NVIDIA's 8 new GPUs are built to handle professional creative workloads

NVIDIA's new GPUs can handle a wide range of creative workloads.

What you need to know

• NVIDIA announced eight new GPUs for desktops, laptops, and servers.
• The new GPUs for desktops, laptops, and workstations are built to handle demanding creative workloads.
• NVIDIA also announced NVIDIA Omniverse Enterprise, a platform for real-time collaboration on 3-D projects.

NVIDIA founder and CEO Jensen Huang announced several new GPUs at his GTC 2021 keynote. NVIDIA announced eight NVIDIA Ampere architecture GPUs for next-gen laptops, desktops, and servers, NVIDIA Omniverse Enterprise, and a new GPU-accelerated software development kit for NVIDIA Maxine, the streaming platform behind NVIDIA Broadcast.

The NVIDIA RTX A5000 and NVIDIA RTX A4000 are desktop GPUs built for professionals. They feature new RT Cores, Tensor Cores, and CUDA cores to help power AI, graphics, and real-time rendering. NVIDIA claims that these new GPUs can handle those task two times faster than previous generations. HP already announced the Z2 Small Form Factor G8 and Z2 Tower G8 that can be powered by these new GPUs.

People with thin and light devices can take a look at the new NVIDIA RTX A2000, RTX A3000, RTX A4000, and RTX A5000 laptop GPUs. These GPUs utilize the latest generations of Max-Q and RTX technologies and work with the NVIDIA Studio ecosystem.

On the data center side of things, NVIDIA announced the NVIDIA A10 GPU and A16 GPU.

NVIDIA outlines some of the benefits that each of the new RTX GPUs have because they're built on the latest NVIDIA Ampere architecture:

• Second-Generation RT Cores: Up to 2x the throughput of the previous generation, with the ability to run concurrent ray tracing, shading and denoising tasks.
• Third-generation Tensor Cores: Up to 2x the throughput of the previous generation, up to 10x with sparsity, with support for new TF32 and BFloat16 data formats.
• CUDA Cores: Up to 2.5x the FP32 throughput of the previous generation for significant increases in graphics and compute workloads.

The desktop GPUs have up to 24GB of GPU memory, which is double the memory of the previous generation. The RTX A4000 has 16GB of GDDR6 memory, while the RTX A5000 has up to 24GB of GDDR6 memory. The RTX A5000 can expand up to 48GB of memory by using NVIDIA NVLink to connect to two GPUs.

The new Desktop RTX GPUs will be available from OEMs and NVIDIA partners later this month. The RTX laptop GPUs should be available in mobile workstations in Q2 2021 from OEMs.

NVIDIA Omniverse Enterprise allows creators to collaborate in real-time in a shared virtual space. It supports 3D design work, which is often challenging to collaborate on in real-time due to the size of files used.

NVIDIA Omniverse Enterprise also includes NVIDIA Omniverse Create, which allows accelerated scene composition in real-time, and NVIDIA Omniverse View, which can be used for collaborative design and visualization of architectural and engineering projects, as well as photorealistic rendering. NVIDIA RTX Virtual Workstation is also part of the NVIDIA Omniverse Create platform. It allows creators to run graphics-intensive 3-D applications from anywhere.

Developers also now have a new SDK for NVIDIA Maxine, the cloud-AI video streaming platform that powers NVIDIA Broadcast. This SDK allows developers to create real-time video-based experiences.