NVIDIA GeForce RTX 5060 Upgrades to GB205 GPU, Marks Shift in Design and Manufacturing
The NVIDIA GeForce RTX 5060 has received a significant hardware upgrade, with new variants now using the larger GB205 GPU die instead of the previous GB206 silicon. The GB205 die, originally designed for the RTX 5070, is being repurposed for the RTX 5060 through core disabling and design adjustments, according to sources.
GB205 vs. GB206: Die Differences and Design Implications
The GB205 GPU die, used in the RTX 5070, features 6,400 CUDA cores, compared to the 4,608 cores in the GB206 die found in earlier RTX 5060 models. However, NVIDIA disables a portion of the GB205’s cores to align with the RTX 5060’s 3,840 CUDA core specification, as noted by MSI and Yeston. This practice is common in the industry, allowing manufacturers to utilize partially defective chips for lower-tier SKUs.
Physically, the GB205 die uses a rectangular package with a distinct mounting hole layout, unlike the square design of the GB206. This difference is visible in MSI’s new RTX 5060 “V1” models, which match the RTX 5070’s mounting pattern.
MSI’s V1 Models: AIB Partners’ Strategy and Market Impact
MSI has introduced four new RTX 5060 models under the “V1” designation, explicitly using the GB205 die. These include the RTX 5060 V1 Gaming, V1 Gaming OC, V1 Ventus 2X, and V1 Ventus 2X OC. The V1 models differ from their predecessors primarily in the GPU’s rectangular mounting hole layout, as reported by Uniko’s Hardware. MSI also noted subtle changes in power consumption and clock speeds, such as a 38MHz reduction in boost frequency and a 5W drop in TDP for the V1 Gaming model.
The V1 naming convention is unusual, as most AIB (Add-in-Board) partners typically retain the same model name across die revisions. MSI’s decision to highlight the GB205 upgrade underscores the significance of the change, though the company has not yet disclosed pricing or availability details.
Technical Adjustments and Industry Context
The GB205-based RTX 5060 variants maintain the same 128-bit memory bus and 8GB GDDR7 configuration as their GB206 counterparts.
The shift to GB205 reflects broader industry trends of repurposing higher-end silicon for lower-tier GPUs. NVIDIA and its partners frequently use this strategy to minimize waste and maximize yield. This approach reduces manufacturing costs but may not significantly impact end-user performance, as the RTX 5060’s specifications remain unchanged.
Broader Implications for Consumers and the Market
The GB205 upgrade does not alter the RTX 5060’s advertised performance metrics, such as 3,840 CUDA cores or 128-bit memory bus. However, the use of a larger die and potential cooling improvements—such as the RTX 5070-class cooler in Yeston’s model—could offer marginal benefits in thermal management. Consumers should note that the 16-pin power connector in Yeston’s variant is non-standard, though it does not affect performance.
For the broader market, the shift highlights NVIDIA’s ongoing efforts to optimize production amid supply chain and cost challenges. The RTX 5060’s GB205 iteration also aligns with the company’s strategy of extending product lifecycles through hardware revisions, rather than launching entirely new series. This approach may delay the anticipated RTX 50 SUPER launch, which has been postponed due to GDDR7 pricing concerns, as noted in one source.
