At the disposal of the publication AdoredTV turned out to be information regarding the performance of the professional graphics accelerator NVIDIA RTX A6000, designed for use in workstations. In SPECviewperf 2020 tests, the NVIDIA RTX A6000 card was compared with its predecessor – NVIDIA Quadro RTX 6000. As it turned out, the new product based on the Ampere architecture is only 11% faster than the solution on the Turing architecture.
The source indicates that it is not yet known how the results reflect a real assessment of the situation, since the test was probably attended by an engineering sample of NVIDIA RTX A6000. So, in the end, the advantage of the novelty may be higher.
NVIDIA introduced the RTX A6000 and A40 graphics cards earlier this month. The manufacturer confirmed last week that it has decided to ditch the Quadro and Tesla brands in new models, as it believes that the difference between cards for workstation and computing solutions is becoming less noticeable. Now the company’s assortment includes three professional cards – A40, A6000 and A100, each of which is designed for its own specific market of users. But only the A6000 has the RTX prefix in its name, indicating that it is the successor to the Quadro RTX 6000, and both are intended for use in workstations.
The performance results of the new professional graphics accelerator can be found below. As you can see, the new product is only slightly superior to its predecessor in professional tasks, for which it should be sharpened. Perhaps in some other professional applications the advantage will be higher.
At the heart of the NVIDIA RTX A6000 is the same version of the NVIDIA GA102 Ampere GPU found in the consumer GeForce RTX 3090 graphics cards. There are also 10,752 CUDA cores, but the video memory is increased to 48 GB (GDDR6). The manufacturer has not yet announced the exact launch date for the RTX A6000 adapters, but the first deliveries are expected to begin in December.
If you notice an error, select it with the mouse and press CTRL + ENTER.