Kog claims 30x faster LLM inference on existing GPUs through low-level reverse engineering.

The French startup argues the belief that GPUs suit agentic workflows poorly is a misconception, and says its Kog Inference Engine reaches up to 30x decoding speedups on standard NVIDIA and AMD datacenter GPUs with no new hardware. The approach is hardware-aware optimization of memory bandwidth use, not a new architecture. TechCrunch Treat 30x as a vendor claim on selected workloads until someone benchmarks it independently. The thesis directly contradicts the custom-inference-silicon narrative, which is why it's worth watching either way.