A well-known hardware leaker who has a reputation for accurately predicting the release dates of Nvidia’s upcoming hardware, @hongxing2020 has stated that the tech company is planning to officially unveil the GeForce RTX 4070 on April 13th.
According to him, this new graphics card will be priced lower than the GeForce RTX 4070 Ti that was released earlier, making it more affordable and attractive to gamers.
The launch of this graphics card will make Nvidia’s Ada Lovelace architecture more accessible to gamers. The move is expected to increase Nvidia’s popularity in the gaming industry.
3.11 update
4070 —>4/13
— hongxing2020 (@hongxing2020) March 11, 2023
However, this information has yet to be confirmed by Nvidia, and it is essential to approach it cautiously. The GeForce RTX 4070 is expected to be a more affordable graphics card than the RTX 4070 Ti, making Nvidia’s Ada Lovelace architecture more accessible to gamers.
The RTX 4070 is also rumored to have a reduced CUDA core count compared to the RTX 4070 Ti but still offers impressive computing performance. Once again, it is essential to note that these details have not been officially confirmed, so they should be taken as rumors until Nvidia makes an official announcement.
The forthcoming Nvidia GeForce RTX 4070 graphics processing unit will employ the identical AD104 graphics processor used in the previously introduced GeForce RTX 4070 Ti.
However, the non-Ti version of the RTX 4070 will have a significantly smaller processor, with only 5888 CUDA cores and a lower speed range of 1920 MHz to 2475 MHz. It contrasts the full-fat version of the RTX 4070 Ti, which features all 7680 CUDA cores. Despite this, the GeForce RTX 4070 will still boast a 12GB GDDR6X memory subsystem with a 192-bit interface.
Even though the GeForce RTX 4070 will have a smaller AD104 GPU than the GeForce RTX 4070 Ti, it is anticipated to provide remarkable computing power of approximately 29 FP32 TFLOPS, which is comparable to that of the current GeForce RTX 3080.
However, the GeForce RTX 3080 has a wider memory bus of 320 bits, allowing for a higher peak bandwidth of 760 GB/s. In contrast, the AD104 GPU in the GeForce RTX 4070 will only have a bandwidth of 504 GB/s with its 21 GT/s GDDR6X memory.
The anticipated compute performance of 29 FP32 TFLOPS of the GeForce RTX 4070 would position it as one of the best graphics cards in the market. It is yet to be seen how the upcoming GeForce RTX 4070 will perform in memory-intensive tasks such as high-resolution rendering with antialiasing when compared to Nvidia’s previous-generation GeForce RTX 3080.
Nvidia can guarantee a good supply of qualified AD104 GPUs necessary to manufacture the graphics card by reducing the GPU configuration of the GeForce RTX 4070. This approach will allow Nvidia and its partners to price the graphics card aggressively.
Since the GeForce RTX 4070 will be placed below the RTX 4070 Ti, which has a $799 MSRP, it is expected to be more affordable. However, the exact price difference is unknown at this time.
It is important to remain skeptical about the information concerning the launch date and hardware specifications of the GeForce RTX 4070 since it has yet to be officially confirmed. Unofficial sources leaked the details, so it is wise to approach the information cautiously rather than assuming it to be entirely accurate.