Old technology resurgence? Memory economics may force NVIDIA to restart production of older GPU models to fill market gaps
In the face of VRAM shortages, NVIDIA is optimizing product supply based on the "revenue per GB of VRAM" model, prioritizing the production capacity of 8GB and flagship models. This move has led to a contraction in the supply of mid-range models with large VRAM, such as the RTX 5060 Ti 16GB. To fill the market gap, NVIDIA may restart the production of older GPUs like the RTX 3060 and use older VRAM such as GDDR5 to maintain the operation of its consumer product line, while the strategic priority of the consumer business has been adjusted in light of AI demand
As the supply of GPU memory continues to be tight, NVIDIA is adjusting its product strategy based on the "revenue per GB of memory" model, prioritizing the supply of high-profit models. This may lead to the marginalization of some mid-range products with large memory, which in turn may force it to reactivate old GPU production lines to fill the market supply gap.
Gigabyte CEO Eddie Lin stated that NVIDIA is prioritizing the supply of the RTX 5060 Ti (8GB) and flagship graphics cards, while the production capacity of mid-range large memory models like the RTX 5060 Ti 16GB may be compressed.
Industry analysis indicates that if the memory shortage persists, NVIDIA's restart of some older GPU production will become a feasible strategy to maintain coverage in various market segments. This shift will not only reshape the product mix available to consumers but also may lead to a new supply pattern in the mid-range graphics card market characterized by "high-profit orientation and tighter memory configuration."
Profit Margin Model Reshapes Product Priorities
Eddie Lin further explained the internal logic of NVIDIA's supply strategy. He pointed out that NVIDIA needs to optimize revenue by adjusting the supply structure while covering all product tiers. Specifically, the company can focus on ensuring supply for the 1st, 3rd, and 5th tiers among five product levels, while compressing the share of the 2nd and 4th tiers, as the latter contribute lower unit memory revenue.
He illustrated this with a specific model: for example, a GPU priced at $300 (like the RTX 5060) can contribute about $35 in revenue per GB of memory; the $400 8GB version (like the RTX 5060 Ti) increases to about $50 per GB. In contrast, a $500 16GB graphics card has a revenue contribution of only about $32 per GB, clearly at a profit disadvantage.
According to analysis from Tom's Hardware, the products currently contributing most significantly to NVIDIA's profit margins are mainly concentrated in the mid-to-high-end models with 8GB of memory (like the RTX 5060 series) and the ultra-high-end RTX 5090 and professional-grade graphics cards. This profit structure will directly transmit to the retail end, determining the actual supply ratio of different models in the market.
Restarting Old Products and Adjusting Supply Strategy
Based on this profit model, recent rumors about NVIDIA possibly restarting RTX 3060 production while raising the suggested retail price of the RTX 5090 seem logical. NVIDIA is expected to fill the supply gap in the mid-to-low-end market with older GPUs, thereby concentrating new generation memory resources on higher-margin new generation product lines.
At the same time, using earlier memory modules like GDDR5 has also become a possible option. This technology is not affected by the current high demand in the AI field and can serve as a "placeholder" solution to maintain the basic operation of consumer-grade product lines during the general memory shortage phase.
Market observers noted that although NVIDIA previously clarified that it would not discontinue the RTX 5070 Ti 16GB version, the actual arrival situation in retail channels will be a key signal for judging the true supply priority of the products. Some configured models may gradually fade out of the mainstream market without any official announcement
Consumer Business Yielding to AI Demand
The fundamental reason for the current shortage of video memory lies in the surge in demand for general-purpose DRAM due to data center construction. Against the backdrop of the continuous booming AI business, NVIDIA's consumer GPU business has clearly become a strategic secondary position.
Since the news of the delay of the RTX 50 SUPER series, doubts about its consumer product strategy have persisted in the market. Recent adjustments in the supply chain further confirm that the priority of the consumer business in the company's overall resource allocation is declining.
It is noteworthy that NVIDIA typically enhances the performance of older generation products by integrating new technologies such as DLSS, providing a technical marketing basis for older GPUs to return to the retail market. It is expected that in the coming months, the market may see a gradual increase in the supply of more older generation graphics card models
