Editing Graphics card (section)

== Multicard scaling ==
Some graphics cards can be linked together to allow scaling graphics processing across multiple cards. This is done using either the PCIe bus on the motherboard or, more commonly, a data bridge. Usually, the cards must be of the same model to be linked, and most low end cards are not able to be linked in this way.<ref>{{cite web|url=http://www.geforce.com/hardware/technology/sli/supported-gpus|title=SLI|work=geforce.com|access-date=2013-03-13|archive-date=2013-03-15|archive-url=https://web.archive.org/web/20130315133439/http://www.geforce.com/hardware/technology/sli/supported-gpus|url-status=live}}</ref> AMD and Nvidia both have proprietary scaling methods, [[CrossFireX]] for AMD, and [[Scalable Link Interface|SLI]] (since the [[Turing (microarchitecture)|Turing]] generation, superseded by [[NVLink]]) for Nvidia. Cards from different chip-set manufacturers or architectures cannot be used together for multi-card scaling. If graphics cards have different sizes of memory, the lowest value will be used, with the higher values disregarded. Currently, scaling on consumer-grade cards can be done using up to four cards.<ref>{{cite web|url=http://techreport.com/review/19404/sli-vs-crossfirex-the-dx11-generation|title=SLI vs. CrossFireX: The DX11 generation|work=techreport.com|date=11 August 2010 |access-date=2013-03-13|archive-date=2013-02-27|archive-url=https://web.archive.org/web/20130227074444/http://techreport.com/review/19404/sli-vs-crossfirex-the-dx11-generation|url-status=live}}</ref><ref>{{cite web|url=http://www.zdnet.com/blog/hardware/nvidia-geforce-gtx-680-in-quad-sli-configuration-benchmarked/19274|title=NVIDIA GeForce GTX 680 in quad-SLI configuration benchmarked|author=Adrian Kingsley-Hughes|work=ZDNet|access-date=2013-03-13|archive-date=2013-02-07|archive-url=https://web.archive.org/web/20130207021036/http://www.zdnet.com/blog/hardware/nvidia-geforce-gtx-680-in-quad-sli-configuration-benchmarked/19274|url-status=dead}}</ref><ref>{{cite news|url=http://www.maximumpc.com/article/features/head_head_quad_sli_vs_quad_crossfirex|title=Head to Head: Quad SLI vs. Quad CrossFireX|work=Maximum PC|access-date=2013-03-13|archive-date=2012-08-10|archive-url=https://web.archive.org/web/20120810150213/http://www.maximumpc.com/article/features/head_head_quad_sli_vs_quad_crossfirex|url-status=live}}</ref> The use of four cards requires a large motherboard with a proper configuration. Nvidia's GeForce GTX 590 graphics card can be configured in a four-card configuration.<ref>{{Cite web|url=https://www.geforce.com/whats-new/guides/how-to-build-a-quad-sli-system#1|title=How to Build a Quad SLI Gaming Rig {{!}} GeForce|website=www.geforce.com|language=en|access-date=2017-12-11|archive-date=2017-12-26|archive-url=https://web.archive.org/web/20171226074533/https://www.geforce.com/whats-new/guides/how-to-build-a-quad-sli-system#1|url-status=live}}</ref> As stated above, users will want to stick to cards with the same performances for optimal use. Motherboards including ASUS Maximus 3 Extreme and Gigabyte GA EX58 Extreme are certified to work with this configuration.<ref name="geforce.com">{{Cite web|url=https://www.geforce.com/whats-new/guides/how-to-build-a-quad-sli-system#2|title=How to Build a Quad SLI Gaming Rig {{!