Microsoft Backs D-Matrix: AI Chip Startup Raises $110 Million in Funding

SA TeamSeptember 5, 2023

Post Views: 4,566

D-Matrix, a pioneering artificial intelligence chip startup headquartered in the heart of Silicon Valley, has successfully secured an impressive $110 million in funding. Notably, this funding round boasts participation from notable investors, including the tech giant Microsoft Corp. This achievement is particularly significant given the current challenges faced by many companies in the semiconductor industry as they seek to secure financial support.

In the highly competitive landscape of AI chip technology, Nvidia stands as a formidable leader. Their stronghold in this market results from an exceptional synergy of cutting-edge hardware and sophisticated software solutions. This robust position, however, has given some potential investors pause when considering backing other AI chip startups. According to sources interviewed by Reuters, the shadow cast by Nvidia’s dominance has influenced investment decisions in certain cases. It’s worth noting that Nvidia has chosen not to provide commentary on this matter.

The Series B funding round was spearheaded by Temasek, a prominent Singapore-based investment firm. Joining them in this endeavor were Playground Global, a venture firm based in Palo Alto, California, and technology giant Microsoft.

Sid Sheth, our CEO, emphasized the significance of this financial support, stating, “This capital infusion comes from experienced backers who understand the intricacies of nurturing a semiconductor enterprise. They possess a proven track record in this domain and are committed to a long-term partnership.”

Sheth revealed that our journey to secure this funding began approximately a year ago in Santa Clara. However, the exact valuation remains undisclosed at this time, although it’s worth noting that we had previously raised $44 million in investments.

D-Matrix specializes in designing chips meticulously crafted to enhance the capabilities of generative AI applications like ChatGPT. Our unique approach involves integrating digital “in-memory compute” technology, which greatly enhances the efficiency of AI computations. Notably, our chip technology is designed to minimize energy consumption during data processing, a vital aspect of generating AI responses, making it exceptionally well-suited for these tasks.

D-Matrix differs from Nvidia in part because its technology focuses on the “inference” element of AI processing rather than competing with Nvidia by developing technology that trains huge AI models.

“We were able to fix the computer architecture,” Playground co-founder Sasha Ostojic remarked. “We have solved the low power requirements and data center needs – (we) built a software stack to deliver the industry’s lowest latency by orders of magnitude.”
Sheth told Reuters that Microsoft has committed to studying the technology for its own usage when it releases next year.

D-Matrix anticipates a revenue of under $10 million for the current year, primarily stemming from customers procuring chips for evaluation purposes. According to Sheth, the company has ambitious plans, expecting to achieve an annual revenue between $70 million to $75 million in just two years, while also reaching a break-even point.