GPU Technology

GPU Technology

Deep technical articles on this topic.

591Articles
591Topics covered
Articles in this category

All 80 articles, sorted alphabetically

Advertisement
ARTICLE · 01

Coalesced Memory Access

Consecutive threads read consecutive addresses = one transaction.

Read article
ARTICLE · 02

CoreWeave

Fastest H100 availability. IPO 2025.

Read article
ARTICLE · 03

Ada Lovelace / Blackwell architecture

Deep-dive on NVIDIA Hopper→Blackwell arc: SM redesign, FP8/FP4, memory bandwidth, NVLink generations, GB200 Grace-Blackwell.

Read article
ARTICLE · 04

Admission Control for LLM Serving

Reject requests to protect SLOs.

Read article
ARTICLE · 05

AI Gateway Overview

AI gateway: central proxy for LLM traffic.

Read article
ARTICLE · 06

AMD Instinct GPUs

How AMD Instinct GPUs (MI300X) compete with NVIDIA for AI workloads.

Read article
ARTICLE · 07

GPU Architecture Overview

How GPUs achieve massive parallelism through streaming multiprocessors, warps, and thousands of cores. The architecture that powers modern AI.

Read article
ARTICLE · 08

GPU asynchronous copy and software pipelining

Deep-dive on the asynchronous global-to-shared copy pipeline that fast GPU kernels are built on: cp.async streams tiles directly into a multi-stage sh…

Read article
ARTICLE · 09

GPU Batching Strategies Overview

Overview of batching approaches for LLM serving.

Read article
ARTICLE · 10

Chunked Prefill

Split long prefills into chunks.

Read article
ARTICLE · 11

Collective Communication Overlap

Overlapping computation with collectives.

Read article
ARTICLE · 12

GPU-Enabled Containers

How to run GPU workloads in Docker containers via nvidia-container-toolkit.

Read article
ARTICLE · 13

Continuous batching

Deep-dive on continuous (in-flight) batching for LLM inference: the scheduling technique that forms a fresh batch every decode iteration, letting fini…

Read article
ARTICLE · 14

GPU Datacenter Cooling Overview

Cooling GPU datacenters: air vs liquid.

Read article
ARTICLE · 15

GPU Cost Optimization

How to reduce GPU costs: rightsizing, spot, reserved, MIG, off-peak scheduling.

Read article
ARTICLE · 16

GPU CUDA Programming

The CUDA programming model: how kernels launch threads organized into blocks and grids, and the __global__/__device__ function distinction.

Read article
ARTICLE · 17

CUDA streams and graphs -- overlap and launch-overhead elimination

Deep-dive on CUDA streams and graphs: streams as ordered work queues enabling concurrency and copy-compute overlap, events for cross-stream synchroniz…

Read article
ARTICLE · 18

GPU Datacenter Deployment

The physical realities of GPU datacenters: 700W+ per GPU, liquid cooling, high-density networking, and the constraints that shape modern AI infrastruc…

Read article
ARTICLE · 19

GPU Datacenter Power Requirements

Powering GPU datacenters: MW scale.

Read article
ARTICLE · 20

Expert Parallelism

How expert parallelism distributes MoE experts across GPUs for efficient MoE training.

Read article
ARTICLE · 21

GPU Fine-Tuning Costs

How to estimate GPU fine-tuning costs based on model, data, method.

Read article
ARTICLE · 22

FlashAttention architecture

Deep-dive on FlashAttention: how tiling queries, keys, and values into on-chip SRAM blocks and using an online (streaming) softmax computes exact atte…

Read article
ARTICLE · 23

GPU Sharing Strategies Overview

Overview of GPU sharing approaches.

Read article
ARTICLE · 24

GPUDirect

How NVIDIA GPUDirect enables direct GPU access to storage and network without CPU involvement.

Read article
ARTICLE · 25

GPUDirect Storage architecture

Deep-dive on GPUDirect Storage: the cuFile API and kernel driver, peer-to-peer PCIe DMA from NVMe and NVMe-oF into GPU HBM, the CPU bounce buffer it e…

Read article
ARTICLE · 26

NVIDIA H100

What makes NVIDIA H100 the workhorse of modern AI training and inference.

