Kopite7kimi’s latest predictions indicate the RTX 4090 will need up to 450W of control, just as the $1,999 RTX 3090 Ti that Nvidia placed on sale in March but offers quite a rare more cores than the last-gen chip: 16,128 paired with 24GB of GDDR6X memory at 21Gbps.
Now that sanity is bearing the market, did you finally find a gaming GPU within your budget? Here’s a motivation to keep on waiting: the Nvidia RTX 4090 might arrive in mid-July.
That’s the term from kopite7kimi, the leaker who perfectly anticipated the specifications of the RTX 3090 and RTX 3090, among other things.
And while that’s only roughly 50 percent more bodies than the RTX 3090, it’s on Nvidia’s new Ada Lovelace layout, as reported. In addition, it crams in a bunch of other hardware, enough that some watchers suspect we may notice double the raw performance of this hardware generation.
Wccftech includes a reasonable breakdown of the math, if you’re curious, though Kimi’s suggesting fewer cores and lower power (450W vs. 600W) for the RTX 4090 today than previously.
Kimi’s been trickling out predictions for months, but we’ve been waiting for them to firm up. As you might expect, the ones proactively marked “April Fools’ Day” didn’t give us much faith!
Speaking of April Fools’, recognize when one Nvidia fan supposed the RTX 4090 would be a 1,000W card with two power cables and 48GB of memory? Last month, Kimi meant Nvidia is testing a 900W card with 48GB of memory that requires two 16-pin power connectors to function.
Nvidia GeForce RTX (Ray Tracing Texel eXtreme) is a high-end professional visual computing platform developed by Nvidia, predominantly used for designing complex large-scale models in architecture and product design and energy exploration, games, scientific visualization, and film and video production.
Nvidia RTX enables real-time ray tracing. Historically, ray tracing had been reserved for non-real-time applications (like CGI in visual effects for movies and photorealistic renderings), with video games relying on direct lighting. However, it has a precalculated indirect contribution to their rendering.
RTX facilitates a new development in computer graphics of generating interactive images that react to lighting, shadows, and reflection. RTX runs on Nvidia Volta-, Turing- and Ampere-based GPUs, explicitly utilizing the Tensor cores (and new RT cores on Turing and successors) on the architectures for ray-tracing acceleration.
In March 2019, Nvidia announced that established GTX 10 series (Pascal) and GTX 16 series (Turing) cards would receive support for subsets of RTX technology in upcoming drivers. However, their lack of dedicated hardware cores for ray tracing will affect functions and performance.
In October 2020, Nvidia announced Nvidia RTX A6000 as the first Ampere-architecture-based graphics card for use in professional workstations in the Nvidia RTX product line.
Nvidia worked with Microsoft to integrate RTX support with Microsoft’s DirectX Raytracing API (XDR). As a result, RTX is currently available through Nvidia OptiX and for DirectX. For the Turing and Ampere architectures, it is also available for Vulkan.