}} GeForce|website=www.geforce.com|language=en|access-date=2017-12-11|archive-date=2017-12-26|archive-url=https://web.archive.org/web/20171226074533/https://www.geforce.com/whats-new/guides/how-to-build-a-quad-sli-system#2|url-status=live}}</ref> A large power supply is necessary to run the cards in SLI or CrossFireX. Power demands must be known before a proper supply is installed. For the four card configuration, a 1000+ watt supply is needed.<ref name="geforce.com"/> With any relatively powerful graphics card, thermal management cannot be ignored. Graphics cards require well-vented chassis and good thermal solutions. Air or water cooling are usually required, though low end GPUs can use passive cooling. Larger configurations use [[Aqueous solution|water solution]]s or immersion cooling to achieve proper performance without thermal throttling.<ref>{{Cite web|url=http://www.nvidia.com/object/slizone_quadsli.html|title=NVIDIA Quad-SLI{{!}}NVIDIA|website=www.nvidia.com|language=en-us|access-date=2017-12-11|archive-date=2017-12-12|archive-url=https://web.archive.org/web/20171212084703/http://www.nvidia.com/object/slizone_quadsli.html|url-status=live}}</ref>

SLI and Crossfire have become increasingly uncommon as most games do not fully utilize multiple GPUs, due to the fact that most users cannot afford them.<ref>{{Cite web|url=https://www.fudzilla.com/news/graphics/38134-crossfire-and-sli-market-is-just-300-000-units|title=Crossfire and SLI market is just 300.000 units|first=Fuad|last=Abazovic|website=www.fudzilla.com|access-date=2020-03-03|archive-date=2020-03-03|archive-url=https://web.archive.org/web/20200303095645/https://www.fudzilla.com/news/graphics/38134-crossfire-and-sli-market-is-just-300-000-units|url-status=live}}</ref><ref>{{Cite web|url=https://thetechaltar.com/is-multi-gpu-dead/|title=Is Multi-GPU Dead?|date=January 7, 2018|website=Tech Altar|access-date=March 3, 2020|archive-date=March 27, 2020|archive-url=https://web.archive.org/web/20200327112043/https://thetechaltar.com/is-multi-gpu-dead/|url-status=live}}</ref><ref>{{Cite web|url=https://www.techradar.com/news/nvidia-sli-and-amd-crossfire-is-dead-but-should-we-mourn-multi-gpu-gaming|title=Nvidia SLI and AMD CrossFire is dead – but should we mourn multi-GPU gaming? {{pipe}} TechRadar|website=www.techradar.com|date=24 August 2019 |access-date=2020-03-03|archive-date=2020-03-03|archive-url=https://web.archive.org/web/20200303033528/https://www.techradar.com/amp/news/nvidia-sli-and-amd-crossfire-is-dead-but-should-we-mourn-multi-gpu-gaming|url-status=live}}</ref> Multiple GPUs are still used on supercomputers (like in [[Summit (supercomputer)|Summit]]), on workstations to accelerate video<ref>{{cite web|url=https://documents.blackmagicdesign.com/ConfigGuides/DaVinci_Resolve_15_Mac_Configuration_Guide.pdf|title=Hardware Selection and Configuration Guide|website=documents.blackmagicdesign.com|access-date=2020-11-10|archive-date=2020-11-11|archive-url=https://web.archive.org/web/20201111195015/https://documents.blackmagicdesign.com/ConfigGuides/DaVinci_Resolve_15_Mac_Configuration_Guide.pdf|url-status=live}}</ref><ref>{{Cite web|url=https://www.pugetsystems.com/recommended/Recommended-Systems-for-DaVinci-Resolve-187/Hardware-Recommendations|title=Recommended System: Recommended Systems for DaVinci Resolve|website=Puget Systems|access-date=2020-03-03|archive-date=2020-03-03|archive-url=https://web.archive.org/web/20200303095515/https://www.pugetsystems.com/recommended/Recommended-Systems-for-DaVinci-Resolve-187/Hardware-Recommendations|url-status=live}}</ref><ref>{{Cite