Upscale AI has raised $190 million in a Series A-1 funding round, valuing the Silicon Valley startup at $2 billion and bringing its total capital raised to $500 million in less than 18 months. The Santa Clara-based company is developing networking technologies designed to address one of artificial intelligence’s biggest infrastructure challenges: enabling AI chips, memory, and storage systems to communicate more efficiently during the training and deployment of large models. The latest round was led by Premji Invest and included participation from new investors Nvidia, Salesforce Ventures, Temasek, and Seligman Ventures. Existing investors Maverick Silicon, Mayfield, Prosperity7 Ventures, StepStone Group, and Tiger Global also participated. The investment follows a $200 million Series A round completed earlier this year and highlights growing investor appetite for infrastructure companies supporting the rapid expansion of AI computing. Founded in 2024, Upscale AI has developed an integrated hardware and software stack aimed at reducing bottlenecks in AI clusters by allowing processors from different chipmakers to communicate seamlessly. More than 80 companies, including AMD, Intel, Google, Meta, and Microsoft, have reportedly adopted the open standard underpinning the company’s technology. “We are trying to get those GPUs to talk the same language,” said co-founder and CEO Barun Kar. As AI workloads become increasingly complex, traditional data centre networks are struggling to keep pace with the scale and synchronization requirements of modern models. Investors see networking infrastructure as a critical component of the next phase of AI growth. “Legacy data center networks were designed for a pre-AI …
Upscale AI has raised $190 million in a Series A-1 funding round, valuing the Silicon Valley startup at $2 billion and bringing its total capital raised to $500 million in less than 18 months.
The Santa Clara-based company is developing networking technologies designed to address one of artificial intelligence’s biggest infrastructure challenges: enabling AI chips, memory, and storage systems to communicate more efficiently during the training and deployment of large models.
The latest round was led by Premji Invest and included participation from new investors Nvidia, Salesforce Ventures, Temasek, and Seligman Ventures. Existing investors Maverick Silicon, Mayfield, Prosperity7 Ventures, StepStone Group, and Tiger Global also participated.
The investment follows a $200 million Series A round completed earlier this year and highlights growing investor appetite for infrastructure companies supporting the rapid expansion of AI computing.
Founded in 2024, Upscale AI has developed an integrated hardware and software stack aimed at reducing bottlenecks in AI clusters by allowing processors from different chipmakers to communicate seamlessly. More than 80 companies, including AMD, Intel, Google, Meta, and Microsoft, have reportedly adopted the open standard underpinning the company’s technology.
“We are trying to get those GPUs to talk the same language,” said co-founder and CEO Barun Kar.
As AI workloads become increasingly complex, traditional data centre networks are struggling to keep pace with the scale and synchronization requirements of modern models. Investors see networking infrastructure as a critical component of the next phase of AI growth.
“Legacy data center networks were designed for a pre-AI world, not for the massive, tightly synchronized scale-up required by modern AI workloads,” said Umesh Padval, Managing Partner at Seligman Ventures.
The fresh capital will be used to expand operations and accelerate commercialisation of the company’s AI networking products. Upscale AI has already grown to several hundred employees within roughly a year of its founding.
The company is led by executives with deep semiconductor and networking experience. Co-founder Kumar Khemani previously founded networking chipmaker Innovium, which was acquired by Marvell for $1.1 billion in 2021. He also held leadership positions at Intel and Cavium.
The latest funding round underscores the intense race to build the infrastructure powering the AI economy. While much attention has focused on chips and foundation models, investors are increasingly backing companies building the underlying networking, memory, and data movement technologies required to support next-generation AI systems.
With demand for AI computing expected to accelerate, startups such as Upscale AI are emerging as key players in the broader ecosystem supporting the future of data centres and high-performance computing.
