The New APEXX T4 PRO-X Workstation: AMD Ryzen Threadripper PRO Meets NVIDIA RTX Blackwell GPUs
Key points about the new APEXX T4 PRO-X Workstation:
- 2050W Power Supply Enables Maximum GPU Capability. Run dual 600W RTX PRO 6000 Blackwell Workstation Edition GPUs or up to four RTX PRO 6000 Blackwell Max-Q Workstation Edition GPUs in a single system, ideal for large-scale AI, rendering, and simulation without performance limitations.
- 96-Core AMD Threadripper PRO for Unmatched CPU Performance. Up to 192 threads and 128 PCIe Gen5 lanes deliver the throughput needed for data-heavy, multi-threaded workflows.
- Blackwell Architecture Built for Scalable AI Development. From local prototyping to enterprise deployment, the unified RTX PRO platform accelerates every stage of professional workloads.
- Optimized for AI Training, Design, and Simulation Workflows. Purpose-built for large model training, high-res content creation, and complex engineering simulation, without cloud dependence.
- The BOXX Advantage: Tuned, Reliable, and Scalable. Application-optimized performance, U.S.-built reliability, and upgrade-ready design ensure the T4 PRO-X meets today's demands and tomorrow’s growth.
The demands of modern professional workloads continue to escalate, requiring hardware that can scale far beyond conventional limits. AI training, rendering, simulation, and design workflows all push systems to their breaking point, exposing the need for GPU power, massive CPU core counts, and reliability. The new APEXX T4 PRO-X delivers on all three. Built with AMD Ryzen™ Threadripper™ PRO 9000 WX-Series processors and a 2050W power supply, it’s designed to support dual NVIDIA RTX PRO™ 6000 Blackwell GPUs or four RTX PRO™ 6000 Blackwell Max-Q GPUs. This combination sets a new standard for workstation performance, enabling faster processing, higher throughput, and sustained operation under maximum load.
Powering AMD Ryzen Threadripper PRO and NVIDIA RTX: The 2050W Workstation Advantage
At the core of the APEXX T4 PRO-X is a 2050W power supply designed to push workstation performance beyond previous limits. Supporting up to two 600W GPUs in one system, it creates the headroom required for next-generation workflows, delivering the kind of performance once limited to multi-node or datacenter deployments. By ensuring consistent, reliable power under maximum load, the T4 PRO-X enables professionals to run rendering, AI training, and complex simulations at full scale without compromise.
Next-Level Performance Delivery
- Integrates dual 600W GPUs in a single workstation.
- Built for continuous, 24/7 operation in production environments.
- Maintains consistent output under sustained maximum demand.
Reliability is not only about power delivery but also stability under real workloads. The T4 PRO-X power supply is engineered for long-duration performance, providing the assurance that demanding projects won’t be interrupted by throttling or instability.
Advanced Cooling and Stability
- Proprietary thermal design ensures quiet, efficient airflow.
- Keeps multi-GPU systems running at peak performance.
- Eliminates performance drops by preventing thermal throttling.
For professionals operating at scale, whether developing AI models, rendering massive datasets, or running multi-threaded simulations, this balance of power and thermal control ensures maximum throughput, faster iteration cycles, and uninterrupted productivity.
Building AI Workflows with NVIDIA RTX Blackwell GPUs
The NVIDIA RTX PRO Blackwell series marks a generational leap in GPU architecture, redefining what is possible for AI, technical, creative, engineering, and design professionals. Built for both workstation and datacenter deployment, Blackwell delivers unmatched scalability, performance, and efficiency across every stage of modern workflows. From local prototyping to enterprise-scale production, it provides a unified platform that accelerates discovery, design, and innovation.
Scaling AI Performance Across Workstation and Server Deployments
The NVIDIA RTX PRO Blackwell lineup delivers a unified GPU platform for every stage of professional AI, design, and simulation workflows:
- RTX PRO 2000 Blackwell. 16GB GDDR7. Compact, power-efficient, and built for pro graphics and AI.
- RTX PRO 4000 Blackwell. 24GB GDDR7. Entry-point for professional AI acceleration and high-performance visualization.
- RTX PRO 4500 Blackwell. 32GB GDDR7. The 4500 Blackwell delivers next-level AI and neural rendering with latest gen Tensor, RT, and CUDA cores.
