Deep Vision, which develops AI accelerator chips and software for edge computing workloads, raises $35M Series B led by Tiger Global
Kyle Wiggers / VentureBeat :
Context & Ripple Effects
Deep Vision's $35M Series B lands mid-way through a 2021 funding wave for edge-AI silicon: AIStorm raised a $16M Series B for low-powered edge AI chips just months earlier, and Deep Vision now pairs its accelerator hardware with the software stack for edge workloads. Tiger Global's involvement is a repeat move, not a one-off — the firm led Deepgram's $25M Series B earlier the same year, part of the aggressive deployment pace that later drew scrutiny when it marked down holdings like Superhuman and DuckDuckGo in 2023.
The raise matters because edge AI chips are capital-intensive: Deep Vision enters a field where newer entrants like Celestial AI's photonic multichip architecture and Luminous Computing's light-based accelerator have since raised at nine-figure valuations, raising the funding bar for everyone in the category.
First-order effects
- Deep Vision gains $35M to push its AI accelerator chips and edge software toward production deployments, directly competing for the same edge-inference sockets AIStorm is targeting with its low-powered chips.
- Tiger Global adds a second AI infrastructure bet in 2021 alongside its Deepgram lead, concentrating its portfolio in the edge-AI stack rather than applications.
Second-order effects
- AIStorm and other low-power edge chipmakers face a better-capitalized rival that can bundle silicon with software, pressuring them to raise larger rounds or differentiate on power efficiency.
- Tiger Global's repeat leadership of edge-AI rounds signals to other crossover funds that the category is financeable, pulling more late-stage money into a segment previously dominated by specialist hardware investors.
Third-order effects
- If Tiger Global's later markdowns of portfolio companies like Superhuman (down 45%) and DuckDuckGo (down 72%) reflect a broader repricing, the 2021-era funding cadence for AI chip startups resets — leaving only the best-capitalized edge-AI players able to sustain the silicon development cycle.
- The pattern points toward edge AI hardware consolidating around a few vertically integrated chip-plus-software vendors, as customers standardize on stacks that can fund multi-generation silicon roadmaps.
The trend: Edge AI accelerator startups are scaling on crossover capital, with Tiger Global's repeat leadership marking edge inference as a mainstream infrastructure bet — and its later valuation cuts as the test of whether that bet holds.