Advertisement
SenseTime to expand computing power amid surging AI model demand
Co-founder Yang Fan expects capacity to grow by up to triple digits annually over the next two years amid surging generative AI demand
Reading Time:2 minutes
Why you can trust SCMP
0
Ann Caoin Shanghai
SenseTime, an artificial intelligence (AI) pioneer in China, is set to expand its computing power by up to a triple-digit percentage annually in the next two years, as the company continues its efforts to increase the use of domestic chips amid an intensified tech war.
Yang Fan, co-founder of SenseTime and president of the SenseCore business group, the company’s AI infrastructure unit, said the company’s computing power would maintain a “high double-digit to triple-digit” annual growth rate in the coming 24 months, signalling a strategy to capitalise on surging demand for generative AI models.
In 2024, the total computing power operated by SenseCore grew by 92 per cent year on year to over 23,000 petaflops. One of its major efforts in recent years has been to increase the adoption of home-grown chips to mitigate risks from the ongoing US-China tech war.
The company launched an upgraded version of SenseCore on Thursday, featuring better performance in computing, among other areas, as well as some industry-wide solutions aimed at accelerating the commercialisation of its infrastructure.
The move reflects SenseTime’s efforts to capitalise on surging AI demand, fuelled by OpenAI’s GPT models and more recently the open-source models from China’s DeepSeek, as it targets its first full-year profit in 2026.
SenseTime, founded in Hong Kong in 2014 and best known for its AI and computer vision software, is also an early mover in building up computing infrastructure. The company began this effort as early as 2018, according to Yang.
Advertisement