Highlights:

  • Nvidia emphasized that the RTX chips boast significant power, capable of accelerating the performance of generative AI models in the Nvidia TensorRT-LLM open-source library.
  • Nvidia disclosed that PC manufacturers, including Acer Inc., ASUSTek Computer Inc., Dell Technologies Inc., HP Inc., Lenovo Group Ltd, Samsung Electronics Ltd., Micro-Star International Co. Ltd, and Razer Inc., are set to unveil new RTX AI laptops either at CES or in the ensuing weeks and months.

Nvidia Corp. made a significant announcement at the 2024 Consumer Electronics Show in Las Vegas, introducing its latest artificial intelligence hardware—the GeForce RTX SUPER desktop graphics processing unit. This technology is touted to empower “supercharged generative AI workloads” specifically tailored for laptop personal computers.

The company affirmed that the newly introduced RTX SUPER GPUs are equipped to amplify generative AI experiences on PCs. These GPUs support Nvidia TensorRT acceleration, catering to the widely used Stable Diffusion XL model for text-to-image workflows. Furthermore, the RTX chips feature support for Nvidia RTX Remix, enhancing generative AI-powered texture tools. Additionally, they integrate with Nvidia ACE microservices, and in video games utilizing Deep Learning Super Sampling technology, they deliver higher-resolution frames from a lower-resolution input.

Nvidia’s announcement highlighted the introduction of the new GeForce RTX 40 SUPER Series graphics cards designed for laptops and workstations. This series encompasses the GeForce RTX 4080 SUPER, 4070 Ti SUPER, and 4070 SUPER models, specifically crafted to deliver top-notch AI performance. The GeForce RTX 4080 SUPER outpaces the GeForce RTX 3080 Ti GPU, demonstrating superior AI video processing at 1.5 times the speed and image generation at 1.7 times the pace. The Tensor Cores within SUPER GPUs achieve an impressive performance of up to 836 trillion operations per second. This capability integrates generative AI functionalities into gaming, content creation, and daily productivity tasks.

The company emphasized that the RTX chips boast significant power, capable of accelerating the performance of generative AI models present in the Nvidia TensorRT-LLM open-source library. In this library, developers can discover pre-optimized versions of numerous widely used large language models for PCs. Among them is Chat with RTX, a new model developed by Nvidia, which was recently released in demo mode this month. This model facilitates AI developers in more accessible and interactive handling of their notes, documents, and other content.

In a statement, Nvidia Chief Executive Jensen Huang emphasized the critical significance of running generative AI locally on PCs. This capability proves indispensable for workloads demanding privacy, low latency, and cost sensitivity. Huang clarified that implementing generative AI on compact devices like PCs or laptops necessitates a substantial base of AI-ready systems and the appropriate developer tools for fine-tuning and optimizing those models.

The CEO added, “With over 100 million RTX AI PCs and workstations, Nvidia is a massive installed base for developers and gamers to enjoy the magic of generative AI.”

Nvidia disclosed that PC manufacturers, including Acer Inc., ASUSTek Computer Inc., Dell Technologies Inc., HP Inc., Lenovo Group Ltd, Samsung Electronics Ltd., Micro-Star International Co. Ltd, and Razer Inc., are set to unveil new RTX AI laptops either at CES or in the ensuing weeks and months. Regarding the RTX chips, they are slated to commence shipping to customers later this month.

AI Software Updates

Nvidia announced that every new A800 40GB Active RTX GPU PC and laptop will include a three-year license for Nvidia AI Enterprise, the company’s AI development software.

In support of developers harnessing the full potential of their latest AI-capable hardware, Nvidia revealed a new AI development toolkit named Nvidia AI Workbench. The toolkit is set to enter beta availability later this month. Nvidia AI Workbench empowers developers to swiftly create, test, and customize generative AI models and LLMs. It facilitates seamless access to models from repositories, including Hugging Face, GitHub, and Nvidia NGC. As highlighted by the company, this capability enables AI development projects to undergo fine-tuning on local RTX systems before being scaled and deployed, whether through on-premises data centers or the public cloud.

The recently unveiled software lineup includes Nvidia AI Foundation Models and Endpoints. This package integrates diverse RTX-compatible AI models and software development kits into the HP AI Studio, a centralized data science platform.

New PC Experiences

Nvidia reported that its in-house developers have been actively leveraging the new capabilities of the RTX GPU to introduce several generative AI applications. These applications are designed to run locally on their machines.

The company highlighted its close collaboration with PC hardware partners in developing applications like Nvidia RTX Remix. This tool is designed for crafting RTX-based remasters of classic video games. Nvidia RTX Remix is set to launch in beta this month, employing generative AI to elevate basic textures from classic games to modern, 4K resolution with physically-based rendering materials. The new Nvidia ACE microservices offer developers access to AI-powered speech and animation models, enabling the integration of intelligent and dynamic digital avatars into their games.

Additionally, TensorRT acceleration for the Stable Diffusion XL Turbo and latent consistency models provides a substantial performance boost of up to 60%, as highlighted by the company. Nvidia DLSS 3 now incorporates frame generation capabilities, utilizing AI to enhance frame rates by up to four times compared to native rendering. This capability is set to be featured in several upcoming video game titles, including “Horizon Forbidden West,” “Pax Dei,” and “Dragon’s Dogma 2.”