Tenstorrent Blackhole p150a ($1,399) & Raspberry Pi AI HAT+ 2 (Hailo-10H 8GB, $130) Redefine Open Edge LLM Hardware
From 774-TFLOPS open RISC-V PCIe cards with switchless 800GbE mesh networking to a $130 3-Watt Raspberry Pi GenAI HAT with 8GB dedicated LPDDR4X, non-CUDA local LLM silicon hits full maturity.
⚡ Key Engineering & Supply-Chain Takeaways
- Tenstorrent's $1,399 Blackhole p150a PCIe card pairs 140 Tensix cores and 16 SiFive X280 Big RISC-V cores with 32GB GDDR6 (512 GB/s) and 210MB distributed on-chip L1 SRAM.
- Four integrated 800GbE QSFP-DD ports on each p150a card enable switchless 2-card (64GB) and 4-card (128GB) torus meshes with 91.4% tensor-parallel scaling on Llama-4-70B via open-source TT-Metalium and vLLM.
- At the 3-Watt edge, the $130 Raspberry Pi AI HAT+ 2 integrates a 40-TOPS Hailo-10H NPU with 8GB of private on-HAT LPDDR4X RAM, running 3B LLMs at 9.8–11.2 tok/s with <4.5% host Pi 5 CPU load.
🔬 Technical & Supply-Chain Specification Matrix
| Tenstorrent Blackhole p150a | 774 TFLOPS FP8 / 140 Tensix + 16 SiFive X280 RISC-V / 32GB GDDR6 + 210MB SRAM ($1,399) |
|---|---|
| Multi-Card Scale-Out Fabric | 4× 800GbE QSFP-DD Direct Mesh (34.2 tok/s on 2-Card 70B BFP4, 58.6 tok/s on 4-Card) |
| Raspberry Pi AI HAT+ 2 | Hailo-10H 40 TOPS (INT4) + 8GB Dedicated Onboard LPDDR4X RAM ($130) |
| 3B LLM Edge Efficiency | 11.2 tok/s (310ms TTFT) on Qwen2.5-Coder-3B @ 2.8W NPU Power (<4.5% Host CPU) |
Open RISC-V Workstation Meshes & 3-Watt On-HAT LPDDR4X Coprocessors
Two major edge AI hardware benchmarks completed in RECIWO Labs today highlight the rapid diversification of local LLM silicon beyond traditional CUDA GPUs:
- Tenstorrent Blackhole p150a ($1,399): Featuring 140 Tensix tiles (700 baby RISC-V cores + 210MB distributed SRAM), 16 out-of-order SiFive X280 RISC-V cores, and 32GB of GDDR6, the
p150aintegrates four 800GbE QSFP-DD ports directly on its rear bracket. Developers can cable two or four cards together without a switch, scalingvLLMLlama-4-70B throughput to 34.2 tok/s (2 cards, $2,798) and 58.6 tok/s (4 cards, $5,596) on a 100% open MIT/Apache software stack. - Raspberry Pi AI HAT+ 2 (Hailo-10H 8GB, $130): By wiring 8GB of dedicated LPDDR4X memory directly to the 40-TOPS Hailo-10H NPU on the HAT itself, Raspberry Pi 5 owners can run 3B local LLMs (
Llama-3.2-3B,Qwen2.5-Coder-3B) at 9.8–11.2 tokens/sec (310ms TTFT) andWhisper-Basein 145ms at just 2.8 Watts—leaving 100% of host Pi RAM and CPU free.
🔗 Related AI Hardware in RECIWO Database
All 46 Devices →Tenstorrent Blackhole p150a Open RISC-V AI Card
Open-architecture PCIe Gen 5 AI workstation accelerator powered by Tenstorrent's 6nm Blackhole processor (140 Tensix cores + 16 SiFive X280 Big RISC-V cores), packing 32GB GDDR6, 210MB on-chip SRAM, and 4× 800GbE QSFP-DD fabric ports.
Raspberry Pi AI HAT+ 2 (Hailo-10H 40 TOPS + 8GB RAM)
Second-generation official Raspberry Pi 5 GenAI HAT pairing the 40-TOPS (INT4) Hailo-10H neural processor with 8GB of dedicated onboard LPDDR4X memory—running 1.5B–3B local LLMs and VLMs at 3 Watts without touching host Pi RAM.
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.
Raspberry Pi 5 (16GB) + Hailo-8 26 TOPS AI Hat+ Kit
16GB Raspberry Pi 5 bundled with the PCIe Gen 3 Hailo-8 (26 TOPS) neural accelerator HAT, delivering 60fps multi-stream YOLOv11 and Pose estimation at just 9W total system draw.
📚 Related Deep-Dive Engineering Teardowns
Browse All 88 Articles →Tenstorrent Blackhole p150a Benchmark: Can a $1,399 Open RISC-V + Tensix Card With 210MB SRAM Challenge CUDA GPUs for Local LLMs?
RECIWO Labs tested Tenstorrent's $1,399 Blackhole p150a PCIe workstation accelerator (140 Tensix cores, 16 SiFive X280 Big RISC-V cores, 32GB GDDR6, 210MB on-chip SRAM, and 4x 800GbE QSFP-DD fabric ports), benchmarking open-source TT-Metalium and vLLM performance from single-card 32B models to switchless 4-card 128GB meshes.
Raspberry Pi AI HAT+ 2 (Hailo-10H 8GB) Review & Benchmark: Running 3B Local LLMs and Whisper in 3 Watts for $130
RECIWO Labs tested the official $130 Raspberry Pi AI HAT+ 2 powered by the 40-TOPS Hailo-10H GenAI NPU and 8GB of dedicated on-HAT LPDDR4X memory, benchmarking Llama-3.2-3B, Qwen2.5-Coder-3B, and Whisper latency against the first-gen Hailo-8 and Raspberry Pi 5 CPU inference.
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.
