Upscale AI raises $190 million to build networking infrastructure for the AI era, hits $2B valuation

2 Sources

Share

Upscale AI closed a $190 million Series A-1 round led by Premji Invest, bringing total funding to $500 million and valuation to $2 billion in under 18 months. The startup is building open-standard networking fabric to connect AI chips from different manufacturers, challenging Nvidia's dominance in AI data center switches as spending approaches $100 billion annually by 2030.

Upscale AI Secures $190 Million in Series A-1 Round

Upscale AI has raised $190 million in a Series A-1 round, pushing its total funding to $500 million and valuation to $2 billion—all achieved in under 18 months.

Premji Invest

, a $15 billion investment fund, led the round with participation from new investors including

Nvidia

,

Salesforce Ventures

,

Temasek

, and Seligman Ventures. Returning backers

Mayfield

,

Tiger Global

, StepStone, Maverick Silicon, and Prosperity7 also joined the funding extension. The Santa Clara-based startup had previously raised $200 million in January during its Series A round led by Tiger Global, Premji Invest, and Xora Innovation.

Source: Fortune

Source: Fortune

Building Open-Standard Networking Fabric for Heterogeneous AI Systems

The company is tackling a critical challenge in AI infrastructure: graphics processing units that power tools like ChatGPT typically can't communicate efficiently unless they're from the same manufacturer. Upscale AI is developing an open-standard networking fabric designed to connect AI chips from different vendors at full speed. "We are trying to get those GPUs to talk the same language," cofounder and CEO

Barun Kar

explained. The company builds hardware, systems, and software that connect AI chips, memory, and storage across one fast network, helping large AI models train and run more efficiently with fewer delays.

Source: Market Screener

Source: Market Screener

Challenging Nvidia's Dominance in AI Data Center Switches

Spending on is forecasted to surpass $100 billion annually by 2030, according to Dell'Oro Group, as Microsoft, Google, Meta, and Amazon race to build out AI infrastructure. The five largest tech companies will spend roughly $660 billion to $690 billion on infrastructure in 2026 alone, nearly doubling what they spent in 2025. Currently, Nvidia's proprietary

NVLink

and

InfiniBand

technologies dominate AI data center networking and lock customers into Nvidia's ecosystem. However, cofounder and Executive Chairman

Rajiv Khemani

sees room for coexistence. "The world in the future is going to be a variety of

heterogeneous AI systems

," he told Fortune. "It's not an Nvidia or our solution. It's a solution where everything coexists with each other."

Positioning as the Cisco for AI Networking Infrastructure

Upscale AI is betting that AI represents the same kind of fundamental computing shift that created networking giants in previous eras. When PCs started connecting to the internet in the 1990s,

Cisco

and Juniper built the plumbing. When the internet moved into the cloud,

Arista

and

Broadcom

took over. "Legacy data center networks were designed for a pre-AI world, not for the massive, tightly synchronized scale-up required by modern AI workloads," Umesh Padval, managing partner at Seligman Ventures, told Fortune. Sandesh Patnam, managing partner at Premji Invest, emphasized the infrastructure gap: "If the compute layer itself was never built for GenAI workloads, look at everything else in that stack. Networking, storage, caching—every single layer of the infrastructure was never built for this workload."

Veteran Leadership and Mounting Competition

Khemani anchors investor conviction as a serial entrepreneur who previously founded Innovium, the only startup in a decade to meaningfully crack Broadcom's lock on networking chip market share before Marvell acquired it for $1.1 billion in 2021. Premji Invest backed Innovium through its late stages. Kar, the company's CEO, was a founding team member at Palo Alto Networks and oversaw Juniper Networks' entire Ethernet portfolio. The two joined forces at Auradine, a blockchain and AI chip startup where Upscale was incubated. However, competitive pressure is mounting. Nexthop AI, a direct rival focused on custom

AI networking

for the same hyperscaler customers, closed a $500 million Series B round in March at a $4.2 billion valuation. Custom chip design at today's most advanced manufacturing nodes costs hundreds of millions of dollars before a product ships, and suppliers increasingly want companies to commit to production slots up to two years in advance. The company has scaled to several hundred employees in under a year and will use the capital to expand its business and speed delivery of its

AI-native networking technology

. When asked whether this is the run to an IPO, Khemani didn't deflect: "An IPO would be an interesting milestone. But we think it's going to be much bigger than that."

Today's Top Stories

© 2026 TheOutpost.AI All rights reserved