NVIDIA Introduces Quadro RTX 6000 Along With Two Other Ray Tracing Models

Nvidia introduced three new Quadro RTX models, the first of which was the 8000 model we talked about in a previous article as well as taking the time to talk about the RTX 6000 model.

RTX 6000, NVIDIA Introduces Quadro RTX 6000 Along With Two Other Ray Tracing Models, Optocrypto
Nvidia introduced the models Quadro RTX 8000 – RTX 6000 and RTX 5000

NVIDIA today introduced the world’s first graphics processor that can display a real-time ray tracing scene: the RTX Quadro graphics card with Turing architecture. The series looks like an absolute beast, with enough processing power to decimate ray tracing scenes in real time.

The NVIDIA Quadro RTX monster graphics card has 48GB of GDDR6 memory and is capable of recording ray tracing scenes in real time with integrated Turing

RTX 6000, NVIDIA Introduces Quadro RTX 6000 Along With Two Other Ray Tracing Models, Optocrypto

It has shown what its Turing GT104 Matrix seems to be and is an absolute monster. At the same time, they announced that this flagship GPU will be part of the new Quadro RTX family. The Quadro RTX can process 10 GigaRays per second, including the RTX 6000, with up to 16 TFLOPs of processing power. It can also offer 500 billion tensor operations per second and up to 100 GB/s bandwidth with NVLink.

Unlike the Pascal GPU, which consisted mainly of shading and arithmetic units, the Turing GPU is a revolutionary advance in design and consists of three different parts. These are the Tensor Core, the RT Core, and the Shader & Compute Core. When the company says core, it actually means segments. The voltage core is specified in 125 TFLOPs of the FP16 (for IA and DL), while the Ray Tracing core is designed at 10 Giga Rays per second.

RTX 6000, NVIDIA Introduces Quadro RTX 6000 Along With Two Other Ray Tracing Models, Optocrypto

The chip consists of 18.6 billion transistors and measures 754mm². The basic clock rate is 1.75 GHz and offers 48 GB GDDR6, the model RTX 6000 has 24 GB memory. It uses a 384-bit bus design and has a 14 Gbps clock rate for a net total of 672 GB/s.

Differences between the three models:

s

GPU MEMORY MEMORY WITH NVLINK RAY TRACING TENSION NUCLEI NUCLEOS
TENSOR
PRICE ($)
Quadro RTX 8000 48GB 96GB 10 GigaRays / sec 4,608 576 10000
Quadro RTX 6000 24GB 48GB 10 GigaRays / sec 4,608 576 6300
Quadro RTX 5000 16 GB 32GB 6 GigaRays / sec 3,072 384 2300

From what we see, the only difference between the RTX 8000 model and the RTX 6000 will be in the amount of memory, 48 versus 24 GB, then it has the same amount of GigaRays, CUDA cores, and Tensor cores.

The Quadro RTX series will be available during the fourth quarter of this year.

 Source: Wccftech