🤖 AI Summary
The ongoing GPU supply crisis shows no signs of resolution as demand from AI data centers continues to outstrip production capabilities. Companies are redirecting resources toward high-bandwidth memory (HBM) chips, essential for AI model training and operation, while also contending with a general shift from DDR4 to DDR5 memory technologies. This has resulted in a dramatic increase in GPU prices, with Nvidia's GeForce RTX 5090 retailing at nearly $5,000, far exceeding its initial launch price. The phenomenon, dubbed "chipflation," affects not only professional consumers but also impacts the broader electronics market, driving up prices for software, gaming consoles, and laptops, and leading to shrinking PC shipments.
Key industry leaders from Micron and SK Hynix anticipate continued constraints, predicting that supply will remain tight until at least 2030. These manufacturers are prioritizing datacenter supplies over consumer GPUs, with HBM expected to account for 30-40% of DRAM production by 2029. The concentration of DRAM manufacturing among just three major companies further complicates recovery efforts. As the AI sector's appetite for memory grows, the ramifications are being felt across the consumer tech landscape, leaving many to question when stability will return to the market.
Loading comments...
login to comment
loading comments...
no comments yet