driver for my GPU I could find was version Radeon 15.7.1. My questions are: How can I get OpenCL2 or newer running on my GPU? Is there maybe some slicing software that only requires OpenCL1? If there is no way to make this work for me, which GPUs should I have a look at to build a budget 3D Modeling PC? Thank you very much for reading this. All the best wishes to you!

Hello all, recently I’ve been able to completely switch from Windows to Linux full time. The only main hurdle I’ve had is getting my AMD GPU to work with Davinci Resolve (which I need for school and work) which fortunately has been resolved (heh heh… see what i did…) by using EndeavourOS (or any Arch distro) and using an OpenCL package from some absolute Chad in the AUR. It works great and I’m superbly happy with it, however, I still have some caveats regarding Arch Linux. I’m still a huge sucker for Point-Release distros, especially Linux Mint (the 21 beta rocks btw) but I have no idea how in the world I’d be able to install the necessary OpenCL packages on a non Arch system. (Obviously for Mint it’d be an Ubuntu based system) Unfortunately, team Red has yet to release another AMDGPU-PRO package for Ubuntu 22.04 so I can’t just install the supported OpenCL stuff from that; essentially… my question is: is there any way to get OpenCL for Resolve installed on Ubuntu based systems or should I just stick with Arch?

Hi, The story is quite simple: one day my DaVinci Resolve gave me this error message: Upon trying to update configuration I discovered that apparently Resolve didn't recognize my discrete GPU at all: The official Blackmagic forum has been less than unhelpful, so I've been wondering if anybody here had the displeasure of experiencing such an error before and might lend me a hand. My system is: AMD Ryzen 7 2700X AMD Radeon RX590 series 16 GB DDR4 RAM Windows 11 Home Evaluation copy. Build 22593.ni_release.220402-1100 (something tells me, that this is the source of the problem). I've tried everything I could by now: reinstalled AMD Adrenalin (multiple times), reinstalled Resolve, tried going back to the previous version of Resolve — no result. I'm hoping somebody knows a workaround for this problem. Help will be greatly appreciated. Thanks everyone in advance!

So, I've been trying to get OpenCL to be un greyed out on my Photoshop for 3 hours and have tried everything from Reddit to Adobe's own community. Here's the deal: GPU-Z says that OpenCL is checked, it's there, but when I open photoshop it says OpenCL is unavailable and in advanced settings it's greyed out, now I also use Affinity, Affinity Photo and Designer and both crash immediately, and the reason I found online was OpenCL. I recently updated to the most modern Game Ready Drivers for my 1070, a card that I believe is still supported, and in GPU-Z it shows that OpenCL exists, I also did a search for OpenCL.dll in my system32 folder, and there were 5 of them? Not sure if that's normal. The thing with photoshop is it clearly detects the graphics card and yet still on OpenCL. Either way I've been trying to fix this forever because I can't make thumbnails, and if I don't fix it soon, I'm f*cked, so any help is majorly appreciated.

what is opencl? is it where you use two gpus ? or.................

Videocardz has just spotted and AMD card in CompuBench sporting 16GB of VRAM, and a clock speed of 1600Mz; along with 64 Compute Units. It looks like another Vega card has cropped up, and has double the VRAM and 400Mhz more on the clock compared to the last TimeSpy leak; where an AMD card had 1200Mhz core clock and 8GB VRAM at 700Mhz ( HBM2 is rated up to 1000Mhz). If clocked at 1600Mhz, it would indicate 13.1 TFLOPS in Single Precision which is not too shabby. https://videocardz.com/69475/amd-radeon-vega-spotted-with-16gb-memory-and-1600-mhz-clock https://compubench.com/device.jsp?benchmark=compu20d&os=Windows&api=cl&D=AMD+6864%3A00&testgroup=info Now before anyone goes and gets a bit hyped we need to look at the VRAM for one. AMD's supplier is SK Hynix, and they have no 8GB stacks ready for production this year it seems; let alone 2.0GBps modules which two are needed of to obtain the 512GB/s Vega is supposed to have. If that's still the case right now, this card found in CompuBench is likely either an Instinct card, or a Radeon Pro; which could allow them to use 4 dies of HBM2, as opposed to only two. Cost is not an option in the professional space, as it is for normal consumers; and HBM2 is rather expensive. We've also seen Raja hold up the Vega die showing 2 HBM2 dies, but that could only be in relation to the consumer gaming cards. Source: SK Hynix Catalogue for 2017 https://www.skhynix.com/static/filedata/fileDownload.do?seq=374 It still a ways off before Computex and we get any solid information directly from AMD sadly. It's all true folks! We have out 16GB, 1600Mhz Vega Card, the Radeon Vega Frontier Edition. http://pro.radeon.com/en-us/vega-frontier-edition/

