Framework Desktop (Ryzen AI Max+ 395 128GB) Unlocks 112GB ROCm Unified VRAM for Under $2,000
Powered by AMD's 16-core Zen 5 + 40-CU RDNA 3.5 'Strix Halo' APU, the 4.5-liter modular Framework Desktop runs 100B+ MoE models locally at 38.6 tokens/sec for $1,000 less than proprietary AI boxes.
⚡ Key Engineering & Supply-Chain Takeaways
- Framework's 4.5L Mini-ITX desktop pairs AMD's Ryzen AI Max+ 395 ('Strix Halo') with 128GB of soldered LPDDR5X-8000 unified memory (256 GB/s) at $1,999.
- Under Linux Kernel 6.14+ and ROCm 6.4 (`gfx1151`), up to 112GB can be allocated directly to the 40-CU Radeon 8060S GPU via BIOS UMA or dynamic TTM page pools.
- Executes Llama-4-70B (Q4_K_M) at 14.8 tokens/sec and 96GB sparse Mixture-of-Experts (MoE) coding models at 38.6 tokens/sec while staying below 28 dBA with Noctua cooling.
🔬 Technical & Supply-Chain Specification Matrix
| APU Silicon | AMD Ryzen AI Max+ 395 (16C Zen 5 + 40-CU Radeon 8060S + 50 TOPS XDNA 2) |
|---|---|
| Unified Memory Pool | 128GB LPDDR5X-8000 (256 GB/s Bandwidth, Up to 112GB GPU-Allocatable) |
| Llama-4-70B (Q4_K_M) Speed | 14.8 tok/s Decode / 195 tok/s Prefill (ROCm 6.4) |
| 100B-Class MoE (Active 14B) | 38.6 tok/s Decode (Fits Entirely in 112GB UMA Partition) |
| Chassis & Reference Price | 4.5L Modular Mini-ITX w/ 400W Flex-ATX PSU / $1,999 USD |
Breaking the x86 VRAM Wall with AMD Strix Halo
For developers who prefer native x86 Linux over macOS or proprietary Arm appliances, the $1,999 Framework Desktop (Ryzen AI Max+ 395 128GB) changes the economics of local LLM hosting. By pairing a 16-core Zen 5 CPU complex and a 40-Compute-Unit Radeon 8060S GPU across a 256-bit LPDDR5X-8000 memory interface, the 4.5-liter Mini-ITX workstation exposes up to 112GB of coherent GPU-accessible memory under ROCm 6.4 (gfx1151).
In RECIWO Labs' benchmark testing, the Framework Desktop ran 70B Q4_K_M models at 14.8 tokens/sec (within 18% of the $2,999 NVIDIA DGX Spark GB10) and excelled on 96GB sparse Mixture-of-Experts (MoE) models that cannot fit inside 64GB Mac mini nodes, sustaining 38.6 tokens/sec at just 132W wall power.
🔗 Related AI Hardware in RECIWO Database
All 38 Devices →Framework Desktop (Ryzen AI Max+ 395 128GB)
4.5-liter modular Mini-ITX workstation powered by AMD's 16-core Ryzen AI Max+ 395 ('Strix Halo') APU with a 40-CU Radeon 8060S GPU and 128GB unified LPDDR5X-8000 memory for local 70B–120B LLM execution.
NVIDIA DGX Spark (GB10 Grace Blackwell Superchip)
Palm-sized personal AI supercomputer powered by the NVIDIA GB10 Grace Blackwell Superchip with 128GB coherent unified LPDDR5X memory and ConnectX-7 200GbE clustering.
Apple Mac mini (M5 Pro 64GB) Private AI Node
Packs 64GB of high-bandwidth unified memory and Thunderbolt 5 into a 5×5-inch footprint, running 32B–70B open-weight LLMs locally at under 35W.
📚 Related Deep-Dive Engineering Teardowns
Browse All 83 Articles →Framework Desktop (Ryzen AI Max+ 395 128GB) Local LLM Benchmark: AMD Strix Halo vs NVIDIA DGX Spark GB10 vs Mac mini M5 Pro
RECIWO Labs benchmarked the $1,999 4.5L Framework Desktop powered by AMD's 16-core Ryzen AI Max+ 395 ('Strix Halo') and 128GB unified LPDDR5X-8000 memory, comparing ROCm 6.4 token generation and prompt prefill speeds against the $2,999 NVIDIA DGX Spark (GB10) and $1,799 Apple Mac mini M5 Pro.
NVIDIA DGX Spark (GB10 128GB) Desk Supercomputer Review: Running 200B FP4 Local LLMs at 140W
RECIWO Labs tested the $2,999 NVIDIA DGX Spark (GB10 Grace Blackwell Superchip, 128GB LPDDR5X) against the 64GB Mac mini M5 Pro and dual RTX 5090 towers across 70B FP8 and 200B NVFP4 models.
Home Private AI Server Benchmark 2026: Apple Mac mini M5 Pro (64GB) vs. Dual-GPU Edge Nodes for 70B Local LLMs
RECIWO Labs tested 32B and 70B local LLMs across the 64GB Mac mini M5 Pro, NVIDIA Jetson Orin NX Super, and RTX 5090 workstations, measuring tokens/sec per watt and 24/7 acoustic noise.
