Standard Kernel
Category: AI Infrastructure
An AI infrastructure startup that builds AI systems to automatically generate ultra-optimized GPU software (kernels) for AI workloads. Standard Kernel was founded in 2025. The company is led by Anne Ouyang. Based in Palo Alto, United States. Team size: 10-50. Total funding raised: $20.0M. Latest round: Seed (Mar 2026). Key investors include Jump Capital, General Catalyst, Felicis, Cowboy Ventures, Link Ventures, Essence VC, CoreWeave, Ericsson Ventures.
- Founded
- 2025
- Headquarters
- Palo Alto, United States
- Team size
- 10-50
- Total funding
- $20.0M
Value proposition
Automates the tedious process of GPU kernel development using AI, unlocking peak hardware performance and efficiency without requiring changes to underlying models or hardware.
Products and solutions
Autonomous Kernel Generation Platform, Adaptive Systems Software (in development)
Unique value
Applies AI to rewrite and optimize the lowest-level software (kernels) that runs AI, replacing a highly manual, scarce-skill engineering process with an autonomous, AI-driven approach.
Target customer
AI-native companies, enterprise partners, and organizations running GPU-intensive AI workloads
Industries served
AI Infrastructure, Enterprise Software, Cloud Computing, Semiconductor & Hardware Acceleration
Technology advantage
Leverages large language models and multi-agent systems to automate kernel development, significantly reducing development cycles and reliance on scarce kernel engineers, while maximizing performance on hardware accelerators like GPUs and TPUs.
How they differentiate
Applies AI to rewrite and optimize the lowest-level software (kernels) that runs AI, replacing a highly manual engineering process with an autonomous, AI-driven approach.
Main competitors
Mako, Herdora, Gimlet Labs
Key partnerships
AI-native companies and enterprise partners (early deployments), Strategic investors including CoreWeave and Ericsson Ventures (indicating potential infrastructure collaborations)
Major milestones
Authored KernelBench, an open-source benchmark for LLM-generated GPU kernels used by Nvidia, Raised $20M Seed Round led by Jump Capital in March 2026
Growth metrics
Early deployments with AI-native companies and enterprise partners
Market positioning
AI infrastructure startup targeting AI-native companies and enterprise partners running GPU-intensive AI workloads
Geographic focus
North America (United States)
Patents and IP
No registered patents disclosed as of latest update. The CEO authored KernelBench, an open-source benchmark for LLM-generated GPU kernels.
About Anne Ouyang
PhD student at Stanford University (on leave). Previously worked on Nvidia's premier kernel engineering team and authored KernelBench. Former Machine Learning Research Development Software Engineer at IBM and Machine Learning Software Engineer at Cadence Design Systems. Holds a BS in Computer Science from MIT.
Official website: https://standardkernel.com/