source: https://www.khronos.org/news/press/khronos-releases-opencl-2.2-with-spir-v-1.2 via: https://www.pcper.com/reviews/General-Tech/Breaking-OpenCL-Merging-Roadmap-Vulkan?utm_source=dlvr.it&utm_medium=twitter += this is one for the history books if it does happen take note to what the Khronos Group writes in the press release that they had with OpenCL 2.2 release: hello! PCPer e-mailed Khronos Group to clarify if the interpretation is correct, this was their answer: there are some questions tho ... OpenCL is not owned by the Khronos Group, but by Apple - is Apple going to sell their rights of the OpenCL API?

source: https://www.khronos.org/opencl/ despite it's name, OpenCL is not an open source standard OpenCL was created by Apple in 2009 and handed to the Khronos Group to maintain it - the standard is open, it means that licenced HW and SW manufacturers can forward proposals to add to the existing standard or propose features for a new version the Khronos Group, for the 1st time, has also released the source code on GitHub to allow manufacturers to fully test their implementation https://github.com/KhronosGroup

Hey LinusTech team, My son is a huge fan of yours Though he's too young to be left unsupervised on an internet forum I find your videos amusing. I especially enjoyed your 8 gamers 1 CPU video and thought I might return the favor. Background : I work at an AI company and we got delivery of new toys today. The servers are the updated SuperMicro SYS-4028GR--TRT2, dual socket with all 10 GPUs being on the same PCIe root complex. I think PLX switches ofc but cant confirm model number. Contrary to popular belief, PCIe switches dont add much latency but this beauty has a penalty of 1 us per switch hop. Having all GPUs on the same root complex is nice for many reasons. I refer the reader to a nice article from Cirrascale -> http://www.cirrascale.com/blog/index.php/exploring-the-pcie-bus-routes/ So in effect by cutting out the Infiniband fabric out of the computation path, we see nice strong scaling across increasing GPU counts. Hugely beneficial for TensorFlow et al 1) The output of nvidia-smi is the most awesome thing I have seen on konsole in the past 2 years J.K. cluster|17:41:59: nvidia-smi 2) The PCIe topology of the GPUs can be easily queried like so : cluster|17:45:40: nvidia-smi topo --matrix 3) You can see in the image above the NVLink field present, though the PCIe version doesnt have it. The specs I Ctrl C+V from the invoice are : 10x PCIex16 NVIDIA® Tesla P100 GPU-NVTP100-16 16GB CoWoS HBM2 PCIe 3.0 -- Passive Cooling Brand NVIDIA / Product Name Tesla P100 Part Number GPU-NVTP100-16 Double-Precision Performance 4.7 TeraFLOPS Single-Precision Performance 9.3 TeraFLOPS Half-Precision Performance 18.7 TeraFLOPS PCIex16 Interconnect Bandwidth 32 GB/s CoWoS HBM2 Stacked Memory Capacity 16 GB CoWoS HBM2 Stacked Memory Bandwidth 720 GB/s Thermal Passive Our other systems on order are these -> http://www-03.ibm.com/systems/power/hardware/s822lc-hpc/ Which have the nice NVLink feature enabled in CPU (IBM POWER8 ) which allows us to page-fault CPU RAM on demand. Useful when your convnet has 10^10 parameters In fact, if one were observing closely, I think OpenPOWER IBM will start to chip away at Intel's dominance in x86 enterprise. 4) Rest of the system specs are not so special as the topology and the pascals. 2x Xeon E5-2643v4 ( you really need high clocked CPUs with non-AVX turbos upto 3.7 GHz to really utilize these GPUs. HIgh core count at low clocks is a strict no-no) 1024 GB RAM, Intel DC S3100 SSDs in RAID 0 for local cacheing & Mellanox ConnectX 4x aggregated EDR infiniband HCAs. 5) Output of deviceQuery ( only the CUDA devs among you will be impressed by this ) : 6 ) Bandwidth Tests : In the 3rd test you see that HBM2 technology shine ( this is not a benchmark ), though it can still be much higher 7) p2p bandwidth & latency [ The inter GPU communication stuff ] This roughly ensures that the added premium the comapny pays for all slots to be on the same root, is actually there. You can see the 1 us addtional hop latency i talked about earlier. 7) And finally a benchmark I'll give just 1, since this isnt a "gaming" GPU though you still can, no game benchmarks. I assure you all it can run Crysis CUDA & NN benchs wont be of interest to this audience. So I give the LuxMark bench v 3.1, which is an OpenCL raytracer. Its a shame there isnt a CUDA version of it, since its common knowledge Nvidia handicaps openCL on their chips. The 1st is from my personal GPU, an EVGA 980 Ti & the 2nd sc-shot is of the server. Enjoy & drool (I cant game on it either, haha)

