Hacker News new | ask | show | jobs
user: matt_d
created: 2014-04-21
karma: 20568

submissions:

PyTorch: A Reference Language
78 points | 9 comments
Tracked Capabilities for Safer Agents
4 points | 0 comments
Attention Decode on AMD MI450 GPUs: A Gluon Kernel Optimization Guide
3 points | 0 comments
Don't Use Ordinary Software to Contain Software-Hacking Agents
1 points | 0 comments
A practical partitioner for distributed simulations on sparse dynamic domains
2 points | 0 comments
NVLink, NVSwitch, and All That
2 points | 0 comments
Continuation-Centric Computing with Arca
2 points | 0 comments
AutoCO: An Online Continuous Optimization System for Phase-Changing DB Workloads
3 points | 0 comments
Reusing Buffers in Multicore OCaml
4 points | 0 comments
Mohabi: Disaggregating and Sandboxing the Firefox JavaScript Engine
11 points | 0 comments
Automated Discovery Has No Universally Superior Harness
1 points | 0 comments
Drawn Apart: Remote GPU Fingerprinting
2 points | 0 comments
ICFP Programming Contest 2026
5 points | 0 comments
Optimization of SASS Stall Counts
2 points | 0 comments
Characterizing Metastable Faults and Failures
1 points | 0 comments
Every Microsecond Matters:Achieving Near SpeedOfLight Latency in GPU Collectives
2 points | 0 comments
SPIR-V on ROCm: A Portable IR for AMD GPUs
5 points | 2 comments
GEMM Performance Measurement Methodology Guidelines
1 points | 0 comments
Faster Algorithms for Structured Matrix Multiplication via Flip Graph Search
1 points | 0 comments
AI for Systems is "AGI-Complete"
3 points | 0 comments
Controlling Reasoning Effort in LLMs
1 points | 0 comments
Making TLA+ and x86 Kiss via Z3Py
4 points | 0 comments
Irreducible Loops
3 points | 0 comments
Fleet: Hierarchical Task-Based Abstraction for Megakernels on Multi-Die GPUs
10 points | 1 comments
Accelerating Block Low-Rank Foundation Model Inference on MemoryConstrained GPUs
10 points | 0 comments
Locality-Aware Automatic Differentiation on the GPU for Mesh-Based Computations
2 points | 0 comments
Triton Plugin Extensions: Enabling TLX and Custom Compiler Passes Out of the Box
3 points | 0 comments
What If the Harness Comes Before Pretraining? A Data Flywheel Perspective
2 points | 0 comments
0 points | 0 comments