PyTorch · YouTube
Printed in Watch · youtube.com
Publishes about 3 a week · last piece Oct 15, 2026
How often
about 3 a week, by the archive: 11 pieces since Sep 24, 2026.
12 weeks agothis week
Latest pieces
- Open Source Inference at Production Scale
- MoRI + vLLM: Wide Expert Parallelism and RDMA KV-Cache Transfer for Disaggregated MoE Serving on AMD
- High Performance Serving Framework for Offline and Real-time Diffusion Models - SGLang-Diffusion
- DeepSpeed Is Not Just ZeRO: Tensor, Sequence, and Expert Parallelism with ZeRO-Level Usability
- PyTorch-Native LLM Serving on TPU: SGLang and vLLM
- What You Cannot Profile, You Cannot Optimize: Learning to Read PyTorch Traces
- Self-Healing Rollouts: Automating Production Fixes With Agentic AI - Kevin Dubois & Daniel Oh
- Optimizing Mixture of Experts LLM inference on ARM CPUs
- Evolving Argo Workflows for AI: New Features, Tuning, and Multi-Tenant Isolation - S. Tian & Y. Su
- Operating Frontier Intelligence at Scale - Chris Aniszczyk & Xiao Zhang
- Tidal Auto-scaling for Training and Inference Based on Kubernetes + KEDA - Jun Zheng & Tan Pei Xiang
Cited this week by
Other sources whose pieces link to it.
- Google Ads Developers
- AI Now Institute
- AlmaLinux
- APNIC Blog
- Arize
- Arm Newsroom
- AWS Compute
- AWS Machine Learning
- Bazel
- Chrome for Developers
- ClickHouse
- David Heinemeier Hansson
- Docker
- Elastic
- Flutter
- Martin Fowler
- Framework
- GitButler
- Godot Engine
- Google Research
- Hamel Husain
- Inside Java
- Nathan Lambert
- JetBrains
- LanceDB
- LangChain
- Laravel
- swyx (Latent Space)
- LLVM
- Modular
- Mozilla
- Microsoft Foundry
- Microsoft 365 Developer
- Mux
- Neo4j
- Netflix TechBlog
- NVIDIA
- NVIDIA Developer
- NVIDIA Robotics
- Nx
- OpenSSF
- openSUSE
- OpenTelemetry
- PlanetScale
- Gergely Orosz (The Pragmatic Engineer)
- Ruby on Rails
- Sebastian Raschka
- Redpanda
- Snorkel AI
- Spring
- Svelte
- Tailscale
- Thorsten Ball
- Tor Project
- VoidZero
- WarpStream
- watchTowr Labs
- Zvi Mowshowitz
In these topics
Pick one in My Edition to read it with its neighbors.