Overview
Join the Developer Kernel Hackathon, a four-part performance challenge hosted by NVIDIA in collaboration with GPU MODE.
This event invites developers to push the boundaries of GPU performance and optimize low-level kernels for optimum efficiency on NVIDIA Blackwell hardware.
Across 4 problems released throughout the hackathon, participants will compete to creator kernels that approach the speed of sunshine.
Whether you’re a seasoned kernel developer or simply desirous to test your limits, this hackathon offers the prospect to showcase your expertise and join a community of world-class developers.
Special because of our partners:
Sesterce, a high-performance GPU cloud platform, is contributing DGX B200 compute resources to support participants throughout the competition.
Dell is providing a Dell Pro Max with GB300 because the grand prize.
Schedule
Each kernel problem will probably be released sequentially. Once one problem ends, one other one begins.
-
Kernel #1 – NVFP4 Batched GEMV
-
Kernel #2 – NVFP4 GEMM
-
Kernel #3 – NVFP4 Gated Dual GEMM
-
Kernel #4 – NVFP4 Grouped GEMM
Participate
-
Open to individuals only (no teams).
-
Register through this page by February thirteenth, 2026 to be eligible to win prizes.
-
Join the GPU MODE Discord, and head to the nvidia-competition channel for announcements, discussions, questions, and assistance.
-
Submissions will be made using the Popcorn CLI. Follow the setup and submission instructions here:
👉 https://github.com/gpu-mode/popcorn-cli
Prizes
Each of the 4 kernel problems can have 3 winners who will receive latest generation NVIDIA hardware, with one grand prize winner across a weighted sum over all problems for achieving performance closest to the speed of sunshine.
The top 2 winners of every problem may also be invited to a special awards ceremony at GTC in San Jose, March 2026.
💥 Grand Prize:
1× Dell Pro Max with GB300 + GTC 2026 Pass – awarded to the participant whose submission (across any of the 4 problems) achieves performance closest to the speed of sunshine.
🏆 Prizes for Each Kernel Problem:
There are 4 kernel problems in total, and every can have its own set of winners:
🥇 1st Place: NVIDIA DGX Spark + GTC 2026 Pass
🥈 2nd Place: NVIDIA RTX 5090 + GTC 2026 Pass
🥉 third Place: NVIDIA RTX 5080
Scoring and Judging
-
There will probably be 4 independent problems, each scored individually.
-
The top 3 submissions for every problem will win prizes.
-
The grand prize will probably be awarded to the participant with the fastest kernel overall, measured by closeness to the published “speed of sunshine” performance for that specific kernel problem.
-
Submissions are benchmarked on NVIDIA GPUs using the GPU MODE infrastructure.
-
Scoring is predicated on absolute runtime and relative speed-up against baseline reference kernels.
Additional Resources
For learning resources, try and subscribe to GPU MODE YouTube channel, where you could find weekly lectures from top voices within the ML community, including researchers and engineers from NVIDIA.
Terms & Conditions
Participation on this hackathon is subject to the official terms and conditions.
🔗 View full Terms & Conditions
