Kog

Description

Builds a hardware‑software inference engine that runs GPUs at high throughput; used by AI developers and teams needing very fast token generation for agentic workflows.

Related

• GroqBuilds cloud infrastructure for fast AI inference using LPU, LPX, and NVIDIA GPUs, serving customers needing scalable inference capacity…
• OLIXBuilds infrastructure components that provide a memory layer for inference, used by developers and systems running frontier AI models to…
• FractileBuilds processors that interleave memory and compute to accelerate inference for frontier AI models, serving thousands of tokens per second…
• Nearfield InstrumentsBuilds automated high-throughput 3D scanning probe metrology systems for process control in the high-end nano-electronics industry, serving…
• DaqstraBuilds software that connects test hardware, unifies data, and orchestrates runs for aerospace, energy, and robotics teams managing…
• Snorkel AIBuilds datasets, evaluation frameworks, environments, and custom AI systems for product and development teams needing domain-specific data…
• U-SpaceBuilds smallsats and constellations and operates complete on-cloud or on-premise mission centers for customers needing fast orbital…
• Cerebras ($CBRS)Cerebras Systems builds AI training and inference platforms, including rack-scale systems for on-premise supercomputers used in medical…
• KERYS SoftwareBuilds a Type‑1 hypervisor that runs multiple locally isolated environments; used by IT, developers, and maintenance teams for secure,…
• Oriole NetworksBuilds all‑optical photonic networking systems for AI and HPC; used to connect GPUs and accelerate AI/ML, algorithmic trading, drug…
• HermeusBuilds fast, powerful aircraft, manufactures F100-PW-229 engines under license, and develops a multipurpose high-Mach mission plane for…
• ModularBuilds composable AI infrastructure, inference endpoints, and open-source language and engine tools for developers deploying text, audio,…