Low-latency NUMA-aware fork-join thread-pool with zero allocations, syscalls, CAS, or false-sharing on the hot path for C++, C, Rust, Zig & Mojo 🍴
-
Updated
Nov 18, 2026 - C++
Low-latency NUMA-aware fork-join thread-pool with zero allocations, syscalls, CAS, or false-sharing on the hot path for C++, C, Rust, Zig & Mojo 🍴
OSVVM Utility Library: AlertLogPkg, CoveragePkg, RandomPkg, ScoreboardGenericPkg, MemoryPkg, TbUtilPkg, TranscriptPkg, ...
Memory consistency modelling using Alloy
High speed C/C++ based behavioural VHDL/Verilog co-simulation memory model
[ECCVW2018] A Memory Model based on the Siamese Network for Long-term Tracking (MMLT)
A runtime where LLM inference is the primitive unit of computation — conversations are processes, prompts are in-memory data, a system image is one JSON | 以 LLM 推理为原始计算单元的运行时
Vyukov MPSC queue in C++20 with a written memory-model proof. Beats mutex, spinlock and boost::lockfree by 3.7–4.3× (WSL2, median of 5 runs). seq_cst cost measured: 8.8 → 20.6 ns per push. CI on x86-64 and AArch64. Push is wait-free; the queue is not lock-free, and says so.
Open AI agent skills implementing Principles of Earned Autonomy: intent interpretation, persistent context and inspectable improvement workflows. Experimental; model- and host-dependent.
A free interactive Python course that shows what your code does in memory. Step through names, frames, references, the GIL, exceptions, imports and special methods as animated diagrams. Plain HTML, CSS and JS.
A low level, modular bytecode, finite register VM focused on safety and extensibility, with plans for a Python like, statically typed language on top.
High-salience events slow the internal clock and reinforce memory. Quiet stretches accelerate the clock and let stale items decay. One model. One knob. Adaptive tempo.
Lean 4 infrastructure for proving program behavior from machine code, with explicit memory, OS resource, and concurrency models.
"Interactive demo: Memory Model V3 with MLP, timestamps and bypass."
Reproducible C# experiments for sparse, overlapping, and dynamic memory engrams
Learn Python by watching what happens in memory: an interactive stack/heap visualizer, sandbox, and 78-level challenge course (EN/DE/FA) — from print() to NumPy, Pandas, asyncio & threading.
How many CPU cycles does one Apple M4 core need to read a NASDAQ ITCH message, update a full order book and decide? 135, verified on 28.7M real messages. Rust, PMU counters, lock-free SPSC rings, pre-registered hypotheses.
To associate your repository with the memory-model topic, visit your repo's landing page and select "manage topics."