Analysis examines rise of Chinese GPU manufacturers driven by AI demand and national tech sovereignty initiatives.
According to Jon Peddie Research, there are 18 GPU developers worldwide today. This represents a significant evolution from the competitive landscape of previous decades. Tens of companies developed graphics cards and discrete graphics processors in the 1980s and 1990s, but intense competition for the highest performance in 3D games drove the vast majority of them out of business. By 2010, only AMD and Nvidia could offer competitive standalone GPUs for gaming and compute, while others focused either on integrated GPUs or GPU IP.
The composition of today's 18 GPU developers breaks down as follows: two companies develop SoC-bound GPUs primarily for smartphones and notebooks, six operate as GPU IP providers, and 11 are GPU developers focused on GPUs for PCs and datacenters, including AMD, Intel, and Nvidia. This growth has been particularly pronounced among China-based companies. The mid-2010s found the number of China-based PC GPU developers increasing rapidly, fueled by the country's push for tech self-sufficiency as well as the advent of AI and HPC as high-tech megatrends. Additional China-based companies like Biren Technology and Tianshu Zhixin are solely focused on AI and HPC for now, so Jon Peddie Research does not consider them GPU developers, though if included, there would be even more GPU designers.
China's investment in GPU development is driven by multiple factors. As Jon Peddie, the head of JPR, explained: "AI training was the big motivator [for Chinese GPU companies], and avoidance of Nvidia's high prices, and (maybe mostly) China's desire for self-sufficiency." From a technical standpoint, GPUs are inherently parallel, meaning there are loads of compute units inside that can be used for redundancy, which makes it easier to get a GPU up and running. Since GPUs are fundamentally parallel, it is also easier to parallelize them in a scale-out manner. Even if Chinese GPU developers lose access to TSMC's advanced nodes, at least some of them could still produce simpler GPU designs at SMIC to address AI/HPC and gaming/entertainment markets. From China's perspective as a country, AI and HPC-capable GPUs may be arguably more important than CPUs, since these technologies can enable new applications such as autonomous vehicles and smart cities.
Despite the growing number of GPU developers, meaningful competition remains limited. While there are many GPU developers, only two can actually build competitive discrete GPUs for PCs. This constraint exists because while it is relatively easy to develop a GPU architecture, it is truly hard to properly implement it and to design proper drivers. CPU and GPU microarchitectures are essentially at the intersection of science and art, representing sets of sophisticated algorithms that can be developed by rather small groups of engineers, but they might take years to develop.