- This item is not returnable
Shipped by NeweggBusiness
ABS AI Workstation - 96GB VRAM (NVIDIA RTX PRO 6000 Blackwell), Intel Xeon W5-2455X, 64GB ECC DDR5 RDIMM, 1TB NVMe M.2 + 1.92TB SATA SSD, Ubuntu (Zaurion ZAW5-2455X-RP6000)
- Item #: 9B-59-991-004
- Mfr. Part #: ZAW5-2455X-RP6000
- Compliance Documents Required for Purchase
- Assembled in USA
- TAA Compliant
- Operating System: Ubuntu
- CPU: Intel Xeon W5-2455X
- GPU: RTX Pro 6000 Blackwell (96 GB GDDR7)
- Motherboard: Gigabyte MW83-RP0
- Memory: 64GB DDR5 Server RAM - ECC, Registered
- OS SSD: 1TB M.2 NVMe (Installed)
- Data SSD: 1.92TB SSD (Installed)
- LAN: 2 x 10GB/s LAN, 1 x 10/100/1000 Mbps Management LAN
- PSU: Single 2000W ATX 80 PLUS Gold power supply
- Barebone: Gigabyte W773-W80
- AI Agent Ready: Optional pre-installed OpenClaw (enterprise AI agent framework)
- Local LLM Support: Optional Qwen 3.5 27B or higher models for on-prem AI inference
The Quiet Way In.
This is the most accessible seat at the family's flagship table: the entry tier of Intel's Xeon W-2400 series paired with the full 96 GB of the RTX PRO 6000 Blackwell — no step up to a bigger CPU required. And it's engineered to run at under 50 dB at full GPU load, so it disappears into a shared office instead of dominating it.
What This Pairing Actually Buys You
An entry-tier workstation CPU next to the family's flagship GPU is a deliberate choice, not a compromise:
96 GB of GDDR7 — Nothing Held Back
- The full 96 GB is identical to what the family's Xeon W9 and Threadripper PRO configurations carry — stepping down the CPU doesn't shrink the GPU.
- Universal MIG partitions the card into isolated instances for concurrent workloads or users.
- Blackwell 5th-gen Tensor Cores with FP4 accelerate agentic and generative AI at architecture-native speed.
- 4th-gen RT Cores drive photoreal, physically accurate rendering for design and simulation work.
Workstation-Grade, Without the Premium
- 30 MB cache, 3.20 GHz base1 — the entry point of Intel's Xeon W-2400 series, built for professional workstation duty, not desktop compromises.
- EMIB packaging connects multiple dies in one socket — the same architecture family that scales up to 60 cores on larger Xeon W SKUs.
- ECC memory and RAS features bring server-grade reliability to a single-CPU workstation.
- Feeds the GPU, doesn't bottleneck it — for GPU-bound AI inference and rendering, VRAM capacity matters more than CPU core count.
Full GPU Load. Under 50 dB.
Internal airflow diversion is the whole trick — heat leaves the case before it becomes noise:
Front: 1× USB 3.2 Gen1, power button, audio/mic jack, and 4× 3.5″/2.5″ SATA hot-swap bays installed (4 more optional — 8 total possible). Rear: legacy COM/VGA for lab and rack instrumentation, USB 3.2 Gen2x2 Type-C plus 3× USB 3.2 Gen2 Type-A, dual 10Gb/s LAN (Intel X710-AT2) plus a dedicated MLAN management port, and the audio/power connectors — a rear panel built for a rack room, wearing a tower's clothes.
The Gigabyte board underneath ships with one GPU installed, but the platform itself is rated for far more: 8× RDIMM slots on 8-channel DDR5, 5× PCIe 5.0 x16 slots (platform-rated for up to 4× dual-slot GPUs), 2× PCIe 4.0 x16 slots, and 2× M.2 SSD slots with only one populated by the 1 TB OS drive.2 Room to grow into, not just a spec sheet to grow out of.
Six Jobs, Same Quiet Desk
The listing's own AI Features, translated into a shared-office reality:
AI Development
Accelerate AI development and inference workloads, and build agentic AI applications.
Data Science
Accelerate your data pipelines.
HPC
Accelerate scientific breakthroughs with enhanced accuracy.
AI-Driven Rendering & Graphics
Deliver stunningly realistic designs and simulations with speed and precision.