- RTX PRO 5000 Blackwell. 48GB GDDR7. Ideal for AI development, LLM inference, and advanced content creation.
- RTX PRO 6000 Blackwell Max-Q Workstation Edition. 96GB GDDR7. Packing a 300W punch into an ultra-efficient design, the Max-Q redefines high-density workstation performance. Optimized for energy and space efficiency.
- RTX PRO 6000 Blackwell Workstation Edition. 96GB GDDR7. The most powerful desktop GPU available, engineered for simulation, rendering, and full-scale AI training.
- RTX PRO 6000 Blackwell Server Edition. Brings RTX performance to the datacenter, enabling high-throughput AI factories and enterprise-scale compute environments.
This range ensures seamless workflow scalability, from local prototyping on the desktop to production deployment in the datacenter.
With this ecosystem, professionals gain a consistent development environment from desktop to datacenter. That means workflows built on the APEXX T4 PRO-X can transition seamlessly into enterprise-scale deployment, ensuring efficiency and scalability at every stage of production.
Unified GPU Architecture with NVIDIA RTX PRO Blackwell
The NVIDIA RTX PRO Blackwell generation is built on a unified architecture designed to accelerate AI, science, and design at every scale, from desktop prototyping to full enterprise deployment. This consistency allows professionals to develop, train, and deploy on the same platform without reconfiguring workflows or compromising performance.
Key advantages include:
- Multi-Instance GPU (MIG) support: Each RTX PRO 6000 can be partitioned into as many as four fully isolated GPU instances with 24GB each, enabling multiple workloads or users to run concurrently without interference.
- Seamless scaling: Workflows can move directly from prototype to production, ensuring that development done locally on the APEXX T4 PRO-X transitions easily to datacenter systems.
- Consistent environments: From workstation to server, professionals work within the same architecture, eliminating the inefficiencies of fragmented toolchains.
The Blackwell ecosystem also integrates tightly with NVIDIA’s broader AI and graphics platforms. As shown in the launch framework, RTX PRO GPUs extend across the entire development pipeline:
- NVIDIA AI Workbench simplifies experimentation and local model training.
- NVIDIA Omniverse powers real-time collaboration and advanced visualization for digital humans, robotics, content creation, and design simulation.
- NVIDIA AI Enterprise provides enterprise-class scalability for deep learning, simulation, and data analytics.
This unified approach ensures that whether the workload is AI development, LLM inference, image/video generation, or simulation, professionals can accelerate every stage without retooling environments or moving between incompatible platforms. The result is a complete development ecosystem designed for speed, scale, and reliability.
Enabling Enterprise AI Workloads with RTX PRO Blackwell Server Integration
The RTX PRO 6000 Blackwell Server Edition extends workstation-class performance into the datacenter, providing a direct path from development on the APEXX T4 PRO-X to full-scale enterprise deployment. By integrating into the world’s most popular enterprise servers, it accelerates the transition from traditional CPU-based infrastructure to GPU-driven computing platforms purpose-built for AI and simulation workloads.
Key advantages include:
- Seamless development-to-deployment workflow – Start with model development or simulation locally on the T4 PRO-X, then move directly into datacenter-scale execution without changing platforms or toolchains.
- Accelerated enterprise adoption – Blackwell-powered servers enable faster migration from legacy CPU systems to modern GPU-accelerated architectures.
- Universal workload acceleration – Supports the full spectrum of professional demands, from agentic and physical AI to advanced design, scientific computing, engineering simulation, graphics, and high-end video production.
For enterprises building AI factories and research pipelines, this integration ensures that innovation scales without friction. Professionals can develop, test, and deploy on a consistent platform, shortening timelines, increasing throughput, and eliminating barriers between desktop performance and datacenter power.
APEXX T4 PRO-X GPU Options: Dual or Quad Blackwell GPUs
The APEXX T4 PRO-X is engineered to scale with the most demanding GPU configurations available, offering either dual RTX PRO 6000 Blackwell GPUs or up to four RTX PRO 6000 Blackwell Max-Q GPUs. These options give professionals the flexibility to balance raw performance, workload requirements, and future scalability within a single system.
Dual RTX PRO 6000 Blackwell GPUs
- Ideal for balanced performance across AI, rendering, and visualization.
- Delivers unprecedented speed for creative workflows, from real-time ray tracing to advanced simulation.
- Provides the power needed for multi-application workflows without compromise.
