OptimizeBenchmarkRepeat
A platform for GPU programming challenges. Write efficient GPU kernels and compare your solutions with other developers.
Want to submit from your IDE? Check out our CLI tool
vector-add.cu
#include <cuda_runtime.h>
__global__ void vectorAdd(const float* A, const float* B, float* C, int N) {
}
Why Tensara?
Tensara provides a unique platform for honing your GPU programming skills through competitive challenges and detailed benchmarking.
Real Hardware Benchmarking
Submissions are run on standardized GPU hardware for fair and accurate performance comparisons.
Competitive Leaderboards
See how your solutions stack up against others on detailed leaderboards for each problem.
Community & Collaboration
Discuss strategies, share insights, and learn from fellow GPU programming enthusiasts.
Powered by
© 2026 Tensara. All rights reserved.
Metric Migration
We're migrating our primary metric from FLOPS to runtime. More info on the changes we've made will be posted soon.