"Black technology" triggers a major reshuffle of storage chip stocks: soaring flash memory stocks stand at a crossroads.

Notice that the two-day sell-off of storage chip stocks reveals a divergence in the artificial intelligence investment frenzy. A breakthrough technology promoted by Alphabet Inc. Class C (GOOGL.US) was seen by analysts as potentially suppressing the demand for certain types of storage chips, while impacting others minimally. On Friday, stocks of flash memory manufacturers, used for running artificial intelligence long-term storage devices, continued to decline, including companies like Kioxia Holdings that had seen significant gains in the past few months. At the same time, stocks of leading companies in high-bandwidth memory (HBM), used for NVIDIA Corporation's artificial intelligence accelerators, were stabilizing. Samsung Electronics regained all lost ground in its stock price, and SK Hynix was close to doing the same. Analysts indicate that the market is beginning to realize that Alphabet Inc. Class C's "TurboQuant" technology has increased the efficiency of AI operations, posing a greater threat to the former type of manufacturers (flash memory). Morgan Stanley analyst Tiffany Yeh and others, wrote in a report, "By reducing memory usage and data transfers, TurboQuant significantly improves inference efficiency. However, it does not reduce the demand for core memory such as HBM. In recent months, investors have been pouring into flash memory and storage product manufacturers, believing that with artificial intelligence moving mainstream, demand will explode. Since the end of August, shares of SanDisk have risen by over 1000%, while Kioxia from Japan has risen by over 600%. Their performance has at times outpaced Samsung, SK Hynix, and Micron Technology, Inc. - the traditional leaders in the storage industry who were once the darlings of AI trading early on. When the market focus was still on training ChatGPT and other large language models, the high-profit HBM chips produced by these companies were a popular investment topic. Now, with investors starting to understand Alphabet Inc. Class C's technological breakthrough, flash memory companies are facing a wave of industry-wide sell-offs that began this week. Alphabet Inc. Class C states that this algorithm can reduce the amount of memory required for specific stages of running large language models by at least six times, helping to reduce the overall costs of artificial intelligence. Investors are concerned that this may reduce the demand for memory from large data center operators like Meta (enterprise scale companies). This, in turn, could lower component prices for smartphones and consumer electronics products. Analyst Jake Silverman noted in a report, "As the need to store model weights in GPU memory, HBM demand and DRAM produced by Micron may not be affected. NAND (flash memory) demand may face more profound long-term impacts." Indeed, after experiencing significant gains in the AI frenzy post-ChatGPT era, tech stocks are being closely scrutinized. Investors are cautious about overvalued stocks due to inflation concerns sparked by the Iran war, and are taking profits based on daily news headlines. Ed Gomes, Chief Investment Officer of SGMC Capital Pte, said, "The development of hardware to support the delivery and application development of artificial intelligence technology is a long-term trend that will span years or even decades, rather than days or weeks. The sell-off triggered by TurboQuant is 'short-term noise' and actually provides a very good buying opportunity."