Nvidia made the first big live-streamed announcement at CES on Tuesday, a day before the primary day of launches. It talked about its GeForce 40-series mobile GPUs and the long-rumored RTX 4070 Ti desktop GPU. AT&T has already planned a six-months-free promotion for the new GeForce Now Ultimate tier, which was a big surprise. The company also shared some news about its commercial tools for cars, robotics, and collaborative design.
RTX 40-series graphics for mobile
Nvidia released a full line of mobile GPUs, from the low-end RTX 4050 to the high-end RTX 4080 and 4090. The RTX 4050 is for cheap discrete graphics that are barely there. The RTX 4060 and 4070 are for mainstream or thin-and-light gaming and graphics laptops.
The new mobile chips use less power thanks to the Ada Lovelace architecture. This means that Nvidia has made a new generation of its Max-Q power-management technology: It has ultra-low voltage DLSS 3 and “tri-speed memory control” to change the memory state on the fly to one that uses less power and more. From what I saw with the 4080 and 4090, DLSS has improved since the last generation. And finally, Advanced Optimus, Nvidia’s way of letting the GPU live on the same bus as the CPU, is starting to catch on. This allows you to use G-Sync on the built-in display and switch to the integrated graphics to use less power without restarting the system. (Every time I hear the phrase “MUX Switch,” a little bit of my soul dies.)
It showed 14-inch laptops that aren’t made for gaming, like the Lenovo Yoga Pro 14 and Asus ZenBook Pro 14, with RTX 4070, 4060, or 4050 mobile chips that will ship in late February starting at $999. Beginning in early February, you can buy gaming laptops like the Alienware x16 with an RTX 4080 or RTX 4090 for as little as $2,000.
GeForce RTX 4070 Ti Desktop
Nvidia first announced the 12GB card as a low-end RTX 4080, but people pointed out that its specs didn’t match those of a xx80-class GPU, so Nvidia “unlaunched” the card. It has since been reborn as the RTX 4070 Ti, which will start shipping on January 5 for $800 and go up from there.
What seems especially interesting is that there doesn’t seem to be an Nvidia-branded Founders Edition version of the card, which there usually is for GPUs at this level. This means there is no guarantee that a card will be sold at that entry-level price. However, we could always count on an Nvidia Founders Edition model to stick to the announced price even if it went out of stock often and stayed out of stock.
GeForce Now Ultimate
Nvidia has also upgraded the back-end cloud servers for its cloud-gaming service from RTX 3080-class GPUs to RTX 4080-class GPUs. This means that its top-tier cloud-gaming service option has also been upgraded. By going with “Ultimate,” Nvidia won’t have to change its name every time it boosts, as it did with “RTX 3080.”
You get the same benefits for the same $20 per month, but the card works better. That can mean 240 frames per second instead of 120 frames per second (the details are unclear). When the new plan is ready, current RTX 3080 subscribers will automatically switch to it. As usual, it will be added to different areas bit by bit.
If your car has Nvidia Drive technology, you may also be able to get GeForce Now as part of its entertainment system. Now, all you need is a way to set up routes based on how good your cell signal is so that you don’t get stuck.
Two important programs that run on RTX GPUs are now part of the family. The beta Eye Contact effect will be added to Nvidia Broadcast. Faking eye contact during videoconferences and presentations is the new big thing I don’t like (Windows also has it). I’ve never seen an implementation that didn’t bother me, and I think at least one of the presenters in the stream was using it because of the unblinking thousand-yard stare that didn’t look at you as much as it looked through you. But maybe that’s just me.
The other feature that could be important is RTX Video Super Resolution, which is meant to make a streaming video on Chrome and Edge better. It uses AI upscaling and artifact reduction to make 1080p video look better on screens with more pixels. That will work on GPUs from the RTX 30 and 40 series.
And this quarter, Nvidia’s Canvas generative AI sketch tool, which can be used on any RTX GPU, will go into beta.
Nvidia also shared some new information about its technologies for developing robots and cars. They include new features in its Isaac Sim environment, like the ability to model multiple humans and arrays of robots (for AI training). Back-end technologies don’t get a lot of attention at CES. That’s more the job of Nvidia’s GTC and GDC conferences, which are geared toward designers and developers. Most of the news was about partnerships and updates on features getting early access. If that’s what you’re interested in, you can find all the details on Nvidia’s website instead of asking me to sort them out for you.
For more such news, stay connected to our homepage NogMagazine.com.