I plugged my display to my MOBO by accident then to my graphics card, i sw some performance dips in CSGO but nothing too bad but when i open adobe premiere pro i get an error. please help. thank you

Hi there, I do a lot of video rendering using my Ryzen 5 2400G in Blender. My build also has a dedicated GTX 750 Ti in which my display is plugged in. I am curious to know if I can ask Blender to use the integrated Vega Graphics to accelerate rendering while I use the 750 ti for normal use, or can I plug in my display to the Vega Graphics and ask Blender to use CUDA acceleration for rendering. I need answers for both of the above cases. Thanks in advance.

Hi guys, I am about to build an editing rig for myself but I'm on a budget. I heard that having some sort of a GPU makes a big difference in editing. This is not gaming focussed but I'll be playing Rocket league on 1080p 60fps. I have the money for either a GTX 1650 or a RX 570. Initially I wasn't including a GPU in my build and was going with a 2400g. But now I'm getting a R3 1200 as it makes up for the addition for the GPU. I would appreciate if you will help me with my build.

hello I am trying to use my Rx 580 8gb to render video on premier but it won't let me select -Mercury open CL engine (grayed out) I was wondering if someone was able to make it work with the same video card, and if so if I can get help enabling it. Thanks System Specs Intel(R) Core(TM) i3-8100 CPU @ 3.60GHz Package (platform ID) Socket 1151 LGA (0x1) Intel Z370 Mobo (asus ROG) 16 GB DDR4 3200mhz Sapphire Raedon RX 580 8gb Windows 10 Pro Adobe Premiere CC 2019 240 gb SSD Note: Already Tried putting the .txt file with the video card name Also enabling and disabling onboard intel graphics

Hi all, I am desperately trying to get OpenCL to work on my 2 AMD ATI Radeon 5700 series graphics cards in crossfire. Can somebody please help me figure out how to install OpenCL? Thanks so much!

HI, I have a gtx 1080. Issue is that GPU-Z shows no check/tick mark on OpenCL box. Everything else like CUDA, PhysX and Direct Compute 5.0 boxes are checked/ticked. I tried OpenCL test on the asus RealBench but couldn't execute due to a LuxMark-64x.exe error. I tried almost all solutions I could find on the internet especially uninstalling the drivers with DUU on safe mode and then installing latest nvidia driver with a clean installation but nothing solved the problem. Can anybody help me on this?

So i watching videos on how to run ATI and Nvidia GPUs in the same PC. Before building my PC i was determined to buy the GTX 970. But then the 1070 arrived and i thought it'll be just better to wait and use a used GPU as it was the only component remaining in my build. So i opted for a used ATI 7870 which i got for $108. System : CPU - Intel® Core™ i7-6700 Processor MotherboardGigabyte GA-H170-Gaming 3 RAM16Gb G.SKILL Ripjaws V DDR4 2400 C15 (2x8GB) GPUSapphire AMD/ATI HD 7870OC Edition 2 GB GDDR5 CaseCooler Master: CM 690 II (ver.2) StorageSamsung 850 Pro 128 GB SSD, WESTERN DIGITAL 1TB Caviar Green HDD PSUCooler Master: GX 650W Since i use my PC for work and medium-light gaming, i wanted to know whether using ATI and Nvidia both in my PC will benefit anything. Just to be clear i mainly concerned about using software applications like AutoCAD, 3DS Max, Adobe Photoshop, Google SketchUp+Vray. I don't wanna bother myself about the advantages of Physx in Arkham City. So keeping that in mind, is there any benefit of using Nvidia and ATI together, ATI as i primary card and Nvidia as a physx card. Are there any other benefits i can take from using them together for work purposes? Also i watched a few videos about Hybrid PhysX but they are basically concentrated towards gaming. Note: If this set up still works i will opt for a used Nvidia card, again. links: http://physxinfo.com/wiki/Hybrid_PhysX http://www.ngohq.com/graphic-cards/17706-hybrid-physx-mod-v1-03-v1-05ff.html

