More than two excruciatingly lengthy years after the RTX 30-series reveal, a brand new technology of graphics playing cards is lastly right here. Nvidia CEO Jensen Huang unveiled the hotly anticipated GeForce RTX 4090 and never one, however two totally different RTX 3080 variants in the course of the “Project Beyond” reveal occasion that kicked off GTC 2022.
These graphics playing cards look completely monstrous, full cease, with the RTX 4090 leaving the RTX 3090 Ti stumbling in its wake. But whereas the efficiency of those next-gen GeForce GPUs guarantees to soften your face, the uncooked speeds are removed from the one attention-grabbing side of this launch. Here are seven must-know info from the RTX 4090 and RTX 4080 reveal, from ray tracing developments to staggeringly excessive new sticker costs for Nvidia’s GPUs.
DLSS 3 and Lovelace’s huge ray tracing enhancements
Nvidia
Nvidia clearly designed its new “Ada Lovelace” structure to scream by ray tracing duties. At its core, the GPU makes use of new “shader execution reordering” to enhance conventional sport rendering by a claimed 25 p.c, however ray-traced scenes two to 3 occasions quicker. Beyond the shader cores, new applied sciences launched within the third technology of devoted ray tracing {hardware} discovered inside Lovelace (resembling Opacity Micro-Maps) additionally intention to drastically velocity up ray tracing efficiency.
But Nvidia’s “RTX” push doesn’t depend on ray tracing brute pressure alone. Huang additionally revealed DLSS 3, working on a brand new technology of tensor cores and designed to interchange the vaunted second-gen model of Nvidia’s AI-enhanced upscaling characteristic.
While earlier DLSS iterations used machine studying to generate new pixels, permitting GeForce graphics playing cards to run video games at decrease resolutions (for larger efficiency) earlier than upscaling them to suit your monitor’s decision, DLSS 3 generates completely new frames to determine optimum upscaled picture high quality. It doesn’t even must render the pixels in these behind-the-scene frames, which (Huang claims) permits DLSS 3 to drastically velocity up video games even when they’re CPU-bound moderately than GPU-bound—then confirmed off Microsoft Flight Simulator working over twice as quick with the know-how lively. Cyberpunk 2077 noticed an excellent larger uplift, going from roughly 22 frames per second with ray tracing on in a demo, then 100fps with DLSS enabled. Whoa.
The proof shall be within the impartial testing pudding, however the eye-melting promise of real-time ray tracing could lastly be achievable with the RTX 40-series.
The GeForce RTX 4090 is quick. Really quick.
Nvidia
Since that is GTC—a developer-focused occasion—Huang didn’t wade an excessive amount of into conventional gaming efficiency, however he teased some tantalizing tidbits. The RTX 4090 is as much as two occasions quicker than the RTX 3090 Ti in conventional video games like Flight Simulator, 3 times quicker in a brand new ray-traced model of Portal revealed on the announcement, and a whopping 4 occasions quicker in Nvidia’s personal “Racer X” ray tracing tech demo (which, to be truthful, is probably going optimized to benefit from Lovelace’s greatest technological developments).
The GeForce RTX 4090 will not be for players
Much just like the RTX 3090 earlier than it, the RTX 4090 could technically be a GeForce-branded graphics card, however its not actually for players. Instead, with its ample 24GB of GDDR6X reminiscence and staggering $1,599 price ticket, it as soon as once more seems like an absolute powerhouse for prosumers who can put the GPU to work in the course of the day, then take pleasure in face-melting framerates in video games at evening.
Again, nevertheless, we’ll want to attend for impartial opinions to make certain. But most players would little doubt be higher off wanting on the new duo of RTX 4080 choices.
What is the RTX 4080 12GB and 16GB?
Nvidia
Nvidia revealed a pair of GeForce RTX 4080 choices. One comes with 16GB of GDDR6X reminiscence and begins at a staggering $1,199—the identical worth as last-generation’s RTX 3080 Ti. Oof. The 12GB RTX 4080, then again, begins at $899.
There’s a large gulf in pricing between the 2 fashions, however Huang didn’t get into technical variations between them, and didn’t even tease gaming efficiency for the 4080 duo. Usually, identically named GPUs solely change within the reminiscence capability provided, however early RTX 40-series rumors steered the 12GB RTX 4080 may truly wield a less-potent, completely totally different GPU than the 16GB mannequin, regardless of rocking the identical title. Lets hope that doesn’t maintain true, because it’s a deeply anti-consumer follow, however the wild pricing hints that it is likely to be. Stay tuned.
RTX 40-series is rather more costly
Now for some unhealthy information. Worries about even-higher worth factors for the RTX 40-series proved correct. The RTX 4090 begins at $1,599, a $99 premium over the RTX 3090. The 16GB GeForce RTX 4080 prices $1,199, the identical worth because the RTX 3080 Ti and an astonishing $400 extra than the $700 RTX 3080. Finally, the mysterious 12GB RTX 4080 begins at $899, a $200 enhance over its predecessor.
Ick.
The GeForce RTX 4090 launches October 12. Availability for the RTX 4080 duo is unknown.
The RTX 30-series is sticking round
Nvidia
The RTX 40-series could also be right here, however Huang stated that the RTX 30-series shall be sticking round at extra mainstream worth factors. It is smart—the autumn of cryptocurrency signifies that GPU makers are at the moment sitting on warehouses stuffed with unsold graphics playing cards. But these new RTX 40-series costs damage, and if the RTX 4090 and 4080 don’t come out of the gate swinging laborious, unsold RTX 30-series GPUs and used graphics playing cards could very effectively be a significantly better worth. We’ll see.
AMD’s RDNA3 counterpunch is coming quickly
Finally, whereas Nvidia beat AMD to the punch with the launch of this new GPU technology, Radeon’s counterpunch ain’t far behind. Mere hours earlier than Jensen Huang’s keynote, Radeon chief Scott Herkelman tweeted that AMD’s next-gen RDNA3 Radeon cards will be revealed November 3, whereas AMD Technology Architect Sam Naffziger penned a weblog put up touting RDNA3’s huge effectivity enhancements—a direct shot on the presumed excessive energy ranges that Nvidia’s RTX 40-series would require to run at peak speeds.
It’s been a bleak couple of years for PC players, however now cryptocurrency is collapsing, provide chain woes are night out, and we’ve obtained a fierce battle for next-gen GPU supremacy brewing. Giddy up—and begin saving these pennies. Those costs are excessive.