NVIDIA, a key player in the graphics card industry, is reportedly facing a significant challenge due to a memory shortage. This supply issue has led the company to inform its board partners that they must source their own GDDR memory chips, marking a shift in how NVIDIA operates with its partners.
NVIDIA Directs Board Partners to Obtain GDDR Memory Independently
The tech giant usually partners with companies like Samsung, Micron, and SK Hynix for its VRAM chips. However, with the booming demand for AI technology, these suppliers are struggling to keep up, impacting NVIDIA’s access to essential memory components. According to well-known leaker “Golden Pig Upgrade,” the scarcity has resulted in NVIDIA no longer providing VRAM chips alongside GPU dies to its partners.
This shift could pose significant challenges, especially for smaller vendors lacking extensive sourcing connections, though larger manufacturers may find it less burdensome. The current memory shortage affects not just GDDR6 and GDDR7 used in the latest GPUs from AMD and NVIDIA but also general system DRAM like DDR4 and DDR5.
Memory Shortage Impact and Future Projections
The shortage in memory supply shows no signs of easing, with recent warnings from industry leaders about further complications. Despite Samsung beginning mass production of 24 GB GDDR7 modules, these are unlikely to see rapid adoption due to ongoing constraints. The RTX 50 Super series, for example, is rumored to be postponed to Q3 2026 because of these issues.
The situation highlights the ongoing challenges in the tech supply chain, with reports suggesting this problem could persist, affecting pricing and availability across the industry.