Given its $6,800 price ticket, you most likely will not see Nvidia RTX 6000 Ada on any lists of the greatest graphics playing cards for gaming. However nobody can deny that the RTX 6000 Ada is a powerhouse for skilled customers and content material creators. The RTX 6000 Ada Era, because it’s formally identified, options Nvidia’s flagship AD102 silicon, which the chipmaker additionally makes use of within the mainstream GeForce RTX 4090.
For the RTX 6000 Ada, Nvidia enabled 142 SMs out of the attainable 144, so the graphics card is simply two SM shy of utilizing the complete AD102 die. The RTX 6000 Ada arrives with 18,176 CUDA cores, 568 Tensor cores for AI workloads, and 142 RT cores for ray tracing duties. With the potential to spice up as much as 2,505 MHz, the graphics card flaunts a powerful FP32 efficiency that tops out at 91.06 TFLOPs. There’s additionally 48GB of 20 Gbps GDDR6 reminiscence onboard with a 384-bit reminiscence interface to land the graphics card at a reminiscence bandwidth of as much as 960 GBps. Word that the TGP (Complete Graphics Energy) on the 6000 Ada is capped at 300W, nonetheless, in comparison with 450W for the RTX 4090 Founders Version.
The earlier RTX A6000 delivered 38.71 TFLOPs FP32 efficiency and 768 GBps of reminiscence bandwidth. Subsequently, the RTX 6000 Ada affords 135% larger FP32 efficiency and 25% extra reminiscence bandwidth than the RTX A6000. The GeForce RTX 4090, with 82.58 TFLOPs FP32 efficiency and 1,008 GBps reminiscence bandwidth, might be the closest rival to the RTX 6000 Ada.
Whereas Nvidia has ported its Ada Lovelace structure over to the corporate’s skilled choices, AMD has but to launch any professional-grade graphics card primarily based on RDNA 3. The one competing product at the moment is the Radeon Professional W6800, relationship again to the RDNA 2 days. Sadly, the Navi 21-based graphics card’s 17.82 TFLOPs FP32 efficiency and 512 GBps reminiscence bandwidth look fairly puny beside the RTX 6000 Ada. However let’s examine what Puget’s benchmarks must say.
RTX 6000 Ada Benchmarks
| Graphics Card | DaVinci Resolve Studio | Adobe Premiere Professional | Topaz AI Suite | Unreal Engine | V-Ray | OctaneRender | Redshift | Blender |
|---|---|---|---|---|---|---|---|---|
| RTX 6000 Ada | 2,975 | 1,498 | 471.2 | 74.0 | 5,399 | 1,146 | 87 | 12,728 |
| GeForce RTX 4090 | 2,960 | 1,505 | 465.5 | 92.5 | 5,564 | 1,292 | 85 | 12,779 |
| RTX A6000 | 2,648 | 1,379 | 445.2 | 46.3 | 2,710 | 627 | 159 | 5,848 |
| Radeon Professional W6800 | 2,135 | 1,017 | 401.6 | 24.7 | N/A | N/A | N/A | 1,982 |
In line with Puget Programs’ thorough evaluation (opens in new tab) of the RTX 6000 Ada, skilled customers can anticipate as much as 2X larger efficiency from the Ada-powered graphics card in comparison with its Ampere predecessor. After all, decreasing workload occasions in half is all the time a welcome enchancment. The outcomes did not come as a complete shock for the reason that RTX 6000 Ada specs insinuate as a lot, nevertheless it’s nonetheless good to have benchmarks that quantify efficiency.
The GeForce RTX 4090 was superior to the RTX 6000 Ada, outperforming its skilled counterpart in six out of eight benchmarks. The GeForce RTX 4090 delivered between 3% to 25% larger efficiency. The Radeon Professional W6800 wasn’t a match for the RTX 6000 Ada, as anticipated. The latter was about 17% to 542% sooner than the previous, relying on the workload.
It is past dispute that the RTX 6000 Ada is a strong possibility for skilled workloads. Sadly, the eye-watering $6,800 MSRP is sufficient to scare off potential consumers, even when the graphics card will possible repay in the long term. Word additionally that not one of the above benchmarks profit from Nvidia’s skilled drivers — Siemens NX for instance sometimes sees huge efficiency good points in comparison with GeForce playing cards.
The GeForce RTX 4090 is a pretty various for the much less severe person because it begins at $1,650. You lose the soundness and software program validation you get from the Nvidia skilled drivers and 24GB of VRAM. Nevertheless, in the event you do not want a blower GPU and use purposes that are not particularly de-optimized by Nvidia’s GeForce drivers, the RTX 4090 ($1,599 MSRP) can lend you its firepower for 76% lower than the price of an RTX 6000 Ada. Little marvel Nvidia does not appear to need extra compact blower variants of the RTX 4090.