Market — GPU Cloud & NeoCloud

Verified claims and product-axis read for **GPU Cloud & NeoCloud**. Every fact below is sourced; every product judgment traces back to underlying signals.

Top products (engine read)

Dedicated GPU compute capacity (neocloud contracts) — neocloud capacity contract

**Opportunity:** Hyperscaler GPU shortage is pushing Microsoft, OpenAI etc. to multi-billion long-term offtake from pure-play GPU clouds; market is asking who the next tier of neocloud winners are beyond the big 4.

This is the headline product of the topic. Nebius just locked a ~$17.4B Microsoft offtake, CoreWeave is so cash-hungry it cut its liquidity floor to $100M, and retail discourse is openly hunting for the next neocloud winner. Hot, with concrete order-book evidence.

GPU cloud price index / cross-provider pricing comparison — price index / comparison tool

**Opportunity:** V100 hourly prices swing from $0.05 to $3.06 across 25 providers — buyers can't tell who is fair. There is an unmet demand for a neutral price index / comparison tool covering hyperscalers and neoclouds.

Strong, repeated demand signal (12 distinct queries about GPU cloud pricing comparison), but no company in this evidence batch is shipping the index. White-space opportunity adjacent to the neocloud boom.

GB200 NVL72 rack-scale Blackwell systems — AI accelerator rack

**Opportunity:** Neoclouds are racing to be first-to-market on GB200 NVL72 capacity; Crusoe is using Iceland clean power as differentiation. Customers want latest-gen Blackwell on tap rather than waiting on hyperscaler queues.

Hot deployment vector for neoclouds — owning early Blackwell rack inventory is the wedge to win the next tier of training contracts. Crusoe+SMCI+NVIDIA visible in this batch.

Decentralized / distributed GPU cloud — distributed GPU marketplace

**Opportunity:** An emerging counter-positioning to the CapEx-heavy neocloud model: aggregate idle consumer/edge GPUs instead of building $B data centers. Beginning to land 'real workloads' per community discussion.

Adjacent / contrarian play to the neocloud category. Single signal so weak in this batch, but worth flagging as a structural alternative if CapEx-funded neoclouds face a downturn (cf. AI-bubble queries in same batch).

Cross-cutting opportunities (industry read)

• **800G/1.6T Silicon Photonics & Co-Packaged Optics (CPO)** — Pluggable copper SerDes is at its reach limit for rack-scale GPU fabrics; 800G→1.6T and CPO are the only paths to keep up with NVLink/UALink rack densities.
• **AI-Optimized Ethernet Switching Fabric (RoCEv2, lossless)** — Hyperscalers want to break NVIDIA/Mellanox InfiniBand lock-in on GPU back-end fabrics; RoCEv2 + UEC-style lossless Ethernet is the consensus alternative but requires retimers, congestion control, and NCCL tuning to match IB.
• **Direct-to-Chip Liquid Cooling & CDU (100kW+/rack)** — Rack densities have crossed the 100kW threshold; air can no longer remove heat from H100/B100/B200 racks. Every colo + every cooling vendor needs a direct-to-chip or rear-door HX SKU on the shelf.

_See [Products](./products) for the full product list and [Strategy](./strategy) for forward-looking judgment._