Map who's challenging Nvidia
TinySet builds a dataset of AI chip startups. TinyResearch weighs which has the strongest case against Nvidia.
TinySet
Dataset · 50 rowsAI chip startups, with chip name, funding raised, target workload and launch status
| Company | Chip name | Target workload | Launch status |
|---|---|---|---|
| EdgeCortix | SAKURA-II | edge AI | in production |
| Ceremorphic | QS 1 | data center training | taped out |
| FuriosaAI | RNGD | data center inference | in production |
| HyperAccel | Bertha 500 | data center inference | announced |
| Cerebras Systems | WSE-3 | data center training | launched |
| SambaNova Systems | SN50 | data center inference | launched |
| Tenstorrent | Blackhole | data center inference | in production |
| MatX | MatX One | data center training | announced |
TinyResearch
ReportCan any AI chip startup take real share from Nvidia, and which has the strongest case?
Nvidia's dominance in AI chips is real, massive, and still expanding in absolute dollar terms — but the structure of competition is fundamentally changing. Nvidia reported $96.2 billion in quarterly revenue in Q2 FY2027, up 106% year-over-year, with Data Center contributing $89 billion [65]. No challenger comes close to that scale. Yet "taking share" is not a monolithic event: it is happening in segmented, workload-specific, and geography-specific ways simultaneously, and the composition of Nvidia's threat environment is more complex than a startup-vs-incumbent frame captures.
