Italian tech website Chips ‘n Bits recently revealed that AMD is planning to prioritize the production of FPGA and GPGPU chips over its next Radeon gaming GPUs (RX 8000) at TSMC. This allocation of production capacity to AI chips rather than graphics cards has sparked concerns about GPU supply for gaming. Given the previous challenges faced by gamers during the crypto craze and the pandemic, the news has left many wondering about the future availability and affordability of graphics cards.
During the pandemic, graphics card prices skyrocketed to absurd levels, creating a significant barrier for gamers. Although the market has stabilized to some extent, many current cards still struggle to offer a reasonable value proposition. AMD’s recent release of the Radeon RX 7800 XT has provided some relief, but the situation remains far from ideal.
Reports suggest that Nvidia has already shifted its focus away from the production of high-end models of its current RTX 40 GPUs and redirected it toward producing more AI chips such as the H100. This move aligns with the company’s strategy to tap into the lucrative AI inferencing market. With limited production capacity at TSMC, Nvidia faces a challenge in meeting the demand for AI chips while simultaneously catering to the gaming GPU market.
Contrary to concerns about GPU supply, Nvidia’s current RTX 40 GPUs are widely available in the market. From the entry-level RTX 4060 to the powerful RTX 4090, gamers can easily find these cards for purchase. While some may argue that the high pricing is linked to the AI boom, there is little evidence to suggest that AI companies are bulk-buying gaming graphics cards.
At present, there is no significant mismatch between supply and demand for gaming graphics cards. The AI boom has not yet caused a distortion in the market. However, if the AI trend continues to grow, there will likely be a shift in production priorities. Nvidia’s roadmap already indicates a delay in the release of its next-generation RTX 50 GPUs, a move that could allow more focus on AI chips by leveraging TSMC’s upcoming 3N node capacity.
Similarly, AMD may face a similar situation as it seeks to establish its presence in the AI market with the Instinct MI300 AI chip. If the chip gains traction, it will encounter the same market forces as Nvidia. However, any temporary impact on gaming GPU availability caused by the AI boom can be mitigated. Graphics cards can temporarily lag behind AI chips by one node or hold back for a cycle, allowing for the gradual release of new and faster GPUs.
To alleviate concerns, it’s worth remembering Nvidia’s Maxwell generation of GPUs released in 2014. Despite being based on the same 28nm production technology as its predecessor, the Kepler GPUs, Maxwell showcased significant performance gains. This example demonstrates that even if gaming GPUs need to fall back one node to prioritize more profitable chips, it does not necessarily mean a halt in technological advancements.
The ongoing AI boom should not be seen as a precursor to a resurgence of exorbitant GPU pricing witnessed during the pandemic. While graphics cards may continue to be expensive due to various factors, such as supply chain disruptions or increased demand, the AI boom itself does not directly cause inflated GPU prices. Therefore, although the current situation is less than ideal for gamers, there is no need to panic.
The allocation of production capacity to AI chips is likely to impact the availability of gaming GPUs in the short term. However, it is important to recognize that this is not an insurmountable challenge. Both Nvidia and AMD have the ability to adjust their strategies and prioritize different product lines without sacrificing the progress of gaming GPU technology. While gamers may face some uncertainties, the industry’s ability to adapt and innovate suggests that there is hope for a brighter future where the AI boom and gaming GPU supply can coexist harmoniously.
Leave a Reply