WHY NVIDIA RTX PRO SERVERS ARE THE BEST CHOICE FOR VISUALISATION & AI

RTX PRO Server is NVIDIA’s latest platform of GPU-accelerated servers. These rackmount servers are the perfect platform for the new generation of Blackwell GPUs, providing best-in-class performance in visualisation and AI workloads. In this article we’ll take a closer look at RTX PRO Servers alongside alternative systems, so you can pick the best solution for your business.

NVIDIA RTX PRO Server Architecture

NVIDIA RTX PRO Servers are based around the powerful combination of NVIDIA Blackwell GPUs, x86 CPUs (either AMD or Intel) and NVIDIA ConnectX-7 SmartNICs. The latter provide rapid network access via Ethernet or InfiniBand. Coming soon, we’ll also be offering systems with embedded ConnectX-8 SuperNICs, which act as both SmartNICs and PCIe switches, boosting performance even further.

8-GPU NVIDIA RTX PRO Server system configuration based on the Intel Xeon 6 platform

As of September 2025, there is only one GPU available, the mighty NVIDIA RTX PRO 6000 Blackwell Server. However, RTX PRO Servers can be customised to your requirements, being scalable from a single GPU all the way up to eight GPUs. The table below summaries the massive improvement in performance and capabilities of the RTX PRO 6000 Blackwell Server versus older GPUs, highlights being over a 12x increase in AI performance and doubling of model size, and close to 4x speed-up in ray tracing performance.

GPU	NVIDIA RTX PRO 6000 Blackwell Server (2025)	NVIDIA L40S (2022)	NVIDIA A40 (2020)
ARCHITECTURE	Blackwell	Ada Lovelace	Ampere
CUDA CORES	24,064	18,176	10,752
TENSOR CORES	752 5th gen	568 4th gen	336 3rd gen
RT CORES	188 4th gen	142 3rd gen	84 2nd gen
MEMORY	96GB GDDR7	48GB GDDR6	48GB GDDR6
ECC MEMORY	✓	✓	✓
MEMORY CONTROLLER	512-bit	384-bit	384-bit
MIG INSTANCES	4	✗	✗
AI PERFORMANCE (FP4)	4 PFLOPS	1.4 PFLOPS	0.3 PFLOPS
MAXIMUM AI MODEL SIZE	153 billion	76 billion	76 billion
RAYTRACING PERFORMANCE	355 TFLOPS	209 TFLOPS	73 TFLOPS

Another key area of improvement is added support for MIG. Multi-Instance GPU fully isolates at the hardware level allowing memory, cache and cores to be partitioned into as many as four independent instances, giving multiple users access to GPU acceleration.

Why choose an NVIDIA RTX PRO Server from Scan?

NVIDIA chose Scan as one of its partners for RTX PRO Servers because of our deep expertise in building GPU-accelerated systems. Our in-house 3XS Systems division has been handcrafting PCs, workstations and servers for more than 30 years. We pioneered the AI dev box back in 2016, so we have a huge amount of experience in building highly reliable systems that deliver the most performance for your budget. Here are the key reasons to buy an RTX Pro Server from Scan.

NVIDIA RTX PRO Server Software Stack

NVIDIA RTX PRO Servers are a lot more than a hardware platform, they also provide the perfect gateway to NVIDIA’s optimised software stack for visualisation and AI workloads.

For visualisation and rendering workloads, RTX PRO Servers deploy technologies such as NVIDIA RTX and DLSS 4 to accelerate workloads including AEC & BIM, CAD & CAE, Media & Entertainment. Experience the very best in real-time ray tracing and AI-powered upscaling and frame generation thanks to NVIDIA’s drivers which are ISV-certified for hundreds of DCC applications.

For digital twins, RTX PRO Servers employ software platforms such as NVIDIA Omniverse to build photorealistic simulations of real environments with accurate physics-informed characteristics. These are ideal for designing retail stores and manufacturing facilities, and training robots.

For AI workloads, RTX PRO Servers use pre-configured NVIDIA frameworks and software platforms such as NVIDIA AI Enterprise to train and inference AI models. However, as powerful as NVIDIA RTX PRO Servers are, for the largest scale AI models used in projects such as generative AI, agentic AI, physical AI, and AI factories we recommend NVIDIA DGX appliances.

Multipurpose & Flexible GPU Servers

Although NVIDIA DGX and HGX servers are the ideal platform for the most demanding AI projects, RTX PRO Servers do have an ace up their sleeves that perhaps isn’t immediately obvious. This is that they are equally suited to running both visualisation and AI workloads.

This is because unlike DGX and HGX systems, RTX PRO Servers include dedicated RT cores that accelerate raytracing. In addition, they support both Linux and Windows Server operating systems, the latter being a crucial requirement for many visualisation applications. This means you could switch from one operating system to another between projects, or for maximum flexibility run multiple virtual machines at once, some running Linux and others Windows Server.

This is a big benefit for businesses because it means that rather than investing in dedicated AI and visualisation servers you can switch between the two workloads at will with an RTX PRO Server.

NVIDIA RTX PRO Server Configuration

As already mentioned NVIDIA RTX PRO Servers from Scan 3XS Systems are available with up to eight NVIDIA RTX PRO 6000 Blackwell Server GPUs. As of September 2025, there are three main configurations available.

Firstly, a 2U chassis with up to four GPUs powered by a single Intel Xeon 6 CPU with up to 80-cores, and up to 2TB of RAM.

Secondly a 4U chassis with up to eight GPUs powered by a pair of Intel Xeon 6 CPUs with up to 160-cores, and up to 4TB of RAM.

Thirdly, a 4U chassis with up to eight GPUs powered by a pair of AMD EPYC 9005 CPUs with up to 256-cores, and up to 3TB of RAM.

Conclusion

NVIDIA RTX PRO Servers are the perfect choice for visualisation and AI workloads thanks to their powerful combination of NVIDIA Blackwell GPUs, NVIDIA ConnectX networking, NVIDIAs unmatched software stack plus the latest AMD and Intel x86 CPUs.

Whether you’re rendering, filmmaking, building digital twins, or training and inferencing cutting-edge AI models NVIDIA RTX PRO Servers deliver unparalleled power and flexibility.