A new challenger is taking aim at Nvidia’s AI hardware crown. After four years in stealth, chip startup Etched publicly revealed its transformer-only ASIC, the Sohu, alongside an eye-popping financial haul: $800 million raised and more than $1 billion in customer contracts. If the numbers hold, this is one of the most ambitious bets on domain-specific AI silicon to date.

What Happened

Etched’s Sohu chip is purpose-built for transformer inference — the core compute workload behind large language models like GPT-4, Claude, and Gemini. Instead of trying to be a general-purpose GPU, Sohu is an ASIC (application-specific integrated circuit) that strips away everything not needed for transformer operations, promising dramatically better performance and energy efficiency per dollar.

The company says its investor list includes Nobel laureates, deep learning pioneers, and quantitative finance firms — signaling both scientific credibility and capital market confidence. The chip is already entering production, and the $1 billion in signed contracts suggests real demand from hyperscalers or large AI labs.

Etched is not alone in this race. Custom chips like Amazon’s Trainium, Google’s TPUs, and startups like Groq and Cerebras have all tried to dethrone Nvidia. But Etched’s laser focus on transformers — the dominant architecture of today’s AI — and its aggressive funding round make it a particularly interesting contender.

Read the full announcement →

My Take

The Sohu chip is a bet that transformers aren’t a passing trend but the foundation of AI for the next decade. That’s a reasonable bet given the industry’s convergence on the architecture, but it carries risk: if a new architecture (like state-space models or something else) displaces transformers, Sohu becomes a very expensive paperweight.

For developers, the implication is clear: we may soon have specialized hardware that runs transformer inference at a fraction of the cost and latency of GPUs. That could lower the barrier for deploying large models in real-time applications, edge devices, and high-volume APIs. However, it also means fragmentation – code optimized for Sohu won’t run on Nvidia, and vice versa. We’ll need to see how well Etched’s software stack integrates with existing frameworks like PyTorch and TensorFlow.

The $800 million figure is staggering for a stealth startup, but it reflects the magnitude of the opportunity. Nvidia’s data center revenue alone is well over $100 billion annually, and even a slice of that market would justify the investment. Etched’s success will hinge on execution: can Sohu deliver the promised performance gains without making developers rewrite their models from scratch?

What to Watch

  • Benchmark results: Independent performance and energy numbers for Sohu versus Nvidia H100/B200 and AMD MI300X. Until those are public, the claims remain just that.
  • Customer names: The $1B in contracts is huge. Which hyperscaler or AI lab is betting on Etched? If it’s a major player like OpenAI or Google, the market will take note.
  • Software ecosystem: Etched’s compiler and runtime support for PyTorch/JAX will determine how quickly DevOps teams can adopt the chip. Proprietary tools could slow adoption.
  • Transformer evolution: If sparse mixture-of-experts or linear-attention models reduce the computational advantage of Sohu’s architecture, the chip’s edge could erode.