Editing Graphics processing unit (section)

==={{anchor|INTEGRATED|Integrated graphics|iGPU}}Integrated graphics processing unit===
[[File:Motherboard diagram.svg|thumb|The position of an integrated GPU in a northbridge/southbridge system layout|616x616px]]
[[File:A790GXH-128M-Motherboard.jpg|thumb|An [[ASRock]] motherboard with integrated graphics, which has HDMI, VGA and DVI-out ports|400x400px]]

''Integrated graphics processing units'' (IGPU), ''integrated graphics'', ''shared graphics solutions'', ''integrated graphics processors'' (IGP), or ''unified memory architectures'' (UMA) use a portion of a computer's system RAM rather than dedicated graphics memory. IGPs can be integrated onto a motherboard as part of its [[Northbridge (computing)|northbridge]] chipset,<ref>{{Cite web|url=https://www.tomshardware.com/picturestory/693-intel-graphics-evolution.html|title = Evolution of Intel Graphics: I740 to Iris Pro|date = 4 February 2017}}</ref> or on the same [[die (integrated circuit)]] with the CPU (like [[AMD APU]] or [[Intel HD Graphics]]). On certain motherboards,<ref>{{cite web|url=https://www.gigabyte.com/products/product-page.aspx?pid=3785#ov|title=GA-890GPA-UD3H overview|url-status=dead|archive-url=https://web.archive.org/web/20150415095629/https://www.gigabyte.com/products/product-page.aspx?pid=3785#ov|archive-date=2015-04-15|access-date=2015-04-15}}</ref> AMD's IGPs can use dedicated sideport memory: a separate fixed block of high performance memory that is dedicated for use by the GPU. {{As of|2007|alt=As of early 2007}} computers with integrated graphics account for about 90% of all PC shipments.<ref>{{cite web |author=Key |first=Gary |title=AnandTech – μATX Part 2: Intel G33 Performance Review |url=https://www.anandtech.com/mb/showdoc.aspx?i=3111&p=23 |url-status=live |archive-url=https://web.archive.org/web/20080531045027/https://www.anandtech.com/mb/showdoc.aspx?i=3111&p=23 |archive-date=2008-05-31 |work=anandtech.com}}</ref>{{update inline|date=February 2013}} They are less costly to implement than dedicated graphics processing, but tend to be less capable. Historically, integrated processing was considered unfit for 3D games or graphically intensive programs but could run less intensive programs such as Adobe Flash. Examples of such IGPs would be offerings from SiS and VIA circa 2004.<ref>{{cite web |author=Tscheblockov |first=Tim |title=Xbit Labs: Roundup of 7 Contemporary Integrated Graphics Chipsets for Socket 478 and Socket A Platforms |url=https://www.xbitlabs.com/articles/chipsets/display/int-chipsets-roundup.html |url-status=dead |archive-url=https://web.archive.org/web/20070526124817/https://www.xbitlabs.com/articles/chipsets/display/int-chipsets-roundup.html |archive-date=2007-05-26 |access-date=2007-06-03}}</ref> However, modern integrated graphics processors such as [[AMD Accelerated Processing Unit]] and [[Intel Graphics Technology]] (HD, UHD, Iris, Iris Pro, Iris Plus, and [[Intel Xe#Xe-LP (Low Power)|Xe-LP]]) can handle 2D graphics or low-stress 3D graphics.

Since GPU computations are memory-intensive, integrated processing may compete with the CPU for relatively slow system RAM, as it has minimal or no dedicated video memory. IGPs use system memory with bandwidth up to a current maximum of 128&nbsp;GB/s, whereas a discrete graphics card may have a bandwidth of more than 1000&nbsp;GB/s between its [[Video random access memory|VRAM]] and GPU core. This [[memory bus]] bandwidth can limit the performance of the GPU, though [[Multi-channel memory architecture|multi-channel memory]] can mitigate this deficiency.<ref name="Coelho">{{cite web |last1=Coelho |first1=Rafael |title=Does dual-channel memory make difference on integrated video performance? |url=https://www.hardwaresecrets.com/dual-channel-memory-make-difference-integrated-video-performance/ |website=Hardware Secrets |access-date=4 January 2019 |date=18 January 2016}}</ref> Older integrated graphics chipsets lacked hardware [[Transform, clipping, and lighting|transform and lighting]], but newer ones include it.<ref>{{cite web |author=Sanford |first=Bradley |title=Integrated Graphics Solutions for Graphics-Intensive Applications |url=https://www.amd.com/us-en/assets/content_type/white_papers_and_tech_docs/Integrated_Graphics_Solutions_white_paper_rev61.pdf |url-status=live |archive-url=https://web.archive.org/web/20071128165723/https://www.amd.com/us-en/assets/content_type/white_papers_and_tech_docs/Integrated_Graphics_Solutions_white_paper_rev61.pdf |archive-date=2007-11-28 |access-date=2007-09-02}}</ref><ref>{{cite web |author=Sanford |first=Bradley |title=Integrated Graphics Solutions for Graphics-Intensive Applications |url=https://www.techspot.com/news/46773-amd-announces-radeon-hd-7970-claims-fastest-gpu-title.html |url-status=live |archive-url=https://web.archive.org/web/20120107200529/https://www.techspot.com/news/46773-amd-announces-radeon-hd-7970-claims-fastest-gpu-title.html |archive-date=2012-01-07 |access-date=2007-09-02}}</ref>

On systems with "Unified Memory Architecture" (UMA), including modern AMD processors with integrated graphics,<ref>{{Cite web |last=Shimpi |first=Anand Lal |title=AMD Outlines HSA Roadmap: Unified Memory for CPU/GPU in 2013, HSA GPUs in 2014 |url=https://www.anandtech.com/show/5493/amd-outlines-hsa-roadmap-unified-memory-for-cpugpu-in-2013-hsa-gpus-in-2014 |access-date=2024-01-08 |website=www.anandtech.com}}</ref> modern Intel processors with integrated graphics,<ref>{{Cite web |last=Lake |first=Adam T. |title=Getting the Most from OpenCL™ 1.2: How to Increase Performance by... |url=https://www.intel.com/content/www/us/en/developer/articles/training/getting-the-most-from-opencl-12-how-to-increase-performance-by-minimizing-buffer-copies-on-intel-processor-graphics.html |access-date=2024-01-08 |website=Intel |language=en}}</ref> Apple processors, the PS5 and Xbox Series (among others), the CPU cores and the GPU block share the same pool of RAM and memory address space. This allows the system to dynamically allocate memory between the CPU cores and the GPU block based on memory needs (without needing a large static split of the RAM) and thanks to zero copy transfers, removes the need for either copying data over a [[bus (computing)]] between physically separate RAM pools or copying between separate address spaces on a single physical pool of RAM, allowing more efficient transfer of data.