Nvidia has quietly introduced a new version of its popular RTX 4070 graphics card, now featuring GDDR6 memory instead of the GDDR6X found in the original model. This move aims to improve supply and meet high demand for the GPU.
Key Specifications
The new RTX 4070 with GDDR6 memory maintains most of the specifications of its predecessor:
- AD104-251 GPU with 5888 CUDA cores
- 12GB memory capacity
- 192-bit memory interface
- 200W TDP
The primary difference lies in the memory speed, which is now 20 Gbps compared to the previous 21 Gbps. This results in a slightly lower memory bandwidth of 480 GB/s, down from 504 GB/s in the GDDR6X version.
Performance Claims
Nvidia asserts that the new GDDR6 variant will offer "similar performance in games and applications" compared to the original GDDR6X model. However, the company has not provided specific benchmarks or comparisons to support this claim.
Availability and Pricing
The RTX 4070 with GDDR6 memory is expected to hit the market in September, available through Nvidia's GPU partners. While pricing details have not been officially announced, it's anticipated that the new variant will be priced similarly to the existing model.
Implications for Buyers
With the introduction of this new variant, potential buyers should pay close attention when purchasing an RTX 4070 to ensure they're getting the desired memory configuration. Both versions will coexist in the market, offering consumers a choice between the two memory types.
Industry Context
This move by Nvidia comes amidst rumors of supply constraints for GDDR6X memory modules. The introduction of a GDDR6 variant could be seen as a strategy to diversify the supply chain and meet market demand more effectively.
As the new RTX 4070 with GDDR6 memory enters the market, it will be interesting to see how it performs in real-world tests and how it impacts the availability and pricing of graphics cards in this segment.