NVIDIA and global workstation vendors showcase exciting systems for generative AI, LLM development and data science

News from SIGGRAPH 2023: NVIDIA and leading manufacturers showcase powerful RTX workstations for the era of AI and digital creativity.These powerful computers are ideal companions for creating AI applications and digital artwork. BOXX, Dell Technologies, HP and Lenovo showcased their high-tech systems here, based on NVIDIA RTX 6000 Ada Generation GPUs and supported by NVIDIA AI Enterprise and NVIDIA Omniverse Enterprise software. But that’s not all, as NVIDIA also unveiled three new Ada Generation GPUs for desktop workstations: the NVIDIA RTX 5000, RTX 4500 and RTX 4000. These GPUs give professionals worldwide access to the latest AI features, graphics technologies and real-time rendering capabilities.

“Few workloads are as challenging as generative AI and digitization applications, which require a full-stack approach to computing,” said Bob Pette, vice president of professional visualization at NVIDIA. “Professionals can now tackle these on a desktop with the latest NVIDIA-powered RTX workstations, enabling them to build vast, digitized worlds in the new age of generative AI.”

The new RTX workstations pack some serious power – they can be equipped with up to four NVIDIA RTX 6000 Ada GPUs, each offering 48 GB of memory. Just think: a single desktop workstation can offer up to 5,828 TFLOPS of AI performance and 192 GB of GPU memory. These models are also true all-rounders. You can use them with NVIDIA AI Enterprise or Omniverse Enterprise, depending on your preference. This means they can handle complex generative AI experiments as well as graphics-intensive tasks with ease. It’s being reported that NVIDIA took a big step today and released NVIDIA AI Enterprise 4.0 – there’s even said to be this thing called NVIDIA NeMo™ included.

According to reports, it’s sort of an all-around worry-free framework that can supposedly be used to conjure up and customize basic models for generative AI. There is also talk of the much-loved NVIDIA RAPIDS™ libraries for data science and a wide range of frameworks, pre-trained models and tools being included. A package for potential AI actions in the business domain, from possible recommendation programs to eventual virtual assistants and potential cybersecurity.

So, too, is Omniverse Enterprise – a potential “superweapon” for an eventual digital revolution in industry. If it’s anything to go by, teams can use this platform to develop 3D workflows in perfect harmony and even leverage OpenUSD applications. As a platform designed to inherently understand OpenUSD, teams from around the world can work on realistic 3D data from a variety of 3D applications. As Jason Schnitzer, the CTO at Yurts, has expressed, “If it comes to fruition, we at Yurts would have a cool generative AI solution on our hands that could adapt to different shapes, potential deployment scenarios, and budgets of our customers. We achieved this by potentially using LLMs for various cool and natural language tricks and integrating RTX 6000 Ada. Whether huge data centers or pocket-sized workstations, Yurts stays true to our platform and offerings alongside NVIDIA for possible scalability.”

So, the brand new NVIDIA AI Workbench is also waiting in the wings, and it could be off to Early Access soon. Then, this simple toolkit could show developers of any experience level how they could create, customize and run generative AI models on PCs or workstations. Then they could smoothly scale it up to data centers, public clouds or NVIDIA DGX clouds. This means that potential AI actions are pre-programmed!

Next-Generation RTX Technology

The new NVIDIA RTX 5000, RTX 4500, and RTX 4000 desktop GPUs feature the latest NVIDIA Ada Lovelace architecture technologies, including:

• NVIDIA CUDA® cores: up to two times higher single-precision floating-point throughput compared to the previous generation.
• Third-generation RT cores: Up to 2x throughput over previous generation with the ability to simultaneously run ray tracing with shading or denoising functions.
• Fourth generation Tensor cores: Up to 2x faster AI training performance than previous generation with enhanced support for FP8 data format.
• DLSS 3: New levels of realism and interactivity for real-time graphics with AI performance.
• Larger GPU Memory: RTX 4000 offers 20GB of GDDR6 memory, RTX 4500 offers 24GB of GDDR6 memory, and RTX 5000 has 32GB of GDDR6 memory – all support error code correction for error-free computing with large 3D models, rendered images, simulations, and AI datasets.
• Advanced Reality Capabilities: Support for high-resolution augmented reality and virtual reality devices to deliver the high-performance graphics needed to create immersive AR, VR, and mixed reality content.

Availability

RTX workstations with up to four RTX 6000 Ada GPUs, NVIDIA AI Enterprise and NVIDIA Omniverse Enterprise will be available from system builders in the fall. The new NVIDIA RTX 5000 GPU is available now from HP and through global channel partners including Leadtek, PNY and Ryoyo Electro. The NVIDIA RTX 4500 and RTX 4000 GPUs will be available in the fall from BOXX, Dell Technologies, HP and Lenovo, and through global channel partners.

Source: NVIDIA