Back

Callosum Secures $10.25 Million to Enhance Multi‑Chip AI Coordination Software

Founded by Cambridge neuroscience PhDs, Callosum claims its multi‑chip AI scheduler can boost enterprise task accuracy, speed, and cost efficiency through cross‑vendor hardware optimization.

Summary

London‑based AI startup Callosum has raised $10.25 million in a round led by Plural with ARIA R&D support. Founded by two Cambridge neuroscience PhDs, Callosum develops software to schedule AI workloads across Nvidia, AMD, AWS, Cerebras, and SambaNova chips. The system reportedly delivers 2× higher accuracy, 7× faster processing, and 4× lower costs for complex enterprise tasks.

Terms & Concepts
  • AI workload: Tasks or processes run by artificial intelligence models, such as training or inference operations, often requiring high computational power.
  • Chip architecture: The design framework of a computer processor that determines how it executes instructions and manages performance.