Attention Optimization, Memory Efficiency, Transformer Acceleration, IO-Aware

Pixel Art Filtering
jorenjoestar.github.io·4h·
Discuss: Hacker News
✂️CUTLASS
Flag this post
Computers Are Getting Much Better at Image Recognition
smithsonianmag.com·1d
👁️Attention Optimization
Flag this post
Third Point Q3 2025 Investor Letter
seekingalpha.com·20h
⏱️Benchmarking
Flag this post
Parallel achieves 70% accuracy on SEAL, benchmark for hard web research
parallel.ai·1h·
Discuss: Hacker News
🌐Distributed Computing
Flag this post
Why stop at 1M tokens when you can have 10M?
news.ycombinator.com·8h·
Discuss: Hacker News
ONNX Runtime
Flag this post
The Learning Loop and LLMs
martinfowler.com·6h·
Discuss: Hacker News
🤖AI Coding Tools
Flag this post
Vibe Check: Claude Skills Need a ‘Share’ Button
kill-the-newsletter.com·1d
🤖AI Coding Tools
Flag this post
Scalable In-Memory Associative Processing for Graph Neural Network Inference
dev.to·2d·
Discuss: DEV
🎯GPU Kernels
Flag this post
Windsurf Codemaps: Understand Code, Before You Vibe It
cognition.ai·2h·
🤖AI Coding Tools
Flag this post
Balancing Cost, Power, and AI Performance
oreilly.com·2h
ONNX Runtime
Flag this post
DCcluster-Opt: Benchmarking Dynamic Multi-Objective Optimization for Geo-Distributed Data Center Workloads
arxiv.org·15h
🔗NCCL
Flag this post
Real-time Semantic Segmentation for AR Glasses: Dynamic Occlusion Handling via Bayesian Fusion
dev.to·11h·
Discuss: DEV
🏎️TensorRT
Flag this post
Honest take: I tested 12+ AI vibe coding tools, but this one actually surprised me
vibe.forem.com·8h·
Discuss: DEV
🤖AI Coding Tools
Flag this post
Predicting Encoding Energy from Low-Pass Anchors for Green Video Streaming
arxiv.org·15h
🔗Kernel Fusion
Flag this post
FG-CLIP 2: A Bilingual Fine-grained Vision-Language Alignment Model
paperium.net·1d·
Discuss: DEV
🧩Attention Kernels
Flag this post
ViSurf: Visual Supervised-and-Reinforcement Fine-Tuning for LargeVision-and-Language Models
paperium.net·2d·
Discuss: DEV
🏎️TensorRT
Flag this post
Unlocking AI Potential: Squeezing Giant Models into Tiny Spaces
dev.to·1d·
Discuss: DEV
📉Model Quantization
Flag this post