The Benchmark Cerebras investment just took a major leap forward, signaling growing conviction around alternative AI chip architectures beyond Nvidia.
This week, Cerebras Systems announced it raised $1 billion in fresh capital at a $23 billion valuation—nearly tripling its valuation from $8.1 billion just six months ago. While the round was led by Tiger Global, a significant portion of the funding came from one of Cerebras’ earliest backers: Benchmark.
According to a person familiar with the deal, Benchmark invested at least $225 million, making the Benchmark Cerebras investment one of the firm’s largest single-company bets to date.
Why Benchmark Raised Special Funds for Cerebras
Benchmark first backed Cerebras in 2016, leading its $27 million Series A when the startup was still an unproven bet on unconventional chip design. Since Benchmark traditionally caps its funds at under $450 million, the firm reportedly raised two separate special-purpose vehicles, both labeled Benchmark Infrastructure, specifically to support the Cerebras investment.
Those regulatory filings underscore just how central Cerebras has become to Benchmark’s long-term AI thesis. Benchmark declined to comment on the deal.
What Makes Cerebras Different From Traditional AI Chips
Cerebras stands out for its extreme approach to hardware design.
The company’s flagship Wafer Scale Engine, unveiled in 2024, measures roughly 8.5 inches per side and contains 4 trillion transistors on a single piece of silicon. Instead of slicing chips from a silicon wafer like traditional processors, Cerebras uses nearly the entire 300-millimeter wafer as one massive chip.
This design enables 900,000 specialized cores to operate in parallel, eliminating the need to move data between multiple GPUs—a major bottleneck in conventional AI clusters. Cerebras claims its architecture allows certain AI inference workloads to run more than 20 times faster than competing systems.
That performance edge is a key reason the Benchmark Cerebras investment continues to scale.
Momentum Builds in the AI Infrastructure Race
The new funding arrives as Cerebras gains traction with some of the world’s largest AI players.
Sam AltmanLast month, the Sunnyvale-based company signed a multi-year agreement worth more than $10 billion to supply 750 megawatts of compute capacity to OpenAI through 2028. The deal is designed to improve response times for complex AI queries, and Sam Altman is also an investor in Cerebras.
Cerebras has been vocal in positioning its systems as faster and more efficient than Nvidia’s GPUs for certain AI workloads—a bold claim in a market Nvidia currently dominates.
IPO Path Clears After Regulatory Hurdles
Cerebras’ journey toward the public markets has not been straightforward.
As of the first half of 2024, G42 accounted for 87% of Cerebras’ revenue, raising concerns due to G42’s historical ties to Chinese technology firms. That relationship triggered a national security review by the Committee on Foreign Investment in the United States (CFIUS), delaying Cerebras’ IPO plans and forcing the company to withdraw an earlier filing in early 2025.
By late last year, G42 had been removed from Cerebras’ investor roster, clearing a major obstacle. According to Reuters, Cerebras is now targeting a public debut in Q2 2026.
Why the Benchmark Cerebras Investment Matters
The size and structure of the Benchmark Cerebras investment highlight a broader shift in AI infrastructure thinking. As demand for compute explodes, investors are increasingly willing to back radically different hardware approaches that promise to break through GPU bottlenecks.
For Benchmark, doubling down on Cerebras isn’t just about one company—it’s a bet that the next phase of AI will require fundamentally new silicon.
Don’t miss out on our latest news—follow us for the latest AI news, breakthroughs, and insights that matter.