AI Infra Interviews logo
← CUDA, Triton & Kernel Engineering / 13
mediumNVIDIA

Sum 100 million floats on the GPU as fast as the hardware allows. Write the kernel and justify each step.

Build a reduction with a checked tail, one partial result per block and a final combine. Count the input bytes, then investigate the instructions, synchronization and numerical behavior that a bandwidth estimate leaves out.

Updated Oct 2026 · Learn how AI infrastructure works through explanations, worked examples and diagrams, then practise applying it to interview questions.

Build a reduction with a checked tail, one partial result per block and a final combine. Count the input bytes, then investigate the instructions, synchronization and numerical behavior that a bandwidth estimate leaves out.

This answer is included with a free account. Sign in to read it and save your progress.

free account · no card needed

Work through the numbers

Change the assumptions in a free calculator, then explain what the result establishes and what still needs measurement.

The concepts behind this question

Use these concept explanations to review the ideas behind this question.

Advanced
💻 Coding for Infra🔒 Premium
Concurrency in Python, Go and C++Concurrency primitives coordinate overlapping work and shared state. Choose threads, processes or asynchronous tasks from the runtime and the work they must perform.
Foundational
⚡ Kernels & Compilers
Kernel FusionKernel fusion combines operations so intermediate results can stay on-chip. It can reduce memory traffic and launch overhead, but may increase resource pressure.
Advanced
🧮 Napkin Math & Capacity🔒 Premium
Bandwidth-Bound Decode ThroughputEstimate decode throughput from bytes per step, then separate aggregate output from per-user pace. A finite-memory example shows how batching, cache reads and quantization change the bound.
Advanced
🧩 GPU & Accelerator Architecture🔒 Premium
NVLink, NVSwitch and PCIeNVLink, NVSwitch and PCIe provide different GPU communication paths. Learn how their topology and bandwidth affect the placement of parallel work.
Explore the full curriculum →
UP NEXT ON YOUR JOURNEY
Next in this trackTranspose a large matrix at close to copy bandwidth. Why is the naive version slow, and what does each fix buy?Next in this trackWhat does torch.compile actually do to your model, and when does it fail to help?Next in this trackWhy do serving engines capture decode steps into CUDA graphs, and what does capture require of the rest of the code?
FEDITOR'S NOTE

Show complete input coverage, a valid shuffle participation mask and block synchronization, then explain how the partials are combined. Separate traffic floors from measured runtime and state the numerical contract.

DISCUSSION · 0

No comments yet — be the first to share your approach.