Design a distributed AI model downloaderAI system designGet a new multi-hundred-gigabyte model checkpoint onto thousands of GPU hosts as fast as possible without overloading the storage it comes from.
Last asked Oct 2026
#gpu
- +1Batched LLM inference API over GPUsAI system designDesign an inference API that batches requests across GPUs efficiently. The discussion goes deep on a few axes: utilization, scaling under spikes, and separating prompt processing…Last asked Oct 2026
+5Design a GPU scheduling platformAI system designShare a large GPU cluster among research and production teams: queue training and inference jobs, enforce quotas and priorities, place jobs on topology-aware hardware, and preempt…Last asked Sep 2026- +1Design a video generation service on a scarce GPU poolAI system designUsers submit text-to-video requests that each need minutes of GPU time, and the GPU pool is limited and changes size. Design the request flow, scheduling and workers, how users…Last asked Sep 2026
- Matmul performance modeling on a fixed GPUML designEstimate wall-clock time for matrix multiplications of varying shapes on a given accelerator from arithmetic intensity and memory movement, then extend the model to a multi-layer…Last asked Sep 2026
Schedule video processing across CPU and GPU workersSystem designVideo processing jobs have stages, some suited to CPUs and some to GPUs. Design how jobs are split into stages and scheduled across the two pools of machines to keep both busy…Last asked Aug 2026
Roofline analysis of a BF16 MLP layerML designApply roofline analysis to an MLP layer running in BF16: count its floating-point operations and the bytes it moves, compute its arithmetic intensity, and decide whether it is…Last asked Aug 2026
GPU and inference performance fundamentalsML designAnswer questions on GPU performance: Amdahl's law, HBM versus on-chip SRAM, matrix-multiply cost and tiling, CPU versus GPU execution, tensor and pipeline parallelism, and where…Last asked Aug 2026- Count distinct strings with your own hash table, then on a GPUCodingGiven an array of strings, return how many distinct strings it contains, without using library hash sets or maps: build your own hash table with a hash function and collision…Last asked Jul 2026
Remove a GPU node and reschedule its podsCodingGiven a cluster of GPU nodes running pods, drain one node and reschedule its pods onto the remaining nodes respecting their resource requirements. Reported on 1point3acres'…Last asked Jul 2026
Command-line tool for a fleet of GPU VMsCodingBuild a command-line tool that manages a fleet of GPU virtual machines. Decide its commands, how it keeps track of the machines and their state, and how it reports errors, and…Last asked Jun 2026
Deterministic-order parallel reduction in CUDACodingWrite a parallel reduction on the GPU (for example a sum over a large float array) whose result does not depend on how threads happen to be scheduled: the additions must happen in…Last asked Apr 2026
Find all GPU idle timesCodingFrom GPU task schedules, compute every idle window where no task runs. Reported on 1point3acres' Fireworks AI bank; reconstructed in our own words.Last asked Jan 2026- Kernel-optimization take-home scored by speedupCodingA two-hour take-home that replaces the coding screen. AI tools are allowed. You get a simulator for a small accelerator and a working but slow kernel; your score is the speedup in…Last asked Dec 2025
Design large-scale model inference serving with back-of-envelope sizingAI system designDesign an inference system for a model used by 100 million daily active users. Do the back-of-envelope estimates for memory and bandwidth, then explain how you would reduce…Last asked Nov 2025- Deep learning frameworks and compilers: from NumPy to PyTorch and JAXML designDiscuss where machine learning frameworks are heading: how they moved from NumPy to PyTorch and JAX, and how PyTorch and JAX differ (eager versus traced execution, functional…Last asked Sep 2025
Multiply two matrices across two GPUsML designMatrices A and B must be multiplied using two GPUs, with both matrices split between them. How do you partition the work (by rows, by columns, by blocks), what has to move between…Last asked Sep 2025- +4Design an LLM inference serving platformAI system designDesign the system that serves several large models to many customers with low latency and high GPU utilization.
- Find eligible GPU nodes for a podCodingFilter a fleet of GPU nodes down to the ones that can host a pod with given CPU, memory and GPU requirements. Reported on 1point3acres' Together AI bank; reconstructed in our own…