2 months ago
London, United KingdomSenior
Responsibilities
- Characterize and simplify communication paths between accelerators by tracing latency and overhead across networking, runtime, memory, and device layers.
- Develop faster data movement across heterogeneous hardware using RDMA, programmable NICs, DPUs, switches, and software-based cross-device memory management.
- Build mechanisms for transferring model state across accelerator types and execution strategies, including KV cache connectors.
- Evaluate scale-up and scale-out fabrics and inform cluster architecture and infrastructure integration.
- Own the technical strategy for reducing communication constraints across heterogeneous accelerator systems.
Requirements
- Deep understanding of networking and communication systems across protocols, hardware, memory systems, and software layers.
- Experience with high-performance data movement such as RDMA, Ethernet fabrics, accelerator interconnects, programmable networking, or similar latency-sensitive systems.
- Strong systems performance instincts and the ability to trace bottlenecks across abstraction layers.
- Ability to question existing abstractions, generalize across unfamiliar technologies, and simplify systems around workload requirements.
Benefits
- Competitive salary determined by skills and experience.
- Equity and ownership.
- Private healthcare.
- Visa sponsorship and relocation benefits.
- In-person work at the London office with provided tools, space, and workplace setup.
Categories
About Callosum
Callosum builds AI infrastructure that unifies heterogeneous compute—software and systems to orchestrate diverse accelerators and the physical clusters they run on—for organizations training and serving models. The privately held company was founded in 2025 and is headquartered in London. It focuses on designing, deploying, and operating mixed-hardware compute fleets across sites and vendors, and raised a seed round in 2026.