Everything engineers need to know about the latest graphics boards for product development, AI and creative design.
PNY has submitted this article.
NVIDIA introduced three new Ada Lovelace architecture-based professional graphics boards at the recently concluded SIGGRAPH 2023 conference: the NVIDIA RTX 5000 Ada Generation, RTX 4500 Ada Generation and the full-height RTX 4000 Ada Generation. All the new Ada products offer enhanced features to enable next-level fluidity and creative iteration for manufacturing and 3D product design.
NVIDIA RTX 5000 Ada Generation
The NVIDIA RTX 5000 Ada Generation, with Gen 3 RTX technology, helps unlock breakthroughs in generative AI that are revolutionizing productivity and providing creative possibilities.
NVIDIA RTX 5000 combines 100 Gen 3 RT Cores, 400 Gen 4 Tensor Cores and 12,800 CUDA cores with 32GB of graphics memory to accelerate rendering, AI, graphics and compute workloads. RTX 5000-powered workstations can help engineers, designers and CAE professionals to achieve success in today’s demanding manufacturing business landscape.
NVIDIA RTX 4500 Ada Generation
Businesses are using GPU-based simulation, HPC and AI to take on complex challenges and open up new possibilities. Generative AI transforms how designers, engineers and CAE specialists develop and manufacture compelling and innovative products.
With 60 Gen 3 RT Cores, 240 Gen 4 Tensor Cores and 7,680 CUDA cores, plus 24GB of graphics memory, the NVIDIA RTX 4500 Ada Generation enables fast ray tracing and AI-powered graphics. Whether you’re a professional designer or just starting out, the RTX 4500 offers the power and flexibility to explore new opportunities for creating the next generation of products.
NVIDIA RTX 4000 Ada Generation
Designers, engineers and CAE professionals require accelerated computing and AI to tackle increasingly complex workloads, time-to-market imperatives and increasing demands for product innovation and quality from consumers. Generative AI is changing entire segments of our economy, with manufacturing playing a leading role as an innovator and trendsetter.
NVIDIA RTX 4000 Ada Generation is a powerful single-slot GPU for professionals, delivering acceleration for real-time rendering, AI, graphics and compute workloads to your desktop. It combines 48 Gen 3 RT Cores, 192 Gen 4 Tensor Cores, and 6,144 CUDA cores with 20GB of graphics memory to effortlessly handle the large datasets and complex visual workloads that exemplify modern manufacturing.
Shared features enhance productivity and creativity
All the new Ada Generation products offer improved features to support manufacturing and 3D product design. The Ada family of products offers engineering teams the ability to extend the power of engineering simulation throughout the design process with accelerated real-time structural, thermal and fluid flow analysis software tools. Engineers can evaluate the impact of design changes instantly to iterate at the speed of light and revolutionize the performance of their designs. The NVIDIA Ada Lovelace CUDA cores bring up to 2x the single-precision floating-point throughput and 2x improved power efficiency compared to previous generations, unlocking new potential for designers and engineers.
RT and Tensor Cores
Gen 3 RT Cores in Ada Generation GPUs provide up to 2x the throughput of the first generation and enable concurrent ray tracing and shading. Design reviews can be accelerated to review high-fidelity visualizations with physically accurate materials, lighting and reflections in real-time. This improved performance unlocks the ability to evaluate new product concepts rapidly and create stunning marketing content directly from CAD geometry.
Gen 4 Tensor Cores provide up to 5x faster AI training performance than the previous generation with FP16 precision. Support for the new FP8 data format for inference provides more than 5x faster performance than the previous generation and reduces data memory usage. When ISVs implement structural sparsity an additional 2x speedup can be obtained.
RTX Ada Generation boards
The new RTX Ada Generation boards feature two video encode engines and two video decode engines, including support for the AV1 video format and the performance required for multi-stream video applications for security and video serving, both of which are desirable in manufacturing use cases.
PCIe Gen4
PCIe Gen4 doubles bandwidth to 31.5 GB/sec for x16 connections, improving data transfer speeds from CPU memory. This is especially helpful for data-intensive tasks such as creating 3D models from large datasets, or working with AI-enhanced applications. Faster PCIe performance also accelerates GPU DMA transfers, resulting in faster IO with GPUDirect Storage.
Increased GPU memory capacity
NVIDIA RTX 5000, 4500 and 4000 Ada Generation graphic boards provide increased GPU memory capacity with ECC (which can be enabled or disabled in software) over their predecessors, providing the memory needed for rendering directly from CD files, engineering simulation and other GPU memory-intensive applications. Support for ECC memory delivers uncompromised computing accuracy and reliability for mission-critical engineering and design scenarios.
Other enhancing features
Ada Generation products also supply four DisplayPort 1.4a connectors, support for multiple 8K monitors, NVIDIA Quadro Sync, Mosaic, and Warp and Blend which enables massive immersive environments for CAVE, video walls and other environments suitable for design reviews and approvals.
Ada Generation boards provide support for the latest high-resolution HMD AR/VR/XR devices, high-performance graphics and generous GPU memory capacity to enable the creation of AR and VR experiences for factory floor training, design reviews, ergonomic studies and more.
The Optical Flow Accelerator and AI-based DLSS Super Resolution with DL denoiser give DLSS 3 a huge boost in frame rates—up to 4x faster than before—without sacrificing image quality.
Other announcements of note to manufacturing and product design
Alongside these new professional graphics boards, NVIDIA announced a major update for NVIDIA Omniverse Enterprise , enabling manufacturers to reinvent collaborative workflows and implement digital twins of entire factories and associated supply chains.
NVIDIA OVX systems and the NVIDIA L40S data center GPU synergistically work with the new workstation GPUs, or provide virtual GPU (vGPU) support, to support Universal Scene Description (OpenUSD). This solution will further simplify manufacturing workloads and enhance the ability of disparate professional applications to seamlessly share critical CAD and other files without tedious and error-prone wading and waiting as applications export and import data.
Availability
The NVIDIA RTX 5000 is available now from PNY’s ecosystem partners. NVIDIA RTX 4500, RTX 4000 and L40S will be available in early fall. To learn more, visit PNY.com.
