75%
Probable
It appears that NVIDIA isn't able to receive sufficient memory for its inventory, which is why it reportedly won't supply the memory chips to its board partners.
NVIDIA Reportedly Tells AIBs to Source Their Own GDDR Memory as It Stops Supplying VRAM Chips With GPUs
NVIDIA gets its VRAM chips from suppliers such as Samsung, Micron, and SK Hynix, but those memory giants "apparently" also have to fulfill the memory demand due to this "AI boom". This has worsened the memory availability to the point where NVIDIA may have stopped getting sufficient memory chips from those companies. This is according to a prominent leaker, "Golden Pig Upgrade", who has reported that NVIDIA has stopped bundling VRAM chips with GPU dies.
For those who aren't aware, NVIDIA ships GPU dies and VRAM chips to its board partners, while the board partners take care of the rest. However, NVIDIA has apparently told its board partners to source the VRAM chips on their own. That said, for major GPU manufacturers, it might not be a hassle, but for many small vendors, sourcing VRAM chips on their own without any good connections will be a huge problem.
Remember that the short memory supply isn't just affecting regular system DRAM, such as DDR4 and DDR5, but also video memory, including GDDR6 and GDDR7. Both these video memories are utilized in the latest AMD Radeon RX 9000 and NVIDIA RTX 50 series GPUs. While NVIDIA has reportedly left its board partners to buy their own video memory, it isn't clear whether AMD has such plans.
Still, keep in mind that this isn't an official report, but considering the good track of the leaker, it's possible, especially when there has been news regarding inflated memory prices.
Keep in mind that the memory availability might worsen even further, as MAINGEAR's CEO recently warned about the same. Even though Samsung has now started mass producing the 3 GB GDDR7 28 Gbps memory modules, it isn't going to be adopted any time soon, as 3 GB chips are likely to end up on the RTX 50 Super series, which is rumored to be delayed to Q3 2026 due to the same memory shortage issue. Meanwhile, NVIDIA has been reaffirming that its GeForce RTX 50 GPUs remain in production.
News Sources: Weibo, via Videocardz
About the author: Sarfraz Khan is a hardware reporter with a focus on PC components and the builder community. With years of experience writing about PC hardware and laptops, his work has been featured on several reputable technology publications. Sarfraz's hands-on experience is demonstrated through his first-person accounts of using and comparing different hardware configurations, providing practical and relatable insights for everyday users. His technical analysis is respected by peers in the enthusiast community and has been cited by specialized hardware sites such as Germany's Igor's Lab.
Follow Wccftech on Google to get more of our news coverage in your feeds.
