2027 memory capacity is reportedly sold out (ign.com)

Article summary

The 2027 memory capacity of major manufacturers Samsung, SK Hynix, and Micron has been sold out, primarily to AI companies, which may lead to increased retail prices and scarcity. This has significant implications for the tech industry, including higher costs for consumers and potential delays in product releases. The demand for memory is also affecting the price of NAND storage, with some SSDs becoming more expensive. The situation is expected to continue into 2028.

Main themes

• Memory shortage
• AI demand
• Tech industry impact
• Price increases
• NAND storage
• Scalability

What commenters say

• The AI bubble may eventually burst, which could lead to a decrease in demand for memory and a subsequent decrease in prices.
• Some argue that the demand for AI is sustainable and will continue to drive the need for memory, while others believe that the costs of AI may become prohibitively expensive for consumers.
• The concept of scalping is not directly applicable to the current memory shortage, as companies are purchasing memory for actual use rather than resale.
• The use of AI is becoming increasingly viable for non-technical individuals to create software and apps, potentially disrupting the traditional software development industry.
• Moore's Law is still applicable in terms of transistor count, but Dennard scaling is what has actually driven performance increases in the past.
• The memory shortage has significant implications for the tech industry, including the potential for increased costs and decreased performance.
• Some commenters believe that the current trajectory of AI improvement will continue, leading to significant advancements in the near future.
• Others argue that the costs of AI may not be sustainable for consumers, and that the industry may need to adapt to find more affordable solutions.