A recent report from VideoCardz states that Nvidia has notified its partners of its plan to launch the GeForce RTX 4070 graphics card in April. This product is expected to be a mainstream performer, utilizing the Ada Lovelace microarchitecture. While the exact launch date has not been specified, Nvidia is aiming for an April release.
The upcoming GeForce RTX 4070 graphics card is expected to share similarities with its counterpart, the GeForce RTX 4070 Ti. It is anticipated that the AD104 graphics processor will power both products and feature 12GB of GDDR6X memory. However, Nvidia is predicted to significantly limit the performance of the GeForce RTX 4070 by reducing its CUDA cores to 5888, with a clock speed range of 1920 MHz to 2475 MHz.
Reports suggest that the upcoming RTX 4070 graphics card will have a peak FP32 compute performance of 29 TFLOPS, which is lower than its counterpart, the RTX 4070 Ti, which offers a higher 40 TFLOPS. The RTX 4070’s lower compute performance translates to lower power consumption, which means it won’t require a high-end cooling system to keep it running. Therefore, the RTX 4070 will likely be a top contender for inclusion in our list of the best graphics cards in the Spring season.
It is anticipated that the NVIDIA GeForce RTX 4070 will be equipped with either AD104-250 or AD104-251 GPU cores, both of which are scaled-down versions of the Nvidia Geforce RTX 4070 Ti, comprising 5,888 CUDA Cores. The cores will operate at a base frequency of 1,920 MHz, which is lower than the frequency of the Ti model, i.e., 2,310 MHz. However, the RTX 4070 will have a Turbo frequency that can achieve 2,475 MHz.
Performance
A reported peak performance of 29 TFLOPs is expected for the NVIDIA GeForce RTX 4070, a considerable drop from the Nvidia Geforce RTX 4070 Ti’s 40 TFLOPs. However, the memory configuration for both GPUs is expected to be the same, offering 12GB of 21Gbps GDDR6X memory and a 192-bit memory interface capable of delivering a bandwidth of 540GB/s. The power consumption for the RTX 4070 is anticipated to be a TDP of 200W, representing an 85W reduction compared to the top-tier model.
The RTX 4070 is anticipated to have 5,888 CUDA cores that improve graphics’ visual appeal, coupled with a maximum boost clock speed of 2,475MHz. In contrast, the RTX 4070 Ti has been confirmed, launched, and benchmarked, featuring 7,680 cores and a higher ultimate boost clock of 2,610MHz.
The RTX 4070’s peak performance, which is reportedly 29TFLOPs, is equivalent to the performance of the previous generation’s RTX 3080 from the Ampere series. This comparison helps to put into perspective the potential performance of the RTX 4070. It’s important to note that this still represents a nearly 50% increase in performance compared to the RTX 3070, which had a peak performance of 20TFLOPs in the previous generation.
Price
The price is a crucial factor when it comes to the RTX 4070. The RTX 4070 Ti was initially sold for an MSRP of $799, but it isn’t easy to find one at that price point. Our review suggested that the card be priced at least $700. Nvidia has faced considerable criticism for its pricing strategy for the Ada Lovelace series, so it will be intriguing to observe whether the company has paid heed to the suggestions and criticisms provided by its customers into account while deciding the price for the RTX 4070. It’s worth noting that the RTX 3060 Ti was launched with an MSRP of $400, while the standard RTX 3060 had a price tag of $330.
The company is likely examining how much inventory its partners have of the GeForce RTX 30 series before choosing the precise launch date. It’s still being determined how much Nvidia intends to charge for the GeForce RTX 4070, but the MSRP will be less than $799. The extent to which the price is lowered will be a significant factor in the card’s appeal.
The AD104 SKU initially expected to have higher specs, has now been renamed the RTX 4070 Ti. As a result, the lower-spec AD104 SKU is now referred to as the RTX 4070. The anticipated specifications for the RTX 4070 have changed multiple times, but according to recent leaks, the GPU is expected to be significantly weaker than previously expected.
Initially, we believed that the graphics card in question was the RTX 4070. However, it turns out that it was the RTX 4070 Ti, which could be a marketing ploy by Nvidia to earn extra profits from the naming conventions of its products. Unfortunately, this development could be more disheartening since the specifications have deviated from what we expected, based on the twitter leaks shared by Kopite7Kimi last August.