Same chip, different box
NVIDIA DGX Station vs GIGABYTE W775-V10-L01
Memory, bandwidth, software stack, price and the models each one can hold. Generated from the two sourced machine records; estimates are labelled.

NVIDIA GB300
NVIDIA DGX Station
496 GB CPU + 252 GB GPU · 748 GB coherent
Open the record
NVIDIA GB300
GIGABYTE W775-V10-L01
496 GB CPU + 252 GB GPU
Open the recordWhat actually differs
- Same GB300 platform. Model recipes carry over; the real differences are chassis, storage, cooling, support and price.
- Both expose 252 GB to models (496 GB CPU + 252 GB GPU · 748 GB coherent vs 496 GB CPU + 252 GB GPU).
- Same memory bandwidth (7100 GB/s), so the same decode ceiling for a model that fits on both.
- No public list price in the record for either machine. Ask the supplier for a configured quote.
Which models fit
Estimate · 4-bit · 8K context · 1 request| Model | NVIDIA DGX Station | GIGABYTE W775-V10-L01 |
|---|---|---|
| Llama 3.1 8B | Fits 12 GB≤ 1251 tok/s | Fits 12 GB≤ 1251 tok/s |
| Qwen3 14B | Fits 16 GB≤ 721 tok/s | Fits 16 GB≤ 721 tok/s |
| Qwen3 32B | Fits 28 GB≤ 338 tok/s | Fits 28 GB≤ 338 tok/s |
| Llama 3.3 70B | Fits 50 GB≤ 164 tok/s | Fits 50 GB≤ 164 tok/s |
| gpt-oss-20b (MoE) | Fits 19 GB≤ 2871 tok/s | Fits 19 GB≤ 2871 tok/s |
| gpt-oss-120b (MoE) | Fits 76 GB≤ 2008 tok/s | Fits 76 GB≤ 2008 tok/s |
| Qwen3 235B-A22B (MoE) | Fits 147 GB≤ 499 tok/s | Fits 147 GB≤ 499 tok/s |
| DeepSeek V3 / R1 671B (MoE, MLA) | Too large 404 GB | Too large 404 GB |
Specifications
Sourced and dated| NVIDIA DGX Station | GIGABYTE W775-V10-L01 | |
|---|---|---|
| Chip / accelerator | GB300 Grace Blackwell Ultra Desktop Superchip | GB300 Grace Blackwell Ultra |
| Platform | NVIDIA GB300 | NVIDIA GB300 |
| Model memory | 496 GB CPU + 252 GB GPU · 748 GB coherent | 496 GB CPU + 252 GB GPU |
| Memory bandwidth | GPU 7.1 TB/s · CPU 396 GB/s · NVLink-C2C 900 GB/s | GPU 7.1 TB/s · CPU 396 GB/s |
| Software stack | CUDA | CUDA |
| Operating system | NVIDIA DGX OS (Ubuntu) · Windows edition announced for Q4 2026 | Ubuntu with NVIDIA AI Developer Tools |
| Storage | 4 × M.2 PCIe Gen 5 slots · drives configured by the partner | 4 M.2 slots · SSDs selected at integration |
| Networking | ConnectX-8 SuperNIC · 2 × 400 GbE QSFP112 · 10 GbE · 1 GbE BMC | 2 × 400 GbE QSFP · 10 GbE · management LAN |
| Form factor | Deskside workstation | Deskside workstation |
| Dimensions | Set by each partner’s chassis | 245 × 500.4 × 531 mm · W × H × D |
| Weight | Set by each partner’s chassis | 29.2 kg net |
| Power rating (not consumption) | 1,600 W platform system-power specification · 20 A circuit specified by NVIDIA | 1,600 W Platinum PSU · NVIDIA platform guidance specifies a 20 A circuit |
| Supplier list price | Price from the manufacturer · No public list price · manufacturer page | Price from the manufacturer · No public list price · manufacturer page |
| Availability | Order through partner manufacturers · configuration and price set by each partner | Check regional availability |
| Model configurations in the Lab | 4 | 4 |
| Setups in the Lab | 3 | 3 |
| Record reviewed | 2026-09-17 | 2026-09-07 |
Before you choose
NVIDIA DGX Station
- 748 GB coherent memory is 252 GB HBM3e on the GPU plus 496 GB LPDDR5X on the CPU, joined by NVLink-C2C. The two tiers have very different bandwidths; large models spill from the fast tier into the slow one.
- NVIDIA does not sell the DGX Station directly: it is ordered through partner manufacturers, each with its own chassis, drives, warranty and price. Partner listings observed in 2026 ran from about $85,000 to $175,000 depending on configuration.
- NVIDIA specifies 1,600 W total system power and a 20 A circuit. Confirm OEM and regional installation requirements before choosing an office outlet. The GB300 can be paired with an RTX PRO 6000 Blackwell workstation GPU for graphics and additional compute.
GIGABYTE W775-V10-L01
- This vendor reference is a barebone workstation platform. A supplier must specify SSDs, OS, display GPU and the final ready-to-use configuration.
- 252 GB GPU plus 496 GB CPU memory; power and circuit requirements need checking before installation.
Own one of these? A measured run (model revision, runtime, context, tokens per second) is worth more than this whole page. Publish it.