Nvidia Quadro

Nvidia Quadro is technology of streaming graphic or a HPC-applications from DPCs.

2020: Representation of Nvidia Quadro Experience

On May 28, 2020 NVIDIA submitted the application realizing the potential of the professional graphic processors Quadro, supporting capture of the screen in 4K, exchange of video, access to drivers and many other things.

The application for GPU Quadro - NVIDIA Quadro Experience — allows professionals to lift in different areas to absolutely other level both creative process, and productivity. Quadro Experience helps professionals to simplify expensive tasks on time, to accelerate workflow and to provide timely updating of the necessary applications.

NVIDIA Quadro Experience simplifies exchange of content: in an application interface it is possible to do screenshots or to write a desktop in 4K. Now commands will be able easily to load content and even to broadcast it directly from the PC or the notebook.

Functions of creation of screenshots and record of video of a desktop are available from any application, and users can load resources from the repository set by them.

The Instant Replay function automatically writes the last 20 minutes of activity of a desktop, allowing users to review the activities and to take pictures of the necessary moments.

Thanks to compatibility with the NVIDIA SHIELD mediaplayer several users can strimit content directly from a desktop at once and to hold the presentations, trainings and meetings.

Quadro Experience will take care of regular performance improvement, stability of work and will inform on emergence of opportunities of your system. The application will automatically inform users on emergence of relevant drivers that they could begin to use the appeared opportunities as soon as they become available.

The description of drivers becomes available in Quadro Experience at once at an exit of the latests version that users could set immediately the fresh version with all updates, optimization and certification of all necessary applications.

Quadro Experience is not only about work. It is possible to switch GPU Quadro to the game mode (Game Mode) and to involve the game potential of the system.

In the application the opportunities similar to GeForce Experience, such as Ansel, Freestyle and Highlights are available. Users can turn on optimal game adjustments (Optimal Playable Settings) for the maximum image quality and frame frequency and strimit a gameplay on NVIDIA SHIELD.

NVIDIA Quadro Experience helps professionals to execute projects more effectively. The Alstom company, the developer of the fast train TGV became one of beta users of the application.

The image is provided to Alstom.

Project analysis and customization are important components of workflow of Alstom — the company often applies Autodesk VRED to project analysis and rendering and Dassault Systemes CATIA to technical expertize. GPU Quadro RTX and Quadro Experience helped Alstom to accelerate work and to direct power of graphic processors to accomplishment of graphic tasks.

To share the graphic content created on GPU Quadro the possibilities of video capture providing high quality of the image are necessary. NVIDIA Quadro Experience allows to write content of professional graphic applications that gives us the chance to improve possibilities of virtual reality of our clients and our internal workflows, speaks Fabien Normand, the expert in virtual reality in Alstom.

Professional work becomes more interactive and gives more opportunities for interaction. Quadro Experience allocates with powerful opportunities of graphics of Quadro of engineers, designers, artists, architects and other professionals.

2018

NVIDIA Quadro RTX 8000, Quadro RTX 6000 и Quadro RTX 5000

On August 13, 2018 the NVIDIA company provided graphic processors based on architecture of Turing - NVIDIA Quadro RTX 8000, Quadro RTX 6000 and Quadro RTX 5000 which provide to professionals hardware accelerated ray tracing, AI, possibilities of shadowing and simulations. Also the company provided Quadro RTX Server – reference architecture for virtual workstations and the systems of rendering with flexibly reconfigured

Graphic processor NVIDIA Quadro RTX

According to the developer, the graphic processors Quadro RTX are intended for the most difficult visual tasks typical for film production and creation of video content, automotive industry, architecture and scientific visualization. They show the high level, thanks to a complex of technologies:

• RT cores provide ray tracing in real time for objects and Wednesdays with physically correct shadows, reflections, refraction of beams and global lighting.
• Tensor cores of Turing reproach training of neuronets and inferens which are used for creation of products and services based on algorithms of artificial intelligence.
• The design of multiprocessors of architecture of Turing locating up to 4608 cores of CUDA provides the speed of calculations to 16 trillion transactions with a floating point in a parallel with 16 trillion integer operations per second for acceleration of difficult simulation of physics of the real world.
• Advanced technologies of programmable shaders increase the speed of work with difficult visual effects and heavy graphic loadings.
• Memory of Samsung 16GB GDDR6 is applied to work with even more difficult projects, massive sets of architectural data, film content in 8K and other data. NVIDIA NVLink allows to integrate two GPU via high-speed interface for expansion of memory to 96 GB and increase in bandwidth to 100GB/page. Hardware support of USB Type-C and VirtualLink, an open industry standard which meets requirements of VR helmets for energy consumption to display and bandwidth and allows to connect them via one connector USB-C.
• Technologies of performance improvement of VR-applications, including Variable Rate Shading, Multi-View Rendering and VRWorks Audio.

Key characteristics of GPU:

Quadro RTX Server defines the standard of rendering on demand in data centers, allowing to configure render-nodes for package and interactive rendering.