web|url=https://helpx.adobe.com/x-productkb/multi/gpu-acceleration-and-hardware-encoding.html|title=GPU Accelerated Rendering and Hardware Encoding|website=helpx.adobe.com|access-date=2020-03-03|archive-date=2020-03-03|archive-url=https://web.archive.org/web/20200303032643/https://helpx.adobe.com/x-productkb/multi/gpu-acceleration-and-hardware-encoding.html|url-status=live}}</ref> and 3D rendering,<ref>{{Cite web|url=https://www.pugetsystems.com/labs/articles/V-Ray-Next-Multi-GPU-Performance-Scaling-1559/|title=V-Ray Next Multi-GPU Performance Scaling|website=Puget Systems|date=20 August 2019 |access-date=2020-03-03|archive-date=2020-03-03|archive-url=https://web.archive.org/web/20200303095517/https://www.pugetsystems.com/labs/articles/V-Ray-Next-Multi-GPU-Performance-Scaling-1559/|url-status=live}}</ref><ref>{{Cite web|url=https://www.redshift3d.com/support/faq|title=FAQ {{pipe}} GPU-accelerated 3D rendering software {{pipe}} Redshift|website=www.redshift3d.com|access-date=2020-03-03|archive-date=2020-04-11|archive-url=https://web.archive.org/web/20200411151101/https://www.redshift3d.com/support/faq|url-status=live}}</ref><ref>{{Cite web|url=https://home.otoy.com/render/octane-render/|title=OctaneRender 2020 Preview is here!|access-date=2020-03-03|archive-date=2020-03-07|archive-url=https://web.archive.org/web/20200307090913/https://home.otoy.com/render/octane-render/|url-status=live}}</ref><ref>{{Cite web|url=https://techgage.com/article/autodesk-arnold-render-gpu-beta-performance/|title=Exploring Performance With Autodesk's Arnold Renderer GPU Beta – Techgage|first=Rob|last=Williams|website=techgage.com|access-date=2020-03-03|archive-date=2020-03-03|archive-url=https://web.archive.org/web/20200303095516/https://techgage.com/article/autodesk-arnold-render-gpu-beta-performance/|url-status=live}}</ref><ref>{{Cite web|url=https://docs.blender.org/manual/en/latest/render/cycles/gpu_rendering.html|title=GPU Rendering — Blender Manual|website=docs.blender.org|access-date=2020-03-03|archive-date=2020-04-16|archive-url=https://web.archive.org/web/20200416212930/https://docs.blender.org/manual/en/latest/render/cycles/gpu_rendering.html|url-status=live}}</ref> [[visual effects]],<ref>{{Cite web|url=https://www.chaosgroup.com/vray/nuke|title=V-Ray for Nuke – Ray Traced Rendering for Compositors {{pipe}} Chaos Group|website=www.chaosgroup.com|access-date=2020-03-03|archive-date=2020-03-03|archive-url=https://web.archive.org/web/20200303095516/https://www.chaosgroup.com/vray/nuke|url-status=live}}</ref><ref>{{Cite web|url=https://www.foundry.com/products/nuke/requirements|title=System Requirements {{pipe}} Nuke {{pipe}} Foundry|website=www.foundry.com|access-date=2020-03-03|archive-date=2020-08-01|archive-url=https://web.archive.org/web/20200801044003/https://www.foundry.com/products/nuke/requirements|url-status=live}}</ref> for simulations,<ref>{{Cite web|url=https://test.foldingathome.org/faqs/gpu2-common/frequently-asked-questions-common-ati-nvidia-gpu2-clients-2/multi-gpu-support/|title=What about multi-GPU support?|access-date=2020-11-10|archive-date=2021-01-18|archive-url=https://web.archive.org/web/20210118015357/https://test.foldingathome.org/faqs/gpu2-common/frequently-asked-questions-common-ati-nvidia-gpu2-clients-2/multi-gpu-support/|url-status=live}}</ref> and for training artificial intelligence.