First large-scale Blackwell B200 AI cluster goes live in Canada
Alpha Compute announced on 8 May 2026 the handover of its first large-scale NVIDIA Blackwell deployment: a 504-chip B200 server cluster located in Canada, now in final testing and ready for AI compute customers.
It is one of the first commercial-scale Blackwell clusters to come online outside the hyperscalers. Blackwell-architecture GPUs pack 208 billion transistors on a custom TSMC 4NP process, with two reticle-limit dies connected by a 10 TB/second chip-to-chip link. NVIDIA claims up to 25× lower cost and energy per token for trillion-parameter model inference compared with the previous Hopper generation.
The deployment lands a week and a half before NVIDIA’s Q1 FY2026 earnings on 20 May, where investors will be watching Blackwell ramp commentary, supply constraints, and forward guidance. The Blackwell successor — Rubin — is already on the roadmap for later in 2026.
Sources: GlobeNewswire · NVIDIA earnings preview · NVIDIA newsroom
