Multi-GPU AI workstation, up to four RTX PRO Blackwell GPUs.
The JNCS AI MAX runs one, two, or four NVIDIA RTX PRO Blackwell GPUs, each on a dedicated PCIe 5.0 x16 slot, on a Threadripper PRO 9000 processor and an ASUS Pro WS WRX90E-SAGE SE board with eight-channel registered ECC memory. It is built for training, multi-user inference serving, and the largest models that fit in a tower.
- 384GBVRAM, 4× RTX PRO 6000
- 16–96cores, Threadripper PRO
- 128GB–1TBDDR5 ECC RDIMM
- 128PCIe 5.0 lanes
GPU configurations and total VRAM
One, two, or four of the same card. Each GPU runs at a full PCIe 5.0 x16 link.
| GPU | 1 GPU | 2 GPUs | 4 GPUs |
|---|---|---|---|
| RTX PRO 4500 Blackwell, 32GB, 200W | 32GB | 64GB | 128GB |
| RTX PRO 5000 Blackwell, 48GB, 300W | 48GB | 96GB | 192GB |
| RTX PRO 5000 Blackwell, 72GB, 300W | 72GB | 144GB | 288GB |
| RTX PRO 6000 Blackwell Workstation Edition, 96GB, 600W | 96GB | 192GB | 384GB |
Workstation specifications
Fixed platform parts are listed as-is; ranges are the options offered in the configurator.
| Processor | AMD Ryzen Threadripper PRO 9000 WX-Series on sTR5: 9955WX (16 cores), 9965WX (24), 9975WX (32), 9985WX (64), or 9995WX (96 cores, 192 threads). 128 PCIe 5.0 lanes, 350W. |
|---|---|
| Motherboard | ASUS Pro WS WRX90E-SAGE SE (SSI-EEB, eight-channel DDR5, four M.2 slots) |
| Memory | Kingston registered ECC DDR5, one module per channel: 128GB (8×16GB), 256GB (8×32GB), 512GB (8×64GB), 768GB (8×96GB), or 1TB (8×128GB) |
| Storage | Samsung 9100 PRO PCIe 5.0 NVMe, 1TB to 8TB per drive, up to four drives. Optional: up to two Seagate Exos 7200 RPM enterprise hard drives, 12TB to 28TB each, as independent volumes or a RAID 1 or RAID 0 pair |
| Cooling | ARCTIC Freezer 4U-M Rev. 2, air cooler for sTR5 |
| Power supply | Auto-sized: ASRock PG-1600G 1600W (80 Plus Gold, 2× 12V-2x6) or SilverStone HELA 3000Rz 3000W (80 Plus Platinum, 4× 12V-2x6) |
| Case | SilverStone WS380-E, SSI-EEB workstation tower with eight hot-swap 3.5-inch bays |
| Operating system | Windows 11 Pro, installed and updated |
Why Threadripper PRO for four GPUs
128 PCIe 5.0 lanes
Four GPUs at x16 need 64 lanes for the cards alone. Threadripper PRO 9000 provides 128 PCIe 5.0 lanes, so every GPU runs at a full x16 link with lanes left for NVMe storage.
Eight memory channels
Registered ECC DDR5 with one module in each of eight channels, from 128GB to 1TB. A common guideline is system memory at least equal to total VRAM; the configurator flags a selection below that.
16 to 96 cores
Five processors from the 16-core 9955WX to the 96-core, 192-thread 9995WX, for data loading, preprocessing, and CPU-side work alongside the GPUs.
3000W power delivery
Two RTX PRO 6000 cards and every four-GPU configuration use the 3000W SilverStone HELA 3000Rz, which has a 12V-2x6 cable for each of four GPUs.
Power and circuit requirements
Estimated draw at full load with a 350W Threadripper PRO processor. Confirm the circuit is in place before delivery.
| GPU configuration | Estimated draw | Power supply | Circuit |
|---|---|---|---|
| 1× any card, or 2× RTX PRO 4500 or 5000 | 650W to 1,050W | 1600W | Standard 15A, 120V |
| 2× RTX PRO 6000 | approximately 1,650W | 3000W | Dedicated 20A, 110–120V required; 30A, 220–240V recommended |
| 4× RTX PRO 4500 or 5000 | 1,250W to 1,650W | 3000W | Dedicated 20A, 110–120V required; 30A, 220–240V recommended |
| 4× RTX PRO 6000 | approximately 2,850W | 3000W | Dedicated 30A, 220–240V required |
A standard 15A, 120V circuit supplies 1,440W continuously. The SilverStone HELA 3000Rz accepts 100–240V input and delivers its full 3000W on a 200–240V circuit.
Assembly, burn-in, and delivery
Each system is assembled by hand at our shop in Fairport, NY, by a shop that has built computers for more than thirty years.
Fit and power validated
The configurator checks physical fit, memory, PCIe slots, and power on every selection. The power supply is sized automatically to the GPU count and estimated draw, with a 12V-2x6 cable for every GPU.
Burn-in under load
Every system is stress-tested under sustained full load and firmware-updated before it leaves the bench, so the first long job is not the first test.
Windows 11 Pro installed
Windows 11 Pro is installed and updated before shipment.
Warranty and support
Every build carries a 3-year warranty and lifetime technical support from the shop that built it.
Quoting and payment
The configurator shows a price for every part and an estimated total, including assembly, burn-in, and testing, as the build is configured. GPU and memory pricing changes frequently, so a technician confirms GPU availability and final pricing by email, with lead time, typically within 1-2 business days. There is no cost and no obligation to request a quote.
Payment for AI workstations is by ACH bank transfer; credit cards are not accepted for these systems. An invoice with bank details is issued once the quote is approved. Systems ship across the United States from Fairport, NY, with local pickup available.
Multi-GPU workstation questions
How much VRAM does a four-GPU workstation have?
What power does a 4× RTX PRO 6000 workstation need?
Does each GPU get a full x16 PCIe slot?
How much system memory does a multi-GPU AI workstation need?
Can it be configured with more than four GPUs, or as a rackmount system?
How is a multi-GPU workstation ordered and paid for?
Related AI workstation builds
RTX 5090 AI workstation
One GeForce RTX 5090, 32GB VRAM, Ryzen 9000, standard 15A circuit.
View →RTX PRO 6000 Blackwell workstation
One or two RTX PRO Blackwell GPUs, up to 192GB VRAM, Threadripper 9000.
View →Local LLM workstation guide
How much VRAM a model needs, and which build provides it.
View →Configure a multi-GPU AI workstation.
Select the GPU and count, the processor, memory, and storage. The power supply and the circuit requirement update with every selection.
Have an exact parts list or a department spec sheet? Send the whole list. For more than four GPUs or rackmount systems, see AI servers and larger systems.