AI chip startups are attacking Nvidia from DRAM, interconnects and new substrates
FinanceYF5 · x · 2026-07-27
- The post summarizes a 2026 map of next-generation AI chip startups and the different ways they try to attack Nvidia’s dominance.
- The listed strategies include eliminating DRAM, removing chip-to-chip interconnects, collapsing the compute/memory split, abandoning general-purpose designs, or even sidestepping a $400M lithography machine.
- The image expands the list with companies in optical, thermodynamic, and superconducting compute, and notes roughly $18B of value now sitting on the interconnect side.
- It frames the field as a search for radically different substrates and data-movement architectures rather than incremental GPU competition.
Related event: AI Chip Startups Target Nvidia's Dominance with $58B Valuation(4 posts)→
More from Infra
- 12 KV Cache Reduction Techniques Every AI Engineer Should Understand, Explained — blaizedsouza · 2026-09-11
- The shadow GPU capacity market is formalizing, with Meta selling excess compute to outside buyers — DavidLinthicum · 2026-09-11
- Engram's random reads don't suit SSDs; CPU-memory over NVLink could serve all 72 GPUs — bookwormengr · 2026-09-11
- 80% of the DIY LLM inference hype posters have already quit — it's brutally hard systems work — abhijithneil · 2026-09-11
- Hugging Face's Ultra Scale Playbook: a free book on training LLMs on GPU clusters — mdancho84 · 2026-09-11
- Is inference latency becoming the biggest bottleneck for production AI agents? — Euphoric_Sea632 · 2026-09-11