Computex 2024 kicks off this week, and Nvidia is at the event showing off a new assortment of tools aimed at making it easier to use AI on its graphics cards.
This includes new tech demos running on laptops packing Nvidia GeForce RTX 40-series mobile GPUs, as well as a new toolkit of SDKs and other tools designed to help devs optimize generative AI models to run on Nvidia cards.
Sure, it's not as exciting as new silicon, but these new releases from Nvidia suggest we'll see even more applications of AI for both gaming and productivity use in 2024 and beyond.
Here's everything you need to know about what Nvidia brought to Computex 2024.
Project G-Assist, an AI assistant
The big showcase demo Nvidia brought to Computex 2024 is Project G-Assist, an AI assistant designed to use generative AI techniques to help players in games.
Nvidia partnered with Studio Wildcard to develop a demo showcasing the tech running in Ark: Survival Evolved. Put simply, it takes text or voice input from the player and uses a large language model (LLM) to try and answer questions about the game, effectively serving as an AI guide or helper.
The Project G-Assist toolkit also includes AI-powered tools designed to help players optimize their PC settings for a given game.
Nvidia RTX AI Toolkit
Computex 2024 is also the place to showcase the Nvidia RTX AI Toolkit, a new package of tools and SDKs designed to help developers run and launch big generative AI models on Windows PCs.
The toolkit encompasses a handful of SDKs and tools for customizing and running AI apps on PCs with Nvidia cards, and the company plans to make it broadly available for developer access this month (June).
Developers can look forward to new tools like an Nvidia AI Inference Manager SDK, and partners like Adobe and Blackmagic Design have confirmed plans to use this new toolkit in their apps to enhance AI performance on Nvidia-equipped PCs.
Windows Copilot Runtime GPU acceleration is getting easier
Nvidia also announced plans to partner with Microsoft in order to make it easier for developers to tap the power of Nvidia graphics cards for use by Windows Copilot and other AI tech.
Microsoft and Nvidia will release an API in developer preview later this year aimed at helping developers of Windows apps utilize GPU-accelerated small language models (SLMs) and retrieval-augmented generation (RAG) features.
Outlook
If you noticed the word developer and preview popping up a lot in this story, it's because Nvidia is all in on pushing AI tools and tech at Computex 2024.
While our on-site team will walk past at least a dozen new laptops packing Nvidia cards at Computex, none are more beefy than a GeForce RTX 4070. No, the new hotness is AI tech demos and Windows Copilot features running locally on laptop GPUs, something we expect to see more of as the year progresses.
More from Tom's Guide
- WWDC 2024 — iOS 18, a new macOS and Apple’s big AI push
- I tested the new Chromebook Plus AI features — has Copilot met its match?
- Tom's Guide asks: What would sell you on a Copilot+ PC?
