#CUDA kernel optimization
113 related articles
·6 min
Not Everything Is Worth a Token: When to Use AI and When to Use Deterministic Code
Over-reliance on LLMs is an overlooked pitfall in AI development. Explore the hidden costs of Token economics, the boundaries between LLMs and deterministic code, and how hybrid architectures balance flexibility and reliability.
Read more →
·5 min
NVIDIA AI-Q Blueprint × OCI: A Practical Guide to Production Deployment of Enterprise AI Agents
Deep dive into NVIDIA AI-Q Blueprint production deployment on Oracle Cloud Infrastructure, covering NIM microservices, RAG architecture, multi-agent orchestration, and OCI GPU selection for enterprise AI agents.
Read more →
·5 min
NVIDIA AI-Q Blueprint × OCI: A Practical Guide to Production Deployment of Enterprise AI Agents
Deep dive into NVIDIA AI-Q Blueprint production deployment on Oracle Cloud Infrastructure, covering NIM microservices, RAG architecture, multi-agent orchestration, and OCI GPU selection.
Read more →
NVIDIA Nsight Tools in Practice: A Per…
·3 min
NVIDIA Nsight Tools in Practice: A Performance Optimization Guide for Neural Reconstruction Pipelines
Learn how to optimize neural reconstruction pipelines using NVIDIA Nsight Systems and Nsight Compute — from CPU-GPU sync blocking to kernel-level profiling for autonomous driving simulation.
Read more →
·3 min
Unsloth Major Update: DiffusionGemma & Gemma 4 MTP Support, 2x Inference Speed Boost
Unsloth v0.1.464-beta adds DiffusionGemma, Gemma 4 MTP, and MiniMax-M3 support, delivering ~2x inference speed boost, new Hub, RAG Q&A, tensor parallelism, and full CUDA/ROCm/Windows coverage.
Read more →
·1 min
Unsloth v0.1.47-beta Released: 2x Faster LLM Fine-Tuning with 70% Less VRAM
Unsloth v0.1.47-beta is out. This 67.9k-star open-source framework fine-tunes Llama, Mistral, and Qwen 2x faster with 70% less VRAM on consumer GPUs.
Read more →
·7 min
PP-OCRv6 Deployment Deep Dive: Invoice Recognition Results and the Font-Missing Pitfall Explained
PP-OCRv6 is PaddlePaddle's SOTA-level OCR model. This article covers Docker deployment, invoice recognition testing, and the critical font-missing pitfall causing entire regions to be skipped.
Read more →
·6 min
14x Speedup: How Manticore Rebuilt the ONNX Embedding Inference Path
Manticore Search restructured its ONNX inference path to achieve 14x faster text embeddings. Deep dive into batching, session reuse, zero-copy memory, and thread tuning for vector search systems.
Read more →
·6 min
Codex Beginner's Guide: Four Product Forms Explained, Plus Comparison with Claude Code and Cursor
In-depth analysis of OpenAI Codex's four usage forms, comparing Codex, Claude Code, and Cursor across price, stability, and frontend/backend fit to help developers choose the right AI programming tool.
Read more →
·3 min
Running Local LLMs on AMD GPUs: Ditching Ollama for LM Studio Delivers 7x Speed Boost
AMD GPU black screens running local LLMs? This post-mortem covers Ollama's 3 fatal flaws and how switching to LM Studio boosted token speed from 5 to 36, with ROCm setup, Speculative Decoding, and GFX version tips.
Read more →
·3 min
Making LLMs Faster and Lighter: A Practical Approach to Reshaping Sparsity for GPUs
Deep dive into Sakana AI and NVIDIA's latest research using TwELL sparse packing format and custom CUDA kernels to convert LLM sparsity into real GPU speedups, achieving 20%+ faster inference/training and significantly lower memory usage.
Read more →
·3 min
DAQIRI Platform Explained: Deep Integration of High-Speed Data Acquisition and Real-Time AI Inference
Deep dive into how the DAQIRI platform embeds NVIDIA GPU-accelerated computing into high-speed data acquisition pipelines, enabling real-time AI inference for industrial inspection, scientific experiments, and autonomous driving.
Read more →
·3 min
Huawei HDC Deep Dive: Pangu 2.0 Goes Open Source and HarmonyOS 7 Agents Reshape the Mobile AI Ecosystem
Huawei HDC unveils Pangu 2.0 full open source and HarmonyOS 7 system-level Agent capabilities. Deep analysis of sparse architecture efficiency, on-device 30B models, and the Agent gateway battle.
Read more →
·3 min
LFM2.5 Local Deployment Hands-On: An 8B Parameter Model That Outperforms GPT-o3s in Tool Calling
Hands-on test of Liquid AI's LFM2.5 local deployment: architecture breakdown, 16GB VRAM troubleshooting, and GraphRAG tool-calling benchmarks vs GPT-o3s.
Read more →
·4 min
Getting Started with Huawei Ascend C Operator Programming: Kernel Functions and Hands-On Development Guide
A systematic guide to Huawei Ascend C operator programming covering kernel functions, three-stage pipeline paradigm, API categories, and a hands-on AddCustom operator walkthrough.
Read more →
·4 min
Cursor Composer2 Training Revealed: A Complete Guide to Distributed Reinforcement Learning Engineering
Deep dive into how Cursor trained Composer2: two-stage architecture, global distributed clusters, MOE numerical alignment, simulation anti-cheating, and more.
Read more →
·3 min
vLLM Deep Dive: How PagedAttention Enables High-Throughput LLM Inference
Deep dive into vLLM's core technologies for high-throughput LLM inference, including PagedAttention memory management, continuous batching, distributed deployment, and comparisons with TensorRT-LLM.
Read more →
Expert Opinions·3 min
Windsurf CEO Deep Dive Interview: Speed Is the Only Moat
Windsurf CEO Varun Mohan shares insights on AI coding IDE pivots, product methodology, async Agent challenges, and differentiation strategy vs Cursor. Speed is the only moat.
Read more →
Tutorials·3 min
Why Learning AI from Scratch Leaves You More Confused — A Clear, Systematic Roadmap for Beginners
Confused learning AI from scratch? This guide breaks down why fragmented learning fails and provides a complete path from Python to deep learning with practical tips.
Read more →
Tutorials·3 min
llama.cpp MTP Acceleration Deployment Guide: Configuration Steps & Real-World Benchmarks
Guide to enabling MTP multi-Token prediction acceleration in llama.cpp, covering CUDA setup, desktop configuration, model selection, and benchmarks showing ~60 Token/s with Qwen3 27B.
Read more →