NVIDIA Gears Up to Release China-Exclusive GeForce RTX 5090DD in August, Featuring Reduced Memory Capabilities

NVIDIA is making strategic moves to re-enter the Chinese market with its anticipated GeForce RTX 5090DD, expected to hit retail stores by August. This release comes after facing challenges due to U.S. export restrictions, which led to multiple design revisions.

Rumors suggest that this new model, using the “double-D” naming convention, is slated to launch following the RTX 5090D’s ban. This gives retailers time to clear existing stock before introducing the 5090DD, which boasts altered specifications aimed at meeting compliance.

The RTX 5090DD maintains the same 21,760 core count as its predecessors, the RTX 5090 and 5090D. However, it sees a reduction in memory to 24 GB on a 384-bit bus, with speeds at 28 Gbps, and retains a TDP of 575W. The GPU will utilize the GB202-240 chip, a step down from the GB202-250 and GB202-300 versions found in its counterparts. Moreover, this card will sport a redesigned PCB, reflecting changes in the VRAM and GPU layout.

A key modification with the RTX 5090DD is the decreased VRAM, likely resulting in a performance drop compared to the standard RTX 5090. This suggests that customers in China will receive a more restricted version than what’s available elsewhere. NVIDIA may position the card with a competitive price tag, speculated to be around $1,200 to $1,500, to capture market interest, though this remains unconfirmed.

While the August release date is tentative, changes could occur based on evolving market conditions. On the AI front, there are whispers of a possible “Blackwell” AI chip featuring GDDR7 memory, but details remain scarce.

Overall, NVIDIA seems to be adapting its offerings to navigate regulatory challenges and regain a foothold in the Chinese market.