Quad RTX PRO 6000 Blackwell Max-Q GPUs
- Maximum parallel processing power in a single workstation.
- Perfect for large-scale AI training, inference, and generative design workloads.
- Future-proof expandability to handle growing datasets, models, and project demands.
With these GPU options, the APEXX T4 PRO-X enables professionals to configure the right balance of power for their workflows while maintaining the ability to scale as workloads evolve.
CPU Leadership: AMD Ryzen™ Threadripper™ PRO 9000 WX-Series
At the heart of the APEXX T4 PRO-X is the AMD Ryzen™ Threadripper™ PRO 9000 WX-Series, delivering unmatched compute power for the most demanding professional workloads. With up to 96 cores and 192 threads, it’s purpose-built for CPU-intensive tasks like simulation, massive datasets, and advanced multitasking. Combined with 128 PCIe Gen5 lanes, the platform eliminates bottlenecks by fully supporting GPU expansion, storage arrays, and high-speed networking cards, ensuring every component operates at maximum throughput.
CPU advantages include:
- Up to 96 cores / 192 threads for parallel performance.
- Handles massive datasets, simulation workloads, and complex multitasking.
- 128 PCIe Gen5 lanes for full GPU, storage, and network scalability.
- Engineered to eliminate bottlenecks across high-performance workflows.
Optimized Performance for AI Research, Content Creation, and Simulation
The APEXX T4 PRO-X is engineered to accelerate workloads that push hardware to its limits. From AI research to media production and engineering simulation, it delivers consistent, application-optimized performance.
| Workflow | Capability | Benefit |
|---|---|---|
| AI Research & Development | Multi-GPU acceleration | Full-scale local training of large models without external clusters |
| On-premise data processing | Protects sensitive datasets and reduces reliance on cloud services | |
| PCIe Gen5 + high-bandwidth memory | Eliminates bottlenecks during model training and inference | |
| Large-scale AI prototyping | Enables development of LLMs, generative AI, and agentic AI in a single workstation | |
| Media & Entertainment | Multi-GPU rendering | Cuts frame render times in half, accelerating delivery schedules |
| Real-time ray tracing | Smooth playback of complex, high-resolution 3D scenes | |
| NVIDIA Omniverse integration | Real-time collaboration with synchronized assets across creative tools | |
| 8K + virtual production workloads | Handles high-resolution editing, compositing, and stage rendering without slowdown | |
| Engineering & Simulation | 96-core CPU + quad GPU acceleration | Speeds up physics-based simulations and generative design |
| Large dataset handling | Scales to manage terabyte-sized datasets in CFD, FEA, and engineering workloads | |
| Parallel CPU/GPU execution | Reduces simulation cycles, enabling more design iterations in less time | |
| Stability for long-running tasks | Reliable for thermal analysis, crash simulations, and molecular modeling |
Scalable Engineering & Flexibility
The APEXX T4 PRO-X is designed to grow with evolving workloads, giving professionals both expansion and deployment flexibility.
Flexibility highlights:
- 12-slot PCIe card bay allows full GPU population alongside high-speed networking and NVMe storage expansion.
- Rackmount or tower deployment enables use in both studio environments and enterprise datacenters.
- Optimized cooling design maintains performance in dense configurations without thermal throttling.
- Quiet operation ensures minimal disruption in production or creative environments.
- Upgrade-ready architecture extends system lifecycle by supporting future GPU and PCIe technologies.
The BOXX Advantage
Built for high-performance computing at scale, APEXX T4 PRO-X reflects the core principles behind every BOXX system: application-tuned performance, long-term reliability, and upgrade-ready architecture.
Performance Purpose-Built for Your Applications
From Autodesk and Unreal Engine to Omniverse and beyond, BOXX systems are purpose-built for the software professionals rely on. GPU configurations are validated for stability and performance across real-world production environments.
Assembled in the US
Every BOXX workstation is built in the U.S. and supported by a team of experts who understand the demands of professional workflows. No off-the-shelf builds. No outsourced support.
Scalable by Design
With upgrade-ready configurations and flexible deployment options, APEXX T4 PRO-X is engineered to grow with your workload, without compromising performance or reliability.
Configure your APEXX T4 PRO-X today
If you're scaling AI development, accelerating simulation, or cutting render times, BOXX high performance workstations are built to meet your workload head-on. Contact our team for expert configuration guidance and see what purpose-built performance can do for you.