Limited Supply of GeForce RTX 5090 Due to Insufficient GPU Availability
Despite the anticipation surrounding the launch of high-end graphics cards, NVIDIA's board partner MSI has announced limited availability of its GeForce RTX 5090 due to insufficient supply of GB202 GPUs from NVIDIA.
Factors Impacting Availability
The overall availability of manufactured products relies on the efficiency of the weakest link in the supply chain. In this case, MSI attributes the limited supply to the number of GPUs provided by NVIDIA.
Graphics processor vendors like MSI, Asus, and Gigabyte purchase NVIDIA processors, which are primarily distributed from Hong Kong. The allocation of processors among vendors depends on factors such as order size, vendor relationships, and ongoing deals.
Insufficient GPU Production
The shortage of RTX 5090s stems not only from demand but also from production constraints. The 750 mm2 size of the GB202 silicon wafer limits the potential yield of fully functional dies, even with high yield rates. TSMC, NVIDIA's chip manufacturer, is unable to meet the demand for these wafers.
Competition for TSMC's N4 Process Node
TSMC's advanced N4 process node is highly sought after. In addition to NVIDIA's Blackwell range, it also produces AMD's Zen 5 chiplets, Strix Halo, and Hawk Point APUs. As a result, there is limited capacity for the production of RTX 50-series chips.
Outlook for RTX 5090 Availability
MSI has confirmed that its RTX 5090 cards will have limited availability upon release. Zotac Korea indicates that RTX 5090s may not be available until early February, and RTX 5080 models have yet to receive a confirmed release date.
UK retailer Overclockers anticipates a "single digit" number of RTX 5090 cards available initially.
It remains unclear whether NVIDIA is intentionally restricting supply to increase revenue or if production limitations are the primary factor affecting availability.