In an unexpected move, Nvidia has discreetly announced a new version of its popular RTX 4070 graphics card. This latest iteration swaps the original GDDR6X memory for standard GDDR6, marking a significant change in the card's specifications.
 |
|---|
| An illustration of the Nvidia Ada Lovelace architecture, showcasing the technology that powers the RTX 4070 graphics card |
Key Changes and Performance Impact
The new RTX 4070 variant comes equipped with 20Gbps GDDR6 memory modules, replacing the 21Gbps GDDR6X found in the original model. This adjustment results in a slight reduction in memory bandwidth, dropping from 504GB/s to 480GB/s. However, Nvidia assures consumers that this change will not significantly impact performance.
According to Nvidia's statement, the company has fine-tuned memory timings and latencies to compensate for the slower GDDR6 modules. They claim that the new variant offers similar performance in games and applications compared to its predecessor. Early benchmarks seem to support this assertion, with performance differences generally falling within a 1% margin across various games and applications.
 |
|---|
| A display of Nvidia graphics cards, including the RTX 4070, highlighting the design and performance of the latest variants |
Availability and Pricing
Nvidia partners are expected to begin selling the new RTX 4070 with GDDR6 memory starting in September. Interestingly, Nvidia has positioned this variant at the same price point as the original GDDR6X model, despite the use of potentially less expensive memory components.
Reasons Behind the Change
The introduction of this new variant appears to be driven by supply chain considerations. Nvidia stated that the move is intended to improve supply and availability to meet strong demand for the RTX 4070. This suggests that the company may be facing constraints in sourcing GDDR6X memory, possibly due to reported quality control issues with Micron's GDDR6X modules.
 |
|---|
| A detailed view of the output ports on a graphics card, reflecting the connectivity options important for consumers in the RTX 4070 lineup |
What This Means for Consumers
For potential buyers, the new RTX 4070 with GDDR6 memory presents an interesting option. While it may not offer any performance improvements over the original model, it could potentially alleviate supply issues and maintain the availability of RTX 4070 cards in the market.
However, consumers should be aware of potential price fluctuations. Some rumors suggest that RTX 4070 prices might increase in the near future, possibly due to the ongoing supply challenges. As such, those considering an RTX 4070 purchase may want to keep a close eye on pricing trends in the coming weeks.
In conclusion, while this new variant doesn't revolutionize the RTX 4070 lineup, it demonstrates Nvidia's efforts to maintain a steady supply of graphics cards in a challenging market environment. As always, potential buyers should carefully consider their needs and budget when deciding between this new variant and other available options in the GPU market.
Update: Wednesday August 21 05:24
The new RTX 4070 GDDR6 variant maintains the same GPU (AD104-250/251-A1), CUDA core count (5888), and base/boost clock speeds (1920 MHz / 2475 MHz) as the original model. The graphics card's TDP also remains unchanged at 200W. NVIDIA has made this new variant available through its AIB partners starting immediately, with no Founders Edition planned. This launch strategy allows NVIDIA to potentially increase production and improve availability of the RTX 4070 without compromising on performance, addressing supply constraints in the market. However, the coexistence of both GDDR6 and GDDR6X variants in the market may lead to some confusion among buyers.