mirror of
https://github.com/allaunthefox/Research-Stack.git
synced 2026-08-12 15:20:36 +00:00
8356558ea7
AMD Ryzen 5 9600X (6C/12T, AVX-512), 31GiB DDR5, 1.8TB NVMe, Radeon iGPU, CachyOS 7.0.3. Mapped 8 capabilities to stack use.
101 lines
2.8 KiB
Text
101 lines
2.8 KiB
Text
title: Q-Fox (Primary Development Machine)
|
||
tags: infrastructure hardware workstation
|
||
created: 20260507
|
||
modified: 20260507
|
||
|
||
# Q-Fox — Primary Development Machine
|
||
|
||
## Overview
|
||
|
||
| Field | Value |
|
||
|-------|-------|
|
||
| **Hostname** | Q-Fox |
|
||
| **Role** | Primary development, compilation, simulation, and orchestration host |
|
||
| **OS** | CachyOS Linux (Arch-based) |
|
||
| **Kernel** | 7.0.3-1-cachyos |
|
||
| **Architecture** | x86_64 |
|
||
|
||
---
|
||
|
||
## CPU
|
||
|
||
| Field | Value |
|
||
|-------|-------|
|
||
| **Model** | AMD Ryzen 5 9600X (Granite Ridge) |
|
||
| **Cores** | 6 physical / 12 logical (SMT) |
|
||
| **Base Clock** | 628 MHz (min) |
|
||
| **Boost Clock** | 5,486 MHz (max) |
|
||
| **L1d Cache** | 288 KiB (6 × 48 KiB) |
|
||
| **L1i Cache** | 192 KiB (6 × 32 KiB) |
|
||
| **L2 Cache** | 6 MiB (6 × 1 MiB) |
|
||
| **L3 Cache** | 32 MiB (unified) |
|
||
| **ISA Extensions** | AVX-512 (F, DQ, IFMA, CD, BW, VL, BF16, VBMI, VBMI2, VNNI, BITALG, VPOPCNTDQ, VP2INTERSECT), AVX2, AVX-VNNI, SHA-NI, AES |
|
||
| **Virtualization** | AMD-V |
|
||
| **TDP** | 65W (nominal) |
|
||
|
||
---
|
||
|
||
## Memory
|
||
|
||
| Field | Value |
|
||
|-------|-------|
|
||
| **Total** | 31.1 GiB (DDR5) |
|
||
| **Type** | DDR5 (AM5 platform) |
|
||
| **Swap** | 30.4 GiB (zram) |
|
||
|
||
---
|
||
|
||
## Storage
|
||
|
||
| Device | Size | Type | Model |
|
||
|--------|------|------|-------|
|
||
| `/dev/nvme0n1` | 1.8 TB | NVMe PCIe 4.0 | MSI M480 PRO 2TB |
|
||
|
||
| Partition | Size | Usage |
|
||
|-----------|------|-------|
|
||
| `/dev/nvme0n1p2` (/) | 1.9 TB | 98% used (45G free) |
|
||
|
||
---
|
||
|
||
## GPU
|
||
|
||
| Field | Value |
|
||
|-------|-------|
|
||
| **Model** | AMD Radeon Graphics (Granite Ridge iGPU) |
|
||
| **Architecture** | RDNA 2 (integrated) |
|
||
| **VRAM** | Shared system memory |
|
||
| **PCI ID** | `12:00.0 VGA compatible controller: AMD/ATI Granite Ridge [Radeon Graphics] (rev c6)` |
|
||
|
||
---
|
||
|
||
## Networking
|
||
|
||
| Interface | Type | Address |
|
||
|-----------|------|---------|
|
||
| `enp14s0` | Ethernet (Realtek) | IPv6 only |
|
||
| `wlan0` | WiFi | 192.168.1.118/24 |
|
||
| `tailscale0` | Tailscale VPN | Mesh overlay |
|
||
|
||
---
|
||
|
||
## Research Stack Relevance
|
||
|
||
| Capability | Stack Use |
|
||
|-----------|-----------|
|
||
| AVX-512 (full suite) | Q16.16 fixed-point vectorization, PIST encode/decode, FAMM delay computation |
|
||
| 12 threads | 11-agent swarm + 1 orchestrator thread |
|
||
| 32 MiB L3 cache | Fits full topological device plan in cache-resident state |
|
||
| 31 GiB RAM | Virtual FPGA cluster (5× Tang Nano 9K) fits in memory |
|
||
| NVMe 4.0 | Fast Lean 4 lake builds, Verilator simulation traces |
|
||
| AVX-512 BF16 | bf4prover bfloat16 proof search acceleration |
|
||
| SHA-NI / AES | Receipt hashing, invariant witness compression |
|
||
| AMD-V | Nested virtualization for sandboxed prover execution |
|
||
|
||
---
|
||
|
||
## Notes
|
||
|
||
- 98% disk usage — cleanup recommended before large builds
|
||
- No discrete GPU — all compute on CPU (sufficient for current stack)
|
||
- CachyOS provides optimized kernel with BORE scheduler and LTO builds
|
||
- zram swap at 30.4 GiB matches physical RAM (1:1 ratio)
|