AI and infrastructure news
- C and C++Article
Turning compiler optimizations into practical tests
Published on March 3, 2026, Matt Godbolt’s blog post gathers an archive of Advent of Compiler Optimizations 2025, useful to C and C++ engineers exploring compiler optimizations.
- C and C++Post
How a debugger works internally: an eight-part series
A blog series published on June 3, 2026, traces the creation of a debugger, from connecting to a process to working with source code and symbols.
- Attention and model architecturePaper
How to validate questions about knowledge graphs
A post published on December 10, 2025 describes SEAL, a framework that extracts the minimum core of a question and uses an agent to validate it against a database.
- C and C++Article
How to visualize a Clang AST in Compiler Explorer
A post published on April 1, 2026, shows how Compiler Explorer can display the Clang AST for a short C++ program.
- Attention and model architecturePaper
KV cache compression could cut memory use by up to 50×
A publication dated June 12, 2026 describes an Attention Matching method for KV cache compression, reporting reductions of up to 50× on some datasets with minimal quality loss.
- C and C++Article
C/C++ compilation: make builds and loading traceable
Published on November 13, 2025, the series covers C and C++ compilers, linking, object files, shared libraries, and loaders. These topics can inform improvements to build and deployment traceability and diagnostics.
- C and C++Article
Compiler construction for ARM: practical foundations
Vladimir Keleshev’s book teaches compiler fundamentals and Assembly through a TypeScript subset that generates code for 32-bit ARM.
- C and C++Article
Undefined behavior in C and C++: reducing risks
Published on June 9, 2026, the article presents a guide to undefined behavior in C and C++ and highlights its relevance to writing and analyzing code in these languages.
- Training and fine-tuningPost
Gradient compression cuts network demand in pretraining
A post published on December 23, 2025, reports using gradient compression in pretraining while maintaining high efficiency with one quarter of the stated network bandwidth.
- CUDA and GPU programmingArticle
GPU computing: from architecture to optimization
An introductory article published on March 28, 2026, presents GPU architecture and computing as context for engineers developing or optimizing software for this hardware.
- C and C++Article
Concurrency in C++20/23: reducing common errors
A guide published on March 29, 2026 covers common multithreading errors in modern C++ and presents practices intended for production-ready code.
- C and C++Paper
Float conversion: comparing algorithms
A March 10, 2026 paper compares floating-point conversion algorithms across architectures and datasets; Schubfach and Dragonbox reach up to ten times the speed of Dragon4.
- C and C++Project
CoreRipperX 2.0 expands CPU stress testing and monitoring
A February 1, 2026 post describes new AVX tests, load modes, and throttling and effective-clock monitoring in CoreRipperX 2.0.
- CUDA and GPU programmingArticle
CuBridge explores LLMs for reconstructing CUDA attention kernels
Published on May 10, 2026, the text presents CuBridge, a framework that uses LLMs to understand and reconstruct efficient attention kernels in CUDA.
- CUDA and GPU programmingDocs
CUDA C++: from a parallel example to a GPU test
In a post published on February 28, 2026, NVIDIA introduces CUDA C++ programming step by step through a simple parallel programming example.
- CUDA and GPU programmingDocs
CUDA C++: first steps in parallel programming
In a post published on November 3, 2025, NVIDIA presents a step-by-step introduction to CUDA C++ through a simple parallel programming example.
- CUDA and GPU programmingPaper
CUDA-L2 automates HGEMM kernel tuning
An article published on December 15, 2025 presents an approach combining language models and reinforcement learning to optimize CUDA kernels for half-precision matrix multiplication.
- CUDA and GPU programmingArticle
CUDA for LLMs: kernels, attention and profiling
Published on February 3, 2026, Manning’s book introduces CUDA programming for NVIDIA GPUs, from basic kernels to Flash Attention, and covers profiling with Nsight Compute.
- SecurityProject
CVE-Genie and automated exploits for known vulnerabilities
The repository description, published on June 15, 2026, presents CVE-Genie as a tool that generates working exploits from known vulnerabilities through a few API calls.
- SecurityProject
cvs-fast-export 2.1 fixes errors with binary blobs
Published on June 10, 2026, cvs-fast-export 2.1 fixes an “invalid delta” error when processing binary blobs in RCS or CVS histories.
- Training and fine-tuningPost
DAPO reports GRPO improvement on AIME, from 30% to 50%
In a paper published on March 1, 2026, DAPO reports that standard GRPO scored 30% on AIME in its experiments, while adding some techniques raised the result to 50%.
