London-based Callosum raised $10.25M to orchestrate AI workloads across heterogeneous chips, challen...
The AMW Read
Updates the infrastructure layer by introducing a multi-chip orchestration player that addresses the structural shift from training to inference-dominated compute economics and sovereign cloud requirements.
London-based Callosum raised $10.25M to orchestrate AI workloads across heterogeneous chips, challenging NVIDIA's GPU monoculture. Their platform delivers 2x accuracy, 7x faster performance, and 4x lower cost by matching each subtask to optimal hardware across Nvidia, AMD, AWS Trainium, Cerebras, and SambaNova. This neuroscience-inspired approach (founded by Cambridge neuroscientists) recognizes that intelligence emerges from diverse specialized systems, not from scaling one architecture. With inference workloads projected to dominate 2/3 of AI compute in 2026 and the inference-optimized chip market reaching $50B+, Callosum's orchestration layer becomes critical infrastructure for enterprises seeking vendor independence and cost optimization. The UK government's ARIA funding signals strategic interest in sovereign cloud capabilities beyond US providers.
#AIInfrastructure #ComputeDiversity #InferenceEconomy #SovereignCloud #MultiCloudAI
