Gids-gekoppelde quiz · Middelmatig Niveau

Triton-taal voor aangepaste GPU-kernelsquiz

Bekijk het kernelmodel op blokniveau van Triton, de afhandeling van gemaskeerde grenzen, de startconfiguratie en de behoefte aan gemeten prestatievalidatie.

Vraag 1 van 8

Hoe beschrijft Triton veel GPU‑werk vergeleken met een thread‑per‑thread CUDA‑kernel?

Bekijk elke vraag8 vragen

Antwoorden en uitleg verschijnen terwijl je de quiz doet.

  1. Hoe beschrijft Triton veel GPU‑werk vergeleken met een thread‑per‑thread CUDA‑kernel?
    • As operations over logical blocks of elements
    • As a single CPU instruction for the full tensor
    • As one kernel per individual value
    • As a SQL query over GPU memory
  2. Waarom maskeren vector‑kernels hun laatste blok?
    • To force the last tile to launch fewer warps
    • To prevent loads or stores beyond the valid tensor extent
    • To pad the underlying tensor until its length is divisible by the block size
    • To skip the entire final program instance when a partial block remains
  3. Wat identificeert tl.program_id meestal?
    • The GPU device ordinal that compiled the function
    • The total number of warps in a program instance
    • The index of the active logical program instance in the launch grid
    • The element offset produced by tl.arange within that instance
  4. Waarom kan een getegelde matrix‑kernel het geheugenverkeer verminderen?
    • It reduces the mathematical number of output elements
    • It eliminates every global-memory access
    • It makes all matrices contiguous automatically
    • It reuses loaded values across multiple calculations
  5. Wat moet een benchmark doen met JIT‑compilatietijd?
    • Count compilation as device execution only
    • Separate first-call compilation from steady-state kernel runtime
    • Exclude all warmup and synchronization unconditionally
    • Assume every shape uses the same compiled binary
  6. Why might a Triton kernel be slower than a library operation?
    • Custom kernels cannot use memory tiling
    • Triton always executes on a CPU
    • The library may already use optimized kernels and launch overhead can dominate
    • GPU kernels never benefit from fusion
  7. What can change numerical results between Triton and a reference implementation?
    • The number of code comments
    • The order of Python imports only
    • The tensor's file name
    • Floating-point operation order and precision
  8. When is a custom Triton kernel most defensible?
    • Profiling identifies a bottleneck and a specialized operation can justify maintenance
    • Before measuring any baseline
    • Whenever a tensor contains more than one value
    • When correctness tests are unavailable