🧠 Cloud & Infrastructure Watch
Gimlet Labs emerged with an $80 million Series A to build a multi-silicon inference cloud that splits AI workloads across CPUs, GPUs, and high-memory systems to improve utilisation and reduce inference waste.
Read MoreHelion confirmed Sam Altman is stepping down as board chair while the company explores a potential power-supply deal with OpenAI that could reserve a large share of future fusion output for AI workloads.
Read MoreNvidia and Emerald AI announced plans with major US power companies to develop flexible AI factories that can modulate energy demand in real time instead of running at fixed peak load.
Read MoreAlphabet president Ruth Porat warned that US energy infrastructure may not be expanding fast enough to keep up with AI demand, reinforcing power as the central bottleneck in the next infrastructure phase.
Read More
💻 Hardware & Semiconductor Moves
Nvidia’s next-gen LPU-based LPX rack revealed to reach ~160kW per rack, signalling a step-change in high-density AI infrastructure design.
Read MoreUpstage entered discussions with AMD to deploy 10,000 MI355 GPUs across South Korea for large-scale AI workloads.
Read MoreEleveight AI deployed 512 Nvidia B300 GPUs at a 2MW data centre in Armenia, expanding frontier GPU cluster footprints.
Read MoreElliott Management took a multi-billion-dollar stake in Synopsys, targeting profitability in a critical AI chip design software provider.
Read More
