Skip to main content

Nvidia might finally fix its VRAM problem — but it will take time

The Razer Blade 14 and 18 on a table.
Luke Larsen / Digital Trends

It’s no secret that some of Nvidia’s best graphics cards could use a little more VRAM. According to a new leak, Nvidia may be addressing that problem in a big way — at least in laptops. The RTX 5090 laptop GPU is now reported to come with 24GB VRAM across a 256-bit memory bus. The downside? These new laptops might not make it to market as soon as we’d hoped.

The information comes from Moore’s Law Is Dead, who cites his own industry sources as he spills the beans on RTX 50-series laptop specs. Up until now, we’ve not heard much about Nvidia’s plans for RTX 50 laptops, indicating that they might be a few months away. The YouTuber agrees with this, saying that Nvidia might be targeting a launch window in the first or second quarter of 2025. This might not affect the entire lineup, though.

The reason why Nvidia might be planning to increase the VRAM on the laptop RTX 5090 is that it’s planning to use newer GDDR7 memory modules for it, with 3GB, up from 2GB. This boosts the memory capacity from 16GB (which the company used in this generation) to 24GB. However, as Moore’s Law Is Dead speculates, that also might be why Nvidia is holding off on the RTX 5090 laptop GPU in particular.

Speculation about the RTX 5090 laptop configuration.
Moore's Law Is Dead

According to the YouTuber, Nvidia’s current plan is to first launch the RTX 5090 and the RTX 5080 in their desktop versions this year. It’s worth noting that most leakers disagree with this timeline — the GPUs aren’t expected to be announced until CES 2025 at this point.

Following the reported desktop launch, Nvidia is said to be unveiling some RTX 50-series laptops at CES 2025 in January, but those will all use the memory modules that are ready now, meaning the 2GB versions. As per the source cited by Moore’s Law Is Dead, the RTX 5090M doesn’t come in a 16GB configuration at all, which is why it won’t be launched alongside the rest of the lineup. Laptop cards like the RTX 5080, RTX 5070, and RTX 5060 might all still make an appearance at CES and launch shortly after with 16GB, 12GB, and 8GB of memory, respectively. They’ll use the 2GB modules, though.

One interesting thing to note is that, while Nvidia will likely use the GB203 die for the desktop RTX 5080 and then use that same die for the RTX 5090 (laptop version), the desktop card will still be limited to just 16GB of memory. In the grand scheme of things, this kind of adds up — if Nvidia is launching desktop GPUs first, the RTX 5080 wouldn’t have been ready with more memory assuming that these 3GB modules aren’t readily available yet. But all of this is still just speculation, and the final specs are unknown to the public.

More VRAM on laptop GPUs is certainly good news for gamers, but it’s hard to know how much of this will turn out to be true. As things stand now, we might not have any concrete information until CES 2025.

Monica J. White
Monica is a UK-based freelance writer and self-proclaimed geek. A firm believer in the "PC building is just like expensive…
After a decade, Nvidia is fixing the worst part of G-Sync
Alan Wake 2 on the Alienware 27 QD-OLED gaming monitor.

It only took 11 years. Nvidia is finally doing away with its proprietary G-Sync module that's been the bane of the best gaming monitors for several years. Although Nvidia eventually started offering G-Sync branding with any variable refresh rate (VRR) tech, display brands have still needed to shell out for a dedicated G-Sync module for proper certification -- and pass along that cost to customers. Going forward, the G-Sync module is no more.

Nvidia's G-Sync tech isn't going anywhere, however. The company announced that it partnered with Mediatek to add every G-Sync feature to Mediatek scalers. If you're unfamiliar, a scaler is basically a small chip inside your monitor that handles resolution scaling, along with a bunch of other processing including the on-screen display (OSD), color adjustments, HDR, and input selection. The G-Sync module itself was a scaler. Now that you'll rarely find a gaming monitor without its own scaler, those features are being rolled into the chip already built into the display.

Read more
Final Fantasy XVI already runs incredibly on PC, but there’s one problem
Clive fighting an enemy in Final Fantasy 16.

The moment Final Fantasy XVI hit Steam, I had the download queued. I plan on doing a closer look at the PC release once it arrives on September 17, but after playing through the opening section of the game, it's already looking like a promising PC port. It's well-optimized on the CPU, there are a ton of graphics options, and it's loaded with PC gaming tech. There's just one problem I'd like to see addressed before launch.

That mainly concerns the game's PlayStation 5 roots, where it ran either at 30 frames per second (fps) in resolution mode, or a wildly unstable 60 fps in its performance mode. Despite the PC release supporting frame rates up to 240 fps and running much smoother on modern hardware, its basis on PS5 still shows up, particularly in cutscenes.
Cutscene woes

Read more
This new patch might finally fix the issues with Intel CPUs
Intel's 14900K CPU socketed in a motherboard.

The problems with Intel's best processors have been going on for months, and Intel's response has been measured. However, at long last, a fix might be imminent -- provided that this one actually helps. A couple of motherboard vendors have already announced BIOS updates that address issues within the microcode of Intel's 13th- and 14th-gen CPUs. With more to come, things might be looking up for the first time in a long time.

ASRock and MSI both announced BIOS updates, and those patches are already available for some motherboards. The following MSI boards are getting the update first:

Read more