- SecurityArticle
Dausos: Surfshark presents a new VPN protocol
On June 15, 2026, Surfshark presented Dausos, a VPN protocol it describes as designed for speed, security, and efficiency, using AEGIS-256X2.
- CUDA and GPU programmingPost
From Numba to Triton: a practical path to GPU programming
Published on December 12, 2025, the publication presents a Jupyter notebook that teaches GPU programming with Numba and progresses to writing kernels with Triton.
- CUDA and GPU programmingArticle
From SIMD to SIMT: what GPU evolution means in practice
In an article published on February 5, 2026, ACM SIGGRAPH Blog explains how GPUs combined thread abstractions with SIMD hardware, becoming flexible processors for graphics, AI, and scientific computing.
- Attention and model architecturePaper
DeepCrossAttention combines Transformer layers based on input
A description published on March 16, 2026 presents DeepCrossAttention, which uses learned, input-dependent weights to combine Transformer layer outputs.
- LLM inferencePost
DEER proposes diffusion drafts to accelerate LLM inference
A publication dated December 18, 2025 describes DEER, which drafts tokens with diffusion models and verifies them with autoregressive models. The text claims up to 5.54× faster inference without quality loss.
- C and C++Article
Defer in C: eight approaches and their trade-offs
An article published on February 5, 2026 compares eight ways to implement defer in C and discusses portability and resource-cleanup patterns.
- HardwareArticle
Fuxi GPU Demonstrated for Lenovo New Vision Z-Engine
On November 30, 2025, a Fuxi GPU with Imagination DXD architecture was demonstrated for Lenovo’s New Vision Z-Engine. The case highlights practical evaluation of hardware and platform compatibility.
- C and C++Post
Ten C projects to strengthen systems programming
A publication dated April 11, 2026 presents ten C projects, from a memory allocator to a mini debugger, in increasing order of difficulty.
- LLM inferencePost
DFlash and DDTree run Qwen3.6-27B on one RTX 3090
A post dated April 23, 2026 reports 73 tokens per second with speculative decoding on a single RTX 3090 and identifies architectural compatibility as a factor in running the model.
- LLM inferenceProject
dflash-mlx v0.1.5 expands runtime and serving features
On May 3, 2026, version 0.1.5 added a unified CLI, RAM and SSD caching, prefix matching, and experimental support for long contexts.
- Reinforcement learningPost
Progressive distillation for improving VLMs
In a post published on January 4, 2026, the author describes the Masters framework and reports gains on two vision-language models using an approach intended to avoid the computational cost of online RL.
- C and C++Article
djbsort adds performance gains with AVX2 and support for more numeric types
The new djbsort version brings performance gains with AVX2 and now sorts int32, uint32, float32, int64, uint64 and float64 arrays, with updated tests, documentation and benchmarks.
- C and C++Post
From C and C++ code to an executable: compilation stages
Published on December 24, 2025, the publication describes how preprocessing, compilation, assembly, and linking turn C and C++ code into an executable.
- Training and fine-tuningPost
DroPE extends context without long-context fine-tuning
In a post published on January 12, 2026, DroPE removes positional embeddings after training and uses brief calibration to extend context. The post reports gains over baselines in 16k retrieval and question answering for models up to 7B.
- AI agentsPaper
Dual-play proposes training LLM reasoning with less supervision
Published on December 15, 2025, the paper presents dual-play: two models take specialized roles, one creating questions and the other solving them, to reduce reliance on external supervision.
- Training and fine-tuningPaper
EB-JEPA explores world models through embedding predictions
Published on February 4, 2026, the article presents EB-JEPA, an open-source modular library for learning representations and world models with architectures that predict in representation space rather than pixel space.
- C and C++Article
eBPF on Linux: Observability Without Changing the Kernel
On March 28, 2026, the eBPF site introduced technology for running sandboxed programs in the Linux kernel without modifying its source code or loading a module, and offered tutorials and community resources.
- Speech: TTS and ASRProject
Prompt-based audio editing: what companies can evaluate
A StepFun post published on November 8, 2025 describes a 3-billion-parameter model for audio editing and multilingual speech generation.
- Reinforcement learningPaper
EditScore applies reward models to image editing
A post dated October 25, 2025 presents a family of reward models for evaluating and guiding complex image edits, with reported results on EditReward-Bench.
arxiv.org
hgpu.org
Let's assess your project
I reply with scope and a proposal.