RTX 4070: New SKU with GDDR6 memory in response to GDDR6X shortage

NVIDIA is about to launch a new version of its GeForce RTX 4070 graphics card, reports MEGAsizeGPU, which will now be equipped with GDDR6 memory instead of the previously used GDDR6X memory. This change is not unexpected, as there have been increasing reports of supply bottlenecks and quality problems in the production of GDDR6X memory modules in recent weeks. The decision to switch to GDDR6 could help NVIDIA circumvent supply chain issues and ensure the availability of its graphics cards.

The switch from GDDR6X to GDDR6 is the result of a number of challenges that NVIDIA has faced in recent months. In particular, quality assurance at one of the main suppliers for GDDR6X, Micron, appears to have affected a batch of memory modules, resulting in a shortage of this memory type. Combined with the already strained production capacity for the “Ada” architecture, which includes the RTX 4070, NVIDIA has been forced to look for alternative ways to meet demand for their graphics cards. GDDR6, although slower than GDDR6X, is a reliable and widely used memory solution. By using GDDR6 memory, NVIDIA could not only improve the availability of the RTX 4070, but also potentially reduce production costs, which could translate into lower prices for end users.

The new RTX 4070 variant with GDDR6 memory should be technically largely identical to the current models. The card is still based on the AD104-250-A1 GPU, which has 5888 CUDA cores, 184 texture units (TMUs) and 64 raster operation units (ROPs). The GPU clock rate also remains unchanged at a base clock rate of 1920 MHz and a boost clock rate of 2475 MHz. However, the decisive change concerns the memory: instead of GDDR6X with a speed of 21 Gbps, the new card will be equipped with GDDR6, which is expected to operate at 20 Gbps. This change will lead to a reduction in memory bandwidth from 504.2 GB/s to 480 GB/s. While this reduction of around 5% may seem small on paper, in practice certain applications, particularly modern games that benefit greatly from memory bandwidth, may show slightly reduced performance. In particular, the Ada Lovelace architecture on which the RTX 4070 is based is designed to benefit from high memory bandwidth.

However, NVIDIA is trying to cushion these potential performance losses with other measures. The RTX 4070 is equipped with a generous 48 MB L2 cache, which, together with advanced memory compression techniques, should help to maximize the efficiency of the memory. There is also the option to overclock the GDDR6 modules to 21 Gbps, which could bring the memory bandwidth back up to the level of the GDDR6X models. However, this overclocking could affect the long-term stability of the card, which could be an important factor for users who use their hardware intensively and over long periods of time. The decision to introduce an RTX 4070 variant with GDDR6 memory could prove to be a smart move to meet current market demands. As GDDR6 memory is available in sufficient quantities, NVIDIA could quickly ramp up production of this new variant, easing pressure on the supply chain. This would be an important step, especially in view of the still tense situation on the global semiconductor market.

In terms of pricing, the new RTX 4070 GDDR6 variant is expected to be slightly cheaper than the current models with GDDR6X memory. While the standard models of the RTX 4070 are currently priced at around 549 US dollars, the GDDR6 variant could be priced at around 499 US dollars. This could make the card particularly attractive for price-conscious buyers who are prepared to accept the slightly lower memory bandwidth. NVIDIA has not yet announced any official information on the availability of the new RTX 4070 GDDR6, but more details are expected to follow in the coming weeks. If the market launch of this variant proves successful, NVIDIA could also extend this strategy to other models in the RTX 40 series in order to further stabilize the supply situation.

The switch from GDDR6X to GDDR6 in the GeForce RTX 4070 shows how flexibly NVIDIA can react to challenges in the supply chain. While the switch to a slower memory type may seem like a step backwards at first glance, it could prove to be a necessary step to ensure the availability of graphics cards while reducing costs for end users. For gamers and PC enthusiasts, it will now be exciting to see how the performance of the new RTX 4070 GDDR6 performs in practice and whether the change will actually make a noticeable difference to the gaming experience. The coming weeks and months will show how the market reacts to this new variant and whether NVIDIA has found a way to successfully overcome the current challenges. Given the ongoing shortages and high demand for powerful graphics cards, the new RTX 4070 GDDR6 could play an important role in restoring the balance between supply and demand.

Source: MEGAsizeGPU