← ⚡ Kernels & Compilers
Core
Triton Programming Model
Triton replaces CUDA's thread with a program that owns a whole block of data, and replaces manual shared-memory staging and coalescing with a compiler that derives them from block shapes. Pointer arithmetic on vectors, masks for the tail, and program-id swizzling for L2 reuse are the three idioms every Triton kernel is built from, and the live exercise at Anthropic, OpenAI and the serving startups is usually one of a fused softmax, a LayerNorm or a matmul in exactly this style.
a free account unlocks the core curriculum tier · no card
RELATED CONCEPTS
PRACTICE THIS IN REAL QUESTIONS
CUDA, Triton & Kernel EngineeringWhen do you write a kernel in Triton, and when do you have to drop down to CUDA?→Napkin Math, Cost & CapacityA kernel does 4 TFLOP and moves 40 GB in one call. On an H100, is it memory-bound or compute-bound?→CUDA, Triton & Kernel EngineeringWrite a fused row softmax in Triton, explain why it is one HBM pass, and say where it stops scaling.→Open-Weights Models & Serving EnginesYou upgraded the serving engine and throughput fell 15 percent. Find the cause.→Open-Weights Models & Serving EnginesYou must serve a frontier open-weights model on non-NVIDIA accelerators. Plan it.→CUDA, Triton & Kernel EngineeringWrite a CUDA vector add and explain the launch: grid math, the bounds check, and why the copies dominate.→