← All hardware
Open RISC-V AI server: 23 PFLOPS from 32 Blackhole chips
Pros
- Open, Ethernet-native scale-out
- No proprietary interconnect lock-in
- Strong price/performance
- Adopted by Cirrascale, Equinix, ai&
Cons
- Smaller software ecosystem
- p150 card tensor-core firmware downgrade noted
- Newer, less battle-tested
- Self-integration effort
✓ Where it shines / best for
- Enterprises and labs building datacenter AI training/inference clusters
- Organizations seeking an open, non-NVIDIA accelerator platform at scale
- Large-model inference serving
✕ Not the best fit for
- Individual developers or small teams (workstation cards fit better)
- Buyers needing public transparent pricing and instant availability
- Graphics/gaming workloads
Features
- ✓ Inference
- ✓ Training
- ✓ Rack Scale
- ✓ Open source
- ✓ AI Accelerator
- ✓ Open Source Stack
- ✓ Risc V
- ✓ Datacenter
- ✓ Ethernet Fabric
Pricing
| Plan | Price | Billing | Notes |
|---|---|---|---|
| Blackhole Galaxy (server) | Contact sales / approx. $100,000+ | one-time | Server-class system; pricing not publicly listed, quote-based |
Pricing verified from the official source. Prices change often — confirm on the vendor's site before buying.
Specifications
| use | Data center inference/training |
| power | Server rack scale |
| memory | 1TB DRAM + 6.2GB SRAM |
| performance | 23 PFLOPS Block FP8 |
| architecture | Blackhole (RISC-V Tensix) |
Sponsored
A full review is being generated for this product and will appear here shortly.