The solution includes GPU Quadro RTX with software of Quadro Infinity and arms professionals with powerful and flexible architecture. Quadro Infinity will provide access to one GPU to several users via virtual workstations that increases density of data centers. End users can also reserve render-nodes and workstations according to the needs, noted in NVIDIA.

As the developer noted, with the preset software for creation of content and rendering of Quadro RTX Server represents the powerful and simple solution for rendering in deployment which can be scaled from small installation to large data center and which is 75% cheaper than a visualization farm based on CPU.

The functionality of Quadro RTX through the graphic NVIDIA RTX platform which includes API for ray tracing, AI, rasterizations, simulations and support of the materials NVIDIA MDL and resources of Pixar of OSD is available to developers.

The expected retail prices: Quadro RTX 8000 from 48 GB of memory: $10,000 Quadro RTX 6000 from 24 GB of memory: $6300 Quadro RTX 5000 from 16 GB of memory: $2300

GPU Quadro RTX will be available in the fourth quarter 2018.

The graphic processor Quadro GV100 with Nvidia RTX technology

On March 28, 2018 the Nvidia company announced the graphic processor Nvidia Quadro GV100 with Nvidia RTX technology which will allow designers to apply ray tracing in real time in applications of professional modeling and creation of content.

Developers of content can work on the projects with correctly built lighting and shadows and to execute difficult rendering up to 10 times faster, than on CPU. Designers and architects can create interactive, photorealistic visualization of massive 3D models directly in real time, noted in Nvidia.

Nvidia Quadro GV100 with Nvidia RTX technology. Photo: blogs.nvidia.com

Quadro GV100, with 32GB of the memory expanded to 64GB at connection of several GPU Quadro using Nvidia NVLink technology. Nvidia Volta founded on graphic architecture, the graphic processor GV100 has performance in 7.4 TFLOPS in transactions of double precision, 14.8 TFLOPS in transactions of single precision and 118.5 TFLOPS in transactions of deep learning. The algorithm of elimination of noise using AI of Nvidia OptiX which is built in Nvidia RTX is executed nearly 100 times quicker, than on CPU, providing rendering in real time without noise, the developer specified.

The Nvidia RTX technology is available to developers through API Nvidia OptiX, API Microsoft DirectX Raytracing and, in the future, through Vulkan – the open cross-platform graphic standard. All three API have the general model of shader programming which allows developers to work with different platforms. GV100 and Nvidia RTX technology provide the high speed of rendering of images of cinema quality. Noise elimination speed using AI in OptiX at ray tracing provides complete visual interactivity in all development process. High speed of transactions of double precision in combination with an opportunity to increase memory up to 64GB on NVLink allows to draw large difficult models. Support of VR with a high graphic and calculating capacity means that designers can apply VR platforms to the analysis of projects and viewing photorealistic scenes in complete scale, noted in Nvidia.

GPU Quadro GV100 is available since April, 2018.^[1]

2017: Quadro vDWS as means of virtualization of servers

On August 18, 2017 the Nvidia company announced an opportunity to use Quadro vDWS software as means of virtualization of server resources, creating productive workstations.

Such opportunities of virtualization are provided with installation in servers of the graphic processors NVIDIA Tesla.

The software technology of NVIDIA Quadro Virtual Data Center Workstation (Quadro vDWS) uses resources of data centers. It is available 120 systems from 33 producers.

Quadro vDWS allows to work productively to territorially spaced commands. Software effectively copes with computing loadings which complexity grows exponential with the advent of technologies for 3D, photorealistic rendering, virtual reality and deep learning.

Hardware support of graphic processors of Tesla on architecture of NVIDIA Pascal allows Quadro vDWS software:

• create difficult 3D and photorealistic projects – to 24 GB of memory of GPU for work with large models;
• increase productivity – the doubled graphic performance in comparison with the previous graphic architecture of NVIDIA;
• an opportunity to execute both graphic, and computing tasks – support of the accelerated processing of graphics and computing tasks (CUDA and OpenCL) accelerates design process and computer simulation;
• increase performance for users of Linux - NVIDIA NVENC increases performance and density of users exempting CPU from resource-intensive coding of video H.264 for users of workstations based on Linux.

Corporations change. Artificial intelligence technologies, photorealism, virtual reality are implemented into workflow, thereby improving interaction of employees. The platform of visualization Quadro develops together with the corporate sector to provide necessary performance. Using software of Quadro vDWS installed on servers based on GPU Tesla, the companies will be able to work with even large volumes of data, to start the heaviest applications and to increase the general mobility of work. To Bob Pette, the vice president for solutions for professional visualization in NVIDIA

With growth of complexity of professional applications for engineering and design, mass applications for daily use Windows 10, it seems, Office 365 or YouTube also need acceleration to provide a full range of opportunities for the user in the virtualized digital workplace.

Quadro vDWS technology support more than 100 server systems from vendors of Cisco, Citrix, Dell, HP, IBM, Lenovo, VMware and others. Support of opportunities of Pascal will appear on September 1, 2017.

Notes