A few days ago, NVIDIA officially released the GeForce RTX 30 series graphics cards. The first launch includes three models: RTX 3090, RTX 3080, and RTX 3070. At present, the domestic official website has updated the price and time to market of the three graphics cards. According to the parameters, the RTX 3070 can achieve faster performance at a price less than the RTX 2080Ti, and netizens have expressed that they are crying in the toilet.
RTX 3080 domestic price starts at 5499 yuan
The GeForce RTX 3070 starts at 3899 yuan and will go on sale in October. Equipped with 8GB GDDR6 video memory, the front dual fan design, the price is less than half of the RTX 2080 Ti, but the average performance is higher, and it is 60% faster than the RTX 2070.
The GeForce RTX 3080 starts at 5,499 yuan and will go on sale on September 17. Replacing the RTX 2080 Ti, the performance is 2 times faster than the RTX 2080. With 10GB GDDR6X video memory, the equivalent frequency is 19GHz, and a stable frame rate of 60FPS can be obtained at 4K resolution.
The GeForce RTX 3090 starts at 11,999 yuan and will be released on September 24. Replacing the original position of the Titan series, it can achieve a stable frame rate of 60FPS under 8K ultra-high resolution when RTX and DLSS are turned on, which is 50% faster than Titan RTX. Three-slot volume, dual-axial streamlined radiator, claims to reduce GPU temperature by up to 30°C compared to Titan RTX, and is 10 times quieter. Equipped with 24GB GDDR6X video memory, it can cope with the most challenging artificial intelligence algorithms and meet the needs of large-scale content creation.
RTX30 series: the strongest GPU today
There are always breaking news before the new product launch, and VideoCardz directly exposed the core specifications of the three new products in one breath. After its official debut, the GeForce RTX 30 series has indeed lived up to expectations and has made leaps in many aspects:
1. New generation process + Ampere architecture: Currently the fastest and most efficient GPU architecture in the world, based on a customized 8nm Samsung process, integrating up to 28 billion transistors.
2. New programmable shader (streaming multiprocessor): The processing power is 30 Shader-TFLOP/s, and the FP32 floating point throughput is twice that of the previous generation.
3. The second generation RT Cores (ray tracing core): The computing power is 58 RT-TFLOPS, which is twice that of the previous generation products, and supports ray tracing, shading and calculation at the same time.
4. The third-generation Tensor Cores (AI): The computing power is as high as 238 Tensor-TFLOPS, which is twice that of the previous generation. It can run AI technologies faster and more efficiently, such as DLSS deep learning super-sampling anti-aliasing.
5. “Black technology” NVIDIA RTX IO: Use the GPU of the graphics card to decompress files, and in the PCIe 4.0 SSD environment, the processing efficiency can be comparable to that of a CPU with more than 20 cores. For gaming applications, NVIDIA RTX IO can speed up scenes and loading times, reducing user wait time. The input/output performance is up to 100 times that of traditional hard disk storage.
6. GDDR6X video memory: Provided by Micron, it supports PAM4 signaling and can provide ultra-high bandwidth of 760Gbps, which is currently the fastest video memory in the world.
7. Dual-axial flow cooling scheme: Different from the previous “gas stove” shape, this time the public version of the RTX 3090/3080 adopts an integrated front and rear dual fan design. Combined with the chassis fan and air duct, the heat dissipation performance is 2 times higher than the previous generation. 55% more, 30% more efficient, 3x more quiet. As for the RTX3070, it is still a dual fan.
8. Exquisite mechanical design and circuit design: The RTX 30 series adopts a more sturdy mechanical structure, a new low-profile leaf spring, and a new 12-pin power interface, leaving more space for components and heat dissipation. The header is compatible with existing 8-pin connectors. The PCB circuit board is half smaller than the previous generation.
9. HDMI 2.1: Higher bandwidth, for the first time supporting a data cable to connect to an 8K HDR TV, enabling ultra-high-resolution games.
10. AV1 decoding: The first discrete GPU to support the new AV1 decoder, 8K HDR network video requires 50% less bandwidth than before.
Overall, the GeForce RTX 30 series has achieved a number of world firsts in the GPU field.