#
memory-bandwidth
Here are 34 public repositories matching this topic...
Provides a set of benchmarks that can be used to measure the memory bandwidth performance of CPU's
-
Updated
Apr 8, 2024 - C
A Fast DNN Accelerator Design Space Exploration Framework.
deep-neural-networks accelerator design-space-exploration memory-hierarchy loop-optimization cost-estimation memory-bandwidth loop-unrolling architecture-search energy-analysis latency-analysis auto-scheduling dataflow-optimization mac-array-utilization
-
Updated
Aug 10, 2022 - Python
Modern Memory Bandwidth and Latency Benchmarks
-
Updated
Jun 10, 2026 - C
Demo code accompanying the talk "Implementing memory locality optimizations in OpenFOAM based code"
-
Updated
Apr 7, 2025 - C++
Main Memory Bandwidth Monitoring
-
Updated
Sep 17, 2018 - C++
General-purpose compile-time Expression Templates library for C++
-
Updated
Sep 29, 2025 - C++
Measure and visualize why LLM inference is slow: bottleneck analysis, model dissection, KV-cache, GEMM/GEMV, quantization, and memory-bound decoding.
visualization webgl benchmark machine-learning inference systems decode performance-analysis profiling quantization gemm webgpu memory-bandwidth prefill kv-cache gemv llm llm-inference memory-bound bottleneck-analysis
-
Updated
Apr 30, 2026 - HTML
Reproducible Pascal GPU Unified Memory benchmark with Nsight and nvprof profiling
benchmark-suite memory-bandwidth page-faults nvprof unified-memory nsight-systems cuda-pascal-unified-memory-gpu nsight-nvprof-pcie-bandwidth cudamallocmanaged cudamemprefetchasync cudapascal
-
Updated
Feb 1, 2026 - Python
-
Updated
Feb 11, 2025 - Python
C++23 benchmarking framework with 6 profiler backends, CUDA GPU support, statistical regression detection, cross-compilation for 5 architectures, and CLI tools for analysis and visualization.
python linux rust raspberry-pi benchmarking cuda ci-cd perf riscv flamegraph profiling google-test performance-testing jetson callgrind gperftools memory-bandwidth nsight bpftrace cpp23
-
Updated
Jul 3, 2026 - C++
Console UI for watching Memory stats
linux rust terminal tui perf performance-monitoring pmu memory-bandwidth memory-monitor dmidecode hardware-monitor ram-monitor ratatui
-
Updated
May 7, 2026 - Rust
Intent-aware KV execution prototype for agentic long-context inference: semantic block selection, dynamic scoring, KV quantization modeling, speculative prefetch simulation, CPU references, and future Triton/CUDA kernels.
cost-model triton memory-bandwidth gpu-kernels mixed-precision prefetching inference-optimization kv-cache sparse-attention long-context paged-attention semantic-routing agentic-ai block-sparse-attention kernel-research block-attention semantic-attention kv-quantization paged-kv kv-cache-optimization
-
Updated
May 29, 2026 - Python
Python lab for exploring memory bandwidth, cache effects, and locality in accelerator workloads
performance-engineering hpc cuda memory-bandwidth gpu-performance roofline-model cache-locality tiling-optimization
-
Updated
Dec 24, 2025 - Python
Measures actual GPU costs for LLM prefill and decode on RTX 2070 (8GB) to validate simulation parameters. Key findings: prefill converges to 30-70 us/token at long sequences; decode is memory-bandwidth-bound (constant with prefix length, 5300-10800 us/token single-request); simulation defaults are correct for server-level amortized batching.
python benchmark gpu inference pytorch decode profiling memory-bandwidth prefill kv-cache llm rtx-2070
-
Updated
Jul 10, 2026 - Python
OpenCL benchmarking tool to measure host-device bandwidth and kernel global memory throughput across GPUs and CPUs.
-
Updated
Mar 16, 2026 - C
Bandwidth-focused LLaMA token-generation latency benchmarking on Apple Silicon M4 vs M2 with TTFT, PTL, KV-cache, quantization projections, and reproducible architecture analysis.
benchmarking pytorch llama mps computer-architecture quantization memory-bandwidth kv-cache apple-silicon llm-inference
-
Updated
May 10, 2026 - Python
Production-grade Triton kernel fusing residual add + RMSNorm + packed QKV projection into a single GPU launch for decoder-only transformer inference (Llama-3, Mistral, Qwen2). +2.4% tok/s, -1.5 GB VRAM on A10G.
cuda pytorch transformer triton llama memory-bandwidth gpu-kernels kernel-fusion rmsnorm llm-inference
-
Updated
Apr 22, 2026 - Python
Improve this page
Add a description, image, and links to the memory-bandwidth topic page so that developers can more easily learn about it.
Add this topic to your repo
To associate your repository with the memory-bandwidth topic, visit your repo's landing page and select "manage topics."