Nvidia has announced a wave of “AI-ready” laptops with its partners using the new “Super” line of GPUs. Nvidia
Electronic chip giant Nvidia took advantage of the Consumer Electronics Show to unveil several products aimed at consumer technology, the automotive market and robotics. It also presents a series of laptops equipped with ChatGPT-type generative artificial intelligence technology.
“The big hit at CES this year will be generative AI and AI PCs,” said Justin Walker, chief product officer at Nvidia.
In the field of consumer AI, Nvidia is positioning the PC as a platform for the individual use of LLMs, such as Meta’s Llama 2 or Stable Diffusion.
Graphics cards for local AI on new computers
The company announced a new series of RTX graphics cards, called the RTX 40 Super series, which it claims can generate videos and images 1.7 times and 1.5 times faster respectively. Several PC manufacturers are launching laptops, which Nvidia describes as “AI-ready”, with these new cards. Acer, ASUS, Dell, HP, Lenovo, MSI, Razer and Samsung have models like this, and will ship this month, Nvidia said.
Workstation cards, three models, 4070 SUPER, 4070i SUPER, and 4080 SUPER, will also go on sale this month for $599, $799, and $999, respectively.
Building on the October release of a programming library, TensorRT-LLM for Windows, Nvidia is emphasizing the personal computer as a workstation for training large language models, in conjunction with cloud services.
Video game NPCs that use 4 AI models
The company’s recently announced AI Workbench software for PC will go live in beta later this month. It allows programmers to grab language models from repositories like Hugging Face and push them to cloud instances, including Nvidia’s DGX Cloud. The RTX cards can then be used on the local PC for “inference and lightweight customization,” the company said.
Still on the consumer market, Nvidia announces improving avatars in games with automatically generated speech. Building on a service called ACE, for “avatar cloud engine”, the company announced “Production Microservices” for ACE which allows game developers to integrate Nvidia programs for voice recognition and facial animations into their games.
As Nvidia’s Mr. Walker explained, this technology will allow the player “to speak to the character in the game,” with their voice passed to automatic speech recognition, “which translates that speech into text, and then we take the text and we feed it into a large language model to generate the character’s response.”
Nvidia to the rescue of iStock on AI Gen
An LLM “can create an answer that’s appropriate and isn’t set in stone – it won’t be the same answer every time,” Walker said. “This response can be vocalized using another model, a text-to-speech model, and then passed to a third model that performs audio lip sync” so that the NPC’s facial muscles move in sync with the vocal response .
“It’s a combination of four different artificial intelligence models to do all of this,” Walker said. “And these AI models, within our platform, are flexible: You can run these models on the cloud, you can run these models locally, on your RTX PC, so it can be powered to help deliver this experience.”
Building on the ACE service, for “avatar cloud engine”, Nvidia announced “Production Microservices” for ACE which allow game developers to integrate Nvidia programs for voice recognition and facial animations into their games. Nvidia
Still for the consumer market, photography giant Getty Images announced a new service, Generative AI by iStock, which it designed using Nvidia’s “Picasso” program. The iStock service “allows anyone to create 4K images from text using an AI model trained on Getty Images’ catalog of licensed, commercially safe creative content,” the two companies said. The service is available immediately on istock.com.
Nvidia strengthens its position in the automotive sector
In addition to initiatives aimed at the general public, Nvidia has focused on news concerning its customers and partners in the automotive market, which always presents concept cars at CES.
Nvidia announced new electric vehicle customers, including Li Auto, which will use its “Drive Thor” computing platform, a board Nvidia describes as a “mainframe for safe and secure autonomous vehicles.” In addition, electric vehicle manufacturers GWM (Great Wall Motor) and Xiaomi will use Nvidia’s “DRIVE Orin” chips to realize ADAS (automated driving assistance) functions.
Mercedes-Benz also presented the “CLA Class Concept”, which uses Nvidia’s “DRIVE Orin” platform. Nvidia noted that Mercedes-Benz also uses “digital twin” technology in its design process through Nvidia’s Omniverse software platform.
And here are all the announcements from CES 2024
Source: “ZDNet.com”