Notes on verified
GPU optimization.
Engineering perspectives on verification, benchmarking, memory bottlenecks, and why a faster kernel is only useful when it's provably correct.
Why AI GPU optimization needs verification
An LLM can generate a faster kernel in seconds. Proving it still computes the right answer is the hard part — and the part that matters.
Read articleHow benchmark noise hides GPU regressions
A single latency number is a coin flip with extra steps. Here's why Kernalin returns "credibly faster / equivalent / regressed" instead of "faster / slower."
7 minUnderstanding GPU memory bottlenecks
Most "slow" kernels aren't compute-bound — they're wasting bandwidth. A tour of the seven ways VRAM quietly costs you money.
8 minTriton vs CUDA — where each one wins
Triton buys you productivity and portability; hand-written CUDA buys you the last mile. A pragmatic take on when to reach for which.
5 minThe economics of GPU optimization
A kernel that runs 30% faster isn't a performance win — it's a line item. How to think about optimization as capacity you get back.
5 minWhy reward hacking is dangerous in code optimization
When you optimize for a metric, you get the metric — not the thing you meant. In GPU code, that gap is where the bugs live.
6 minArticles are sample content illustrating Kernalin's engineering perspective.
Optimize with AI. Prove it with Kernalin.
Join the early-access program and bring verification-first optimization to your GPU workloads.