Read article
ARTICLE · 27

H100 SM architecture

Deep-dive on H100 SM: warp scheduler, CUDA cores, 4th gen tensor cores, shared memory, TMA async copy, cluster launch, FP8/INT8.

Read article
ARTICLE · 28

GPU HBM architecture

Deep-dive on GPU HBM tier: bandwidth, banks, coalescing, L2/L1 caches, roofline model, async copy (TMA), memory-bound tuning.

Read article
ARTICLE · 29

LLM Inference Optimization Overview

Techniques for LLM inference optimization.

Read article
ARTICLE · 30

InfiniBand + NVLink architecture

Deep-dive on GPU cluster fabric: NVLink intra-node, InfiniBand inter-node, GPUDirect RDMA, NCCL + SHARP collectives, topology-aware placement.

Read article
ARTICLE · 31

GPU kernel fusion -- fewer kernels, less memory traffic

Deep-dive on GPU kernel fusion: the memory-bound HBM-traffic problem, fusing operations to keep intermediates on-chip (registers/SRAM), elementwise/ep…

Read article
ARTICLE · 32

LLM Canary Deployment

Canary: send small % to new LLM.

Read article
ARTICLE · 33

LLM Deployment Pattern Overview

Deployment patterns for LLM serving.

Read article
ARTICLE · 34

LLM FinOps

FinOps discipline for LLM apps.

Read article
ARTICLE · 35

LLM Guardrails

Guardrails for LLM production.

Read article
ARTICLE · 36

LLM Load Testing

Load-testing LLM serving stacks.

Read article
ARTICLE · 37

GPU Serving Architecture for LLMs: The Full Stack

A 2500-word walkthrough of a modern GPU serving stack for LLMs: ingress, mesh, router, engine, autoscaler, pod, fabric, object store, and observabilit…

Read article
ARTICLE · 38

LLM Serving Stacks

The main LLM serving stacks and their trade-offs.

Read article
ARTICLE · 39

Lookahead Decoding

How Lookahead Decoding speeds LLM generation without separate draft model.

Read article
ARTICLE · 40

GPU Memory Hierarchy

How GPU memory is organized: registers, shared memory, L1/L2 cache, HBM. Bandwidth matters more than latency.

Read article
ARTICLE · 41

NVIDIA MIG architecture

Deep-dive on Multi-Instance GPU: compute and memory slices, GPU Instance and Compute Instance hierarchy, profile geometry on A100/H100, Kubernetes GPU…

Read article
ARTICLE · 42

MoE All-to-All Communication

All-to-all: token shuffle for MoE.

Read article
ARTICLE · 43

MoE Serving Overview

Serving Mixture-of-Experts models.

Read article
ARTICLE · 44

NCCL collectives

Deep-dive on NCCL collective communication: all-reduce and collective ops, bandwidth-optimal ring and latency-optimal tree algorithms, NVLink/NVSwitch…

Read article
ARTICLE · 45

nvidia-smi

How nvidia-smi provides quick GPU status, and the key views for debugging.

Read article
ARTICLE · 46

NVLink and NVSwitch architecture

Deep-dive on NVLink/NVSwitch: link anatomy and memory semantics, the non-blocking crossbar, SHARP in-switch reductions, NCCL topology mapping, link-fl…

Read article
ARTICLE · 47

GPU occupancy architecture

Deep-dive on GPU occupancy: the ratio of active warps per SM to the hardware maximum, determined by whichever per-SM resource — registers per thread, …

Read article
ARTICLE · 48

P2P KV Cache Transfer

Peer-to-peer KV cache transfer between GPUs.

Read article
ARTICLE · 49

Paged KV cache architecture

Deep-dive on PagedAttention-style KV cache management: physical block pools and block tables, copy-on-write prefix caching, continuous batching under …

Read article
ARTICLE · 50

PCIe and GPU

How GPUs connect to CPU via PCIe, and why PCIe bandwidth matters for GPU workloads.

Read article
ARTICLE · 51

Prefill/Decode Disaggregation Architecture in Depth

A 2500-word walkthrough of prefill/decode disaggregation: router, prefill pool, KV transfer, decode pool, continuous batching, independent SLOs, autos…

Read article
ARTICLE · 52

Pipeline Parallelism

How pipeline parallelism splits model layers across GPUs, with micro-batches for utilization.

