Moore Threads, a Chinese firm, has declared that it has become the first national company to deliver a local, “fully featured” GPU solution. The startup hired the majority of its core engineering staff from Nvidia, with the usual suspects (Microsoft, Intel, Arm, and others) also on the lookout for technical expertise. “Moore Threads” is more than just a catchy name; the company’s website promises to “increase the number of concurrent threads every two years.”
If that seems familiar, it’s because the papers are called after Moore’s Law, which states that the number of transistors in an integrated circuit doubles every two years. As we know, Moore’s law is named after Gordon Moore, co-founder of Intel.
The company claims that its GPU solutions are based on fully domestic intellectual property and manufacturing expertise, and that they can handle a wide range of GPU workloads, such as 3D graphics computing, AI training and inference computing, high-performance parallel computing, and acceleration for ultra-high-definition video codecs. The business intends for its GPUs to be smoothly integrated with the infrastructure designs of partner companies, including Chinese-developed CPU and platform solutions – an integration process that the company has already begun. Moore Threads believes it already has a plethora of expertise for every stage of GPU production – and that this is what allows it to claim that it can design, produce, sell, and support its GPUs over the full product life cycle.
Interestingly, Moore Threads was formed as recently as October 2020 – a year has passed since its start and development of this GPU solution. The firm has already gone through three investment rounds in a single year, with notable investors such as Sequoia Capital China, ByteDance (of TikTok fame), and Tencent all participating at some time. Moore Threads’ most recent Series A fundraising round totaled $313 million, intending to allow the firm to begin mass production and manufacture of its first GPU chip, as well as continuing IP research and development for its GPU SoCs.
If Moore Threads’ assertions are correct, then this is undoubtedly a watershed point in China’s efforts to become technologically independent of Western countries in terms of silicon-bound technology. With pressure piling on China’s industrial skills in the wake of the US-China trade war, the country’s technical independence appears to be dependent on its ability to reinvent the wheel.