Multiple negative factors trigger a correction in Hong Kong storage concept stocks. GigaDevice once dropped nearly 7% during trading.

Financial Associated Press March 27 (Editor Hu Jiarong) Affected by fluctuations in global capital markets, Hong Kong stocks related to memory chips collectively fell today, with Zhaoyi Innovation briefly dropping nearly 7%.

As of the time of writing, Zhaoyi Innovation (03986.HK) is down 3.50%, and Lanke Technology (06809.HK) is down 1.66%.

During the same period, Southern Asset’s 2x Long Samsung Electronics (07747.HK) and Southern Asset’s 2x Long SK Hynix (07709.HK) both fell over 7%.

In terms of news, according to the CFMS|MemoryS 2026 Flash Memory Industry Summit, prices for storage products have experienced a significant increase for three consecutive quarters, and it is expected that the rate of increase will slow down starting in the third quarter of 2026, with some specific product line prices showing differentiation.

For customers, securing storage capacity is more important than locking in prices. Compared to the booming AI market, the consumer market represented by smartphones is entering a period of pain, with costs rising rapidly and sales volumes expected to decline by about 10%, with some smartphone price drops reaching as high as 30%.

At the same time, the overnight U.S. stock market saw a sharp decline in the memory chip sector. Micron Technology saw its stock decline for the sixth consecutive trading day, with a single-day drop of 6.97%, retracting over 23% from its historical high set on March 18; SanDisk, Seagate, and Western Digital fell by 11.02%, 8.33%, and 7.70%, respectively.

Market sentiment is further under pressure, partly due to industry attention triggered by Google’s release of the TurboQuant AI memory compression algorithm on March 26.

It is reported that this algorithm is specifically designed for KV cache scenarios in large language model inference processes, compressing 16bit or 32bit cache data down to 3bit through innovative compression technology, reducing memory usage to 1/6 of its original level, and enabling long-context inference with zero precision loss without requiring model retraining or fine-tuning. Test results show that its 4bit version achieves about 8 times the inference speed on NVIDIA platforms compared to the 32bit baseline, achieving a groundbreaking balance between compression efficiency, precision retention, and inference performance.

This technology features plug-and-play capabilities and has been adapted for mainstream open-source models such as Gemma and Mistral, which can be widely applied in AI servers, edge computing, and mobile devices, significantly reducing the computational and memory costs of deploying large models.

Google officially released the revolutionary TurboQuant AI memory compression algorithm on March 26, 2026. This technology is optimized specifically for large model KV cache scenarios, compressing 16bit or 32bit model cache data down to 3bit, achieving a reduction in memory usage to 1/6 of its original level, while ensuring zero precision loss in long-context inference without the need for retraining or fine-tuning the model.

Test data indicates that its 4bit version achieves 8 times the inference speed on NVIDIA chip platforms compared to the traditional 32bit baseline, successfully achieving a groundbreaking balance between compression ratio, precision loss, and performance.

Institutions say that the decline in AI application deployment costs is expected to stimulate broader demand growth

In response to market concerns, several industry research institutions have pointed out that the core optimization of TurboQuant AI focuses on cache efficiency in the inference stage, without touching on core application scenarios for storage chips such as HBM high-bandwidth memory and model weight storage. From a medium to long-term perspective, the decline in AI application deployment costs is expected to stimulate broader demand growth, thereby expanding incremental space for the storage chip market.

Current stock price fluctuations are mainly influenced by short-term emotional disturbances, and the industry’s fundamentals have not undergone substantial changes. Leading storage manufacturers, with their continuous technological iteration capabilities, capacity layout advantages, and resilience in the supply chain, still have a solid foundation for steady development.