Positron Secures $230 Million in Series B Funding for High-Speed Memory Chips
Semiconductor startup Positron has recently announced that it has successfully raised $230 million in Series B funding, according to exclusive information obtained by JS. The company intends to utilize this capital to accelerate the deployment of its high-speed memory chips, which are considered a crucial component for chips used in AI workloads, as per sources familiar with the situation.
Among the investors in this funding round is the Qatar Investment Authority (QIA), the sovereign wealth fund of the country. Sources have indicated that the QIA has been increasingly focusing on developing AI infrastructure.
The Series B funding for the Reno-based startup comes at a time when hyperscalers and AI companies are looking to lessen their dependence on Nvidia, a long-standing leader in the industry. For instance, OpenAI, one of Nvidia’s major customers, has reportedly been exploring alternatives to Nvidia’s latest AI chips since last year.
Meanwhile, Qatar, through the QIA, is ramping up efforts to establish itself as a key player in sovereign AI infrastructure, a goal that was highlighted at the recent Web Summit Qatar in Doha. The country recognizes the importance of compute capacity in maintaining competitiveness on the global economic stage and is positioning itself as a prominent AI services hub in the Middle East, thus generating interest in startups like Positron.
Qatar’s strategy is already taking shape through significant commitments, including a $20 billion AI infrastructure joint venture with Brookfield Asset Management announced in September.
With this recent funding, Positron has now raised a total of over $300 million in its three-year existence. The startup previously secured $75 million in funding last year from investors such as Valor Equity Partners, Atreides Management, DFJ Growth, Flume Ventures, and Resilience Reserve.
The company’s first-generation chip, Atlas, produced in Arizona, is claimed to deliver performance comparable to Nvidia’s H100 GPUs while consuming less than a third of the power. Positron is specifically focused on inference computing, essential for running AI models in real-world applications, rather than training large language models. This positions the company well as demand grows for inference hardware amidst a shift in focus towards deploying AI models at scale.
Positron’s Strengths Extend Beyond Memory Capabilities
Sources have revealed to JS that in addition to its memory capabilities, Positron’s chips excel in high-frequency and video-processing workloads.
JS has reached out to Positron for further details on this development.