Read article
ARTICLE · 53

GPU Preemption Handling

Handling GPU preemption in training + serving.

Read article
ARTICLE · 54

GPU Profiling

How to profile GPU workloads: Nsight Systems for timeline, Nsight Compute for kernel-level detail.

Read article
ARTICLE · 55

Ring Attention

How ring attention enables million-token context by distributing sequence across GPUs.

Read article
ARTICLE · 56

RLHF Pipeline on GPU

Full RLHF pipeline: SFT → RM → PPO.

Read article
ARTICLE · 57

GPU Scheduling

How multiple workloads share a GPU: time-slicing, MPS (Multi-Process Service), and MIG (Multi-Instance GPU).

Read article
ARTICLE · 58

GPU Scheduling With Slurm

How Slurm schedules GPU jobs on HPC clusters via GRES (Generic RESources).

Read article
ARTICLE · 59

Multi-Region LLM Serving

Deploy LLM across regions.

Read article
ARTICLE · 60

GPU shared memory architecture

Deep-dive on GPU shared memory: why the memory hierarchy makes on-SM shared memory the linchpin of performance, the 32-bank structure and how warp-wid…

Read article
ARTICLE · 61

SLO-Aware Scheduling

Schedule LLM requests to meet SLOs.

Read article
ARTICLE · 62

Speculative Decoding Architecture in Depth

A 2500-word walkthrough of speculative decoding: draft/target models, generate + verify, accept + reject sampling, correctness proof, acceptance rate,…

Read article
ARTICLE · 63

Static Batching

Static batching: fixed batch size per iteration.

Read article
ARTICLE · 64

StreamingLLM

How StreamingLLM enables efficient inference on very long conversations.

Read article
ARTICLE · 65

Switch Transformer Architecture

Google Switch Transformer: foundational MoE.

Read article
ARTICLE · 66

Tensor core architecture

Deep-dive on GPU tensor cores: WMMA/MMA, tile size, accumulator dtype, input dtype, sparsity, multi-buffering, kernel selection.

Read article
ARTICLE · 67

GPU Tensor Cores

How tensor cores accelerate matrix multiplication for deep learning, precision modes (FP16, BF16, FP8, INT8), and when they kick in.

Read article
ARTICLE · 68

GPU Time Slicing

Software GPU time slicing.

Read article
ARTICLE · 69

GPU Tensor Memory Accelerator (TMA): Architecture Deep-Dive

How Hopper's Tensor Memory Accelerator moves tensor tiles asynchronously via descriptors and mbarriers, freeing compute warps…

Read article
ARTICLE · 70

Google TPU

How Google TPU compares to GPU for training and inference on Google Cloud.

Read article
ARTICLE · 71

NVIDIA vGPU

vGPU: virtualization + graphics + compute.

Read article
ARTICLE · 72

GPU warp specialization architecture

Deep-dive on GPU warp specialization: splitting a thread block's warps into producer warps that drive cp.async/TMA copies and…

Read article
ARTICLE · 73

ZeRO sharding architecture

Deep-dive on ZeRO: the 16-bytes-per-parameter memory math, stages 1/2/3, reduce-scatter and all-gather volume, prefetch overlap, hybrid sharding, ZeRO…

Read article
ARTICLE · 74

Latitude.sh

Bare metal cloud. Reserved + on-demand GPUs.

Read article
ARTICLE · 75

Optimizer State Offloading to CPU/NVMe

Move optimizer state off GPU. Trade PCIe bandwidth.

Read article
ARTICLE · 76

GPU Register Pressure

Registers per thread. Limits threads per SM.

Read article
ARTICLE · 77

SambaNova RDU

Reconfigurable Dataflow Unit. Alternative arch.

Read article
ARTICLE · 78

Second-Hand GPU Market

Enterprise churn feeds resale.

Read article
ARTICLE · 79

SHARP

Aggregate on switch. Faster all-reduce.

Read article
ARTICLE · 80

Warp Scheduling on GPU

32 threads = 1 warp. SIMT execution model.

Read article