Gimlet Labs raises $300M to build a multisilicon inference cloud
Gimlet's Series B
Gimlet Labs has announced a $300M Series B led by Andreessen Horowitz, with participation from Arm, Samsung Ventures, Tiger Global, and others. The company is building a 'multisilicon cloud' for AI inference, disaggregating workloads across GPUs, near-memory compute, dataflow architectures, and CPUs to achieve 3-10x faster performance for frontier workloads. With power as the critical bottleneck, Gimlet claims 5-10x speedups for the same power footprint, addressing the industry's need for both high throughput and low latency.
By breaking up the model and running it across different types of accelerators, we are able to achieve 5-10X speedups for the same power footprint, or similar throughput improvements for the same latency.