How the NVIDIA RTX 4000 SFF Ada Generation GPU meets design and engineering needs.
PNY has submitted this article.
The demand for smaller footprint workstations has been skyrocketing, fueled by an array of factors. Advancements in cutting-edge hardware, the extreme miniaturization of components, an increase in remote and mobile workforces, and the need for more efficient use of space in modern offices and data centers requires organizations to consider the size of their workstations. The demand for compact, high-performance computing solutions has never been higher. In addition, rising energy efficiency mandates and global cost pressures mean organizations must incorporate workstations that are both energy efficient and cost effective.
Important features to consider when selecting small form factor workstations
Organizations are now adding small form factor (SFF) workstations into data centers and offices to meet space and energy efficiency requirements. Data-intensive, multi-application workflows are now common, and engineering staff frequently perform artificial intelligence (AI), computer-aided design (CAD) and computer aided engineering (CAE) simulations which require modern graphic processing units (GPUs) to provide the necessary computing power and performance to support the implementation of advanced technologies.
Organizations must consider their unique needs when selecting SFF workstations. Let’s explore some areas where modern GPU technology, such as the NVIDIA RTX™ 4000 SFF Ada Generation technology, can meet the size, energy efficiency, and performance needs of modern organizations.
Performance
Built on the revolutionary NVIDIA Ada Lovelace GPU architecture, the RTX 4000 SFF brings new levels of performance, features and capabilities to small form factor workstations. Delivering 2x Gen-1 single-precision (FP32) performance, the RTX 4000 SFF enables two FP32 data paths, doubling the peak FP32 operations and making it ideal for interactive CAE applications that can bring design validation into the beginning of the engineering process, not the end of it. Enhanced Tensor Cores accelerate the latest AI-enhanced applications and work in conjunction with updated RT Cores to deliver significant improvements in graphics and rendering with NVIDIA’s DLSS 3.
Ray tracing
Ray tracing is a method of graphics rendering that simulates the physical behavior of light and adds realism to graphic renders. The RTX 4000 SFF provides up to 2x improved ray-triangle intersection throughput compared to the previous generation for strong ray tracing performance, with full support for leading APIs like NVIDIA OptiX, Microsoft DXR and NVIDIA Vulkan ray tracing, through 3rd generation NVIDIA RT Cores. A single RTX 4000 SFF board can render complex assemblies or designs with physically accurate shadows, reflections and refractions to empower users with instant insight, directly from CAD files. Systems based on the RTX 4000 SFF can power interactive design workflows to provide immediate feedback for unprecedented levels of productivity. These enhanced RT Cores combined with new Shader Execution Reordering (SER) technology dynamically reorder inefficient workloads, dramatically improving shader performance to accelerate end-to-end ray-traced image rendering performance.
Faster throughput
Adequate throughput is required as more CAD, CAE, and other engineering applications add AI-enabled or enhanced features. In the RTX 4000 SFF, 4th generation NVIDIA Tensor Cores support the Fine Grained Structured Sparsity feature that delivers more than 2x throughput for tensor matrix operations compared to the previous generation. A new FP8 precision mode is also available. Finally, independent floating-point and integer data paths allow more efficient execution of workloads using a mix of computation and addressing calculations.
High quality graphic frames
Videos and games require high performance to effectively render high quality graphic images. The RTX 4000 SFF supports NVIDIA’s Deep Learning Super Sampling 3 technology (DLSS 3). The Ada Lovelace GPU architecture features a new Optical Flow Accelerator and AI-based DLSS Super Resolution with DL denoiser that boosts DLSS 3.0’s frame rates up to 4 compared to the previous GPU architecture while maintaining or exceeding native image quality.
Augmented reality / virtual reality
Augmented reality (AR) and virtual reality (VR) require high quality graphics for the user. The RTX 4000 SFF has full AR/VR support for the latest high-resolution head-mounted display (HMD) devices. It also provides high-performance graphics, and 20 GB of GPU memory to enable the creation of immersive augmented reality (AR) and virtual reality (VR) experiences for photorealistic design reviews or ergonomic optimization studies.
Data intensive engineering and design tasks
Organizations need GPUs with high bandwidth capabilities to support CAD, CAE and other simulations. The RTX 4000 SFF includes PCIe Gen 4 which provides twice the bandwidth of PCIe Gen 3, improving data-transfer speeds from CPU memory for data-intensive engineering and product design tasks. Built with 20GB of higher speed GDDR6 memory, RTX 4000 SFF provides an ideal memory footprint to address complex designs or assemblies in latency-sensitive pro engineering and product applications.
Support for remote work
The NVIDIA Encode (NVENC) engine, now in its 8th generation, provides better encoding quality than software-based x24 encoders when engineers or designers need to share their work with remote colleagues. The card also incorporates AV1 video encoding, which is 40 percent more efficient than H.264 encoding for 4K HDR video, essentially providing better quality at the same bitrate bandwidth. NVDEC, now in its 5th generation, supports hardware-accelerated decoding for MPEG-2, VC-1, H.264 (AVCHD), H.265 (HVEC), VP8, VP9 and AV1. This enables interactive collaboration with remote colleagues across environments ranging from Zoom or Teams-based video conferencing, which can also use a variety of AI-powered features using the RTX-powered NVIDIA Broadcast app. In addition, other performance enhancements of the NVIDIA Ada Lovelace architecture can deliver next-generation collaborative 3D workflows across design, creative, and engineering teams with NVIDIA Omniverse™ Enterprise.
Multi-application or collaborative engineering
PNY provides four mDP to DP adapters with the RTX 4000 SFF to support up to four 5K monitors at 60Hz, or dual 8K displays at 60Hz per card. The RTX 4000 SFF supports HDR color for 4K at 60Hz for 10/12b HEVC decode and up to 4K at 60Hz for 10b HEVC encode. Each mini-DisplayPort connector can drive ultra-high resolutions of 4096 x 2160 at 120Hz with 30-bit color, giving engineering and design professionals the screen real estate required for even the most demanding multi-application engineering or product design contexts, or to provide interactive windows into collaborative engineering or design workflows powered by NVIDIA Omniverse.
Energy efficiency
With rising energy efficiency mandates and global cost pressures, the RTX 4000 SFF provides energy efficiency at an unprecedentedly low 70 W of TGP (Total Graphics Power) while housed in a small footprint to meet modern computing needs.
GPU performance comparison
Organizations need to consider GPU performance when evaluating products. The sample below shows an example of a 2.1x performance improvement when comparing graphics run on the NVIDIA RTX 4000 SFF as compared to the NVIDIA RTX A2000 12GB.
Meeting space and energy needs
Organizations need to fit workstations in smaller areas while improving system performance and reducing energy usage and costs. Engineers and design professionals increasingly use CAD, CAE, and AI applications with visually compute intensive workloads requiring next-generation workstations. The NVIDIA RTX 4000 SFF GPU provides a small footprint and low energy usage while meeting the requirements for today’s professional workflows, including unprecedented AI/ML and 3D design and simulation processing performance.
PNY provides support for a wide range of NVIDIA GPU and NVIDIA networking products. For more information on how PNY can help in choosing the right NVIDIA RTX GPU for CAD and CAE engineering workstations, visit PNY.
* Intel i9-12900K | 64GB DRAM | Windows 11 Enterprise x64 | NVIDIA driver 526.99 | Graphics test results based on SPECviewperf 2020 4K energy subtest scores