RAM prices suddenly dropped

Page Wccftech monitored RAM price fluctuations at many retail systems in the US and discovered for the first time that memory prices showed signs of a sharp decrease after many months of continuous increases.

Specifically at the Amazon system, the Corsair Vengeance DDR5 RAM line with 32 GB capacity and 6400 MHz speed is recorded at a retail price of about 380 USD, while the highest price ever reached 490 USD. Similar reductions also appear on some lower models, for example the 16 GB DDR5-5200 version retails for $220, while last week, this RAM model cost $260.

The site also said similar reductions were recorded at Newegg systems and a number of other systems, but the majority were recorded with Corsair products.

This is a rare time when RAM prices show signs of decreasing after a long period of price increase. Since 2025, consumers and computer hardware manufacturers have been struggling with “memory winter”, when RAM and SSD products are largely collected by data centers to serve the race for AI, causing RAM prices in the consumer market to increase many times. For example, with the RAM models mentioned above, at their cheapest, they cost less than 100 USD.

In addition to adjustment measures from the manufacturer, one of the reasons for this price reduction may come from the TurboQuant algorithm revealed by Google researchers this week. “Memory price drop may be triggered by inventory sell-off after TurboQuant was announced,” Wccftech comment.

Accordingly, the TurboQuant algorithm is capable of compressing the cache of large language models by six times without reducing response quality. The introduction of this algorithm is likened to Google’s “DeepSeek moment”, in terms of optimization capabilities that help AI run smoothly on poor hardware at a cheaper cost.

TurboQuant is still in the testing phase and is expected to be presented in detail at next month’s ICLR 2026 conference. However, as soon as it appeared, it caused the shares of major suppliers such as SK Hynix, Samsung, and Micron to drop sharply.

However, according to some analysts, this may just be a temporary reaction of the market to new news. In the long term, demand for RAM is still high and buyers still cannot access this component at the same price as before.

According to the page TomsguideIn theory, compression algorithms can help reduce the huge demand on RAM, but in reality the opposite can be seen. As the algorithm becomes more effective, large companies and users increasingly need to use it to exploit that efficiency. This also happened when DeepSeek launched last year with low-cost model training capabilities, causing Nvidia’s shares to plummet due to predictions of reduced chip demand. However, after that, chip demand with Nvidia continued to increase, even stronger than before.