http://www.anandtech.com/show/9792/amd-sc15-boltzmann-initiative-announced-c-and-cuda-compilers-for-amd-gpus Called the Boltzmann Initiative, AMD is overhauling their entire approach to HPC, especially when it comes to GPGPU acceleration. Today AMD announced it will be implementing CUDA, OpenMP, C++ compilation for its GPUs. The potential boost this brings to AMD's compute performance (as OpenCL is historically a fair deal worse than CUDA in the same workloads, be it for drivers or other reasons) is quite large, especially since theoretically AMD's FirePros have the compute advantage. At least now the cost of switching to AMD will be decreased for the HPC community: less libraries to rewrite, inter-architectural compatibility, etc..) It seems AMD has finally come to its senses about programming standards that have proven more effective than OpenCL. The implications for this are actually great for AMD's gaming prospects too since so much of advanced PhysX is based in CUDA, giving AMD far better access than ever before to Nvidia's feature sets. Being able to compile C++ code directly to the GPU will also save a ton in development time and skill ramp up for new developers. Finally the red team sees reason... Maybe there is hope. EDIT: while not stated in the article itself, I'm pretty sure the only way you can have a commercially used cross-compiler for CUDA is to have a CUDA license. It's a derived fact, not one provided in stone.

Okay first of all i know this question has been asked a lot of times but the more i read the more i get confused. I started with searching for a good GPU, then ended up with two of them in my budget ie r9 390 and gtx 970. But then i came across OpenCL vs CUDA. :unsure: I know Adobe uses CUDA more efficiently but people are also saying that OpenCL will have a better development in future. I am going to buy a new pc with the following specfications Intel i7 6770k asrock z170 motherboard 16gb DDR4 ram and other basic stuff like psu, hdd's etc. Now, this is what I'm going to work on Video editing in Adobe AE and Premiere Pro And also a bit of gaming on 1080p or 1440p. Gaming is not much important for me as much as video editing is. But these days upcoming titles are using much more VRAM than they used to. I won't be able to buy another Computer for another 4-5 years after i buy this. So i really need to decide which one is better to go with. I watched the video of Jayztwocents on AMD R9 390 vs Nvidia GTX 970 Performance per dollar king... and he clearly mentioned that the 390 wins. But here i get confused again with the OpenCL vs CUDA. I really need to get a PC which i can look upto and say "yes this is a good system FOR ME, for the next 4-5 years. Please tell me what should i go with Look upto the VRAM of AMD over Nvidia, keep faith that OpenCL will yeild better outputs with Adobe softwares in future and see to it as a "Future proof" or Trust Nvidia upon CUDA efficiency with softwares like Adobe. :blink: And yes last but not the least, Even if i buy an Nvidia card or AMD i will SLI or Crossfire it in the next 1.5-2 years. Please answer this and solve this topic forever PS- Don't start over the 3.5gb VRAM issue of 970. That will only come into the scenario if i play games like Shadow of Mordor which has HD textures. so keep that topic aside.

Ok, so I'm becoming a streamer. I have a question tho. If I'm rendering somthing(not the game I'm playing at that moment obviously:P) but like the stream itself as it happens, or a video.. Some lower end cards don't say "cuda support" Ik they have cuda cores... Will any graphics card with cuda cores or Stream processors be able to accelerate streaming/video rendering without having to use a display output from it? Can I just throw in a R5 230, not use the HDMI cable, and render a video with it? Or can you only use the card that your using to power the display ? EX: I'm streaming, if it needs gpu horsepower, can I dedicate the R5 to the stream, and my r7 265 to the game I'm playing... or does it not work like that? My ASRock fatal1ty z97 killer supports cf(which I'm not talking about ) So it has a 2nd pci x16 slot, so that is were I would put the r5 230. Can I also play a game, and render a video with the acceleration of my r5 230? Or once again.. only a card that is powering the display? Please awnser my questions I use xplit, twitch, and Camtaisa. Maybe a few side things.. but primarily, can I have the r5 do the work for me?

What is better for rendering videos? (On consumer grade graphics cards like gtx 970 vs r9 290x) What programs are compatible with each technology?

I have an NVIDIA GeForce GPU, and so I know that I can use CUDA cores when rendering, but is it better to use CUDA or OpenCL? What are the advantages and disadvantages of each?

While reading about the 4770k i stumbled upon this paragraph "You can always use OpenCL and code against the HD 4600 graphics OpenCL allows you to utilize the graphics processor to do non-graphics related activities(GPGPU is vernacular) The graphics processor used can be a discrete card or it can be integrated in the CPU." Any idea how much someone would benefit from this and how it can be done ?