A repo for llm on ncnn

Based on Nano-vLLM, a simple replication of vLLM with self-contained paged attention and flash attention implementation

test pl ddr；run in c++; self defined memory controller for pl ddr

Cookbook of SGLang - Recipe

tinyGPU: A Predicated-SIMD processor implementation in SystemVerilog

Large Language Model Inference Acceleration: A Comprehensive Hardware Perspective

Release of stream-specialization software/hardware stack.

一个基于nano banana pro🍌的原生AI PPT生成应用，迈向真正的＂Vibe PPT＂; 支持上传任意模板图片；上传任意素材&智能解析；一句话/大纲/页面描述自动生成PPT；口头修改指定区域、一键导出 - An AI-native PPT generator based on nano banana pro🍌

Premium Email Blog

cuTile is a programming model for writing parallel kernels for NVIDIA GPUs

Tile-Based Runtime for Ultra-Low-Latency LLM Inference

"Paper2Slides: From Paper to Presentation in One Click"

A Plugin-Based Multi-Agent System for In-Editor Academic Writing, Review, and Editing

[MLSys'25] QServe: W4A8KV4 Quantization and System Co-design for Efficient LLM Serving; [MLSys'25] LServe: Efficient Long-sequence LLM Serving with Unified Sparse Attention

CXL-DMSim: A Full-System CXL Disaggregated Memory Simulator With Comprehensive Silicon Validation

Chisel examples and code snippets

Preview Code for Continuum Paper

Code for the paper “Four Over Six: More Accurate NVFP4 Quantization with Adaptive Block Scaling”

A framework for efficient model inference with omni-modality models

RISC-V SystemC-TLM simulator

DRAMsim3: a Cycle-accurate, Thermal-Capable DRAM Simulator

A simulator for SK hynix AiM PIM architecture based on Ramulator 2.0

Sample codes for my CUDA programming book

The copy of materials from UPMEM website.

NeuPIMs: NPU-PIM Heterogeneous Acceleration for Batched LLM Inferencing

Processing-In-Memory (PIM) Simulator

The official repository for the gem5 computer-system architecture simulator.