Video Content & Streaming
Boost performance for live media and video pipelines.
Game Development
Drive new possibilities for game development with AI.
Scene photos are AI-generated illustrations of typical deployments, not photographs of this exact product.
Fine-Tune Past 96 GB with Phison aiDAPTIV+
An optional add-on that offloads GPU memory to flash, reducing the VRAM full fine-tuning actually requires:
| Target Model Size | VRAM Needed (Without aiDAPTIV+) | VRAM Needed (With aiDAPTIV+) |
|---|---|---|
| ~13B | 120–200 GB | 32 GB |
| ~34B | 250–350 GB | 96 GB — this workstation's GPU |
| ~70B | 500–700 GB | 192 GB |
Phison aiDAPTIV+ SSD is an optional add-on and is not included by default. VRAM figures are vendor-published estimates for full fine-tuning.
Where This Sits in the Family
Every rung of this family's 96GB tier carries the identical RTX PRO 6000 Blackwell — only the CPU platform changes:
Intel Xeon W5-2455X
The Xeon W-2400 series entry point. The most accessible door onto the full 96 GB GPU in this family.
Intel Xeon W9-3575X
Higher core-count Xeon W tier on the same platform and the same 96 GB GPU.
AMD Threadripper PRO
A server-pedigree AMD CPU alternative, also paired with the 96 GB RTX PRO 6000 Blackwell — compare options on this listing.
Zaurion Duo Aqua
The same family's dual-GPU tower for workloads that need more than one 96 GB card.
As Configured
| ZAW5-2455X-RP6000 | |
|---|---|
| CPU | Intel Xeon W5-2455X — Intel Xeon W-2400 series, 30MB cache, 3.20GHz base1 |
| GPU | NVIDIA RTX PRO 6000 Blackwell — 96 GB GDDR7 |
| Memory | 64 GB DDR5 ECC Registered (RDIMM) |
| Storage | 1 TB M.2 NVMe (OS) + 1.92 TB SSD (Data) |
| Networking | 2× 10 Gb/s LAN + 10/100/1000 Mbps management LAN |
| Motherboard | Gigabyte MW83-RP0 (Gigabyte W773-W80 barebone) |
| Power | Single 2000 W ATX 80 PLUS Gold |
| Acoustics | Under 50 dB at full GPU load (internal airflow diversion) |
| OS | Ubuntu |
| Optional | Phison aiDAPTIV+ SSD; OpenClaw/NemoClaw agent framework; Qwen 3.5 27B+ local LLM package |
Quick Answers Before You Buy
Why is this the most accessible way onto the 96 GB GPU tier?
The Xeon W5-2455X is the entry point of this family's Xeon W-2400 CPU tier, paired with the same 96 GB RTX PRO 6000 Blackwell offered on the higher Xeon W9 and Threadripper PRO configurations — the GPU doesn't shrink when the CPU steps down.
How quiet is it really?
Under 50 dB at full GPU load, via internal airflow diversion — engineered to sit in a shared office without becoming a distraction.
Does it come with an AI agent framework or local LLM pre-installed?
Not by default. OpenClaw/NemoClaw and a Qwen 3.5 27B+ local LLM package are optional add-ons configured at order time.
What is Phison aiDAPTIV+ and do I need it?
An optional SSD that offloads GPU memory to flash, reducing the VRAM full fine-tuning requires — useful for fully fine-tuning models larger than 96 GB alone can hold. Not included by default.
Is it eligible for government / public-sector purchasing?
It is TAA compliant and assembled in the USA, with compliance documents available for purchase review.
How is it supported?
1-year parts and labor limited warranty backed by ABS manufacturer support. As a Build to Order system, it is custom built and stress-tested after purchase.
Warranty & Returns
Warranty, Returns, And Additional Information
Warranty
- Limited Warranty period (parts): 1 year
- Limited Warranty period (labor): 1 year
Return Policies
- Return for refund within: non-refundable
- Return for replacement within: non-replaceable
- This item is covered by NeweggBusiness.com's Manufacturer Only Return Policy.
- Read full return policy for details.
Manufacturer Contact Info
- Manufacturer Product Page
- Manufacturer Website
- Support Phone: 1-800-685-3471
- Support Email: customercare@abs.com
- Support Website
- View other products from ABS













