NVIDIA has announced the release of a wide range of hardware and software products powered by generative AI. The new products are aimed to unlock the full potential of generative AI on Windows 11 devices. The newly launched NVIDIA RTX 40 series GPUs are capable of running a wide range of applications with peak performance. The integrated Tensor Cores significantly boost the AI performance across demanding applications.
The GeForce RTX 4080 SUPER is capable of generating AI video over 1.5X faster and images over 1.7X faster when compared with the GeForce RTX 3080 Ti. The Tensor Cores on SUPER GPUs deliver up to 836 trillion AI operations per second. This technique brings transformative AI capabilities for gaming and also to enhance productivity.
The company announced that Acer, ASUS, HP, Dell, Lenovo, MSI, Razer, and Samsung will start shipping new laptops powered by RTX 40 series graphics by the end of January. You will be able to experience generative AI capabilities out of the box.
NVIDIA RTX desktops and mobile workstations that are powered by the NVIDIA Ada Lovelace architecture also deliver the performance that are needed to meet the growing challenges of enterprise workflows. The RTX GPU-powered workstations are capable of delivering NVIDIA AI Enterprise software such as TensorRT and NVIDIA RAPIDS.
This will provide simplified, secure generative AI and data science development. NVIDIA has bundled a 3-year license for NVIDIA AI Enterprise along with RTX A800 40 GB graphics processor. This provides a workstation development platform for AI and data science.
NVIDIA also released AI Workbench, which is a unified toolkit that enables developers to create, test, and customize trained generative AI models and LLMs. The AI tool enables developers to collaborate flexibly and migrate projects to any GPU-enabled environment. Moreover, the system also provides streamlined access to GitHub.
The AI models can be optimized to leverage the full advantage of Tensor Cores on RTX GPUs via NVIDIA TensorRT. The company extended Tensor RT to text-based apps with the help of TensorRT-LLM. It’s an open-source library to accelerate large language models.
The latest update to TensorRT-LLM adds Phi-2, which is capable of running up to five times faster when compared with other inference backends. The PC developers can now deliver generative AI apps on top of 500 AI-powered PC games and apps that are accelerated by RTX GPUs.
NVIDIA and its partners have released new generative AO-enabled apps and services. This includes NVIDIA RTX Remix, which is a platform for the creation of RTX classic game remasters. NVIDIA ACE microservices will also be released to include generative AI speech and animation models. This will enable developers to add intelligent, dynamic digital avatars to games. You will be able to establish connectivity with PC LLMs via Retrieval Augmented Generation (RAG). The system is accelerated by TensorRT-LLM, which enables users to interact with their notes, documents, and content.
