Continues after advertising
Manufacturers will raise SSD prices once again, seeking to recover from the sequence of quarters with falling profitability due to low demand. According to a recent market trend report, SSDs with Samsung and Western Digital NAND chips, for example, will be between 10% and 15% more expensive, with adjustments for corporate models reaching 25%.
It is important to highlight that, although the current adjustment comes from a consecutive sequence of increases and is the second in 2024, it is only the first in the current fiscal year, which generally starts between April and May.
Also according to TrendForce, SSD manufacturers continue to recover from the losses of recent years in the storage device segment. Therefore, the projection for the coming quarters is a total accumulated adjustment of up to 40%, assuming that companies are able to make use of SSDs even with the new estimated price range.
AI-accelerated update window
However, the new adjustment is not necessarily a surprise. With the arrival of AI PCs, the expectation is that there will be a considerable leap in Artificial Intelligence applications in the domestic segment as early as 2024, with an even greater intensification between 2025 and 2028.
| Segment | Readjustment projection | |
| 1st Quarter 2024 | 2nd Quarter 2024 | |
| Memory Cards (UFS / eMMC) | 25% to 30% | 10% to 15% |
| Enterprise SSDs | 23% to 28% | 20% to 25% |
| Home SSDs | 23% to 28% | 10% to 15% |
| NAND chip wafers | 23% to 28% | 5% to 10% |
| Projected total readjustment | 23% to 28% | 13% to 18% |
A factor that is often underestimated – and fatally undersized – when we talk about AI technologies is storage, and this applies to both SSDs and memories. In addition to the volume of data processed in transit, requiring increasingly larger and faster RAM memories, practically the entire AI workload depends on LLM models, which are extremely large by definition.
Continues after advertising
For comparison purposes, a simple chatbot tool like NVIDIA’s Chat With RTX requires an initial download of 40GB just for the already scaled-down language models. Considering that installation will require even more space, just the initial configuration process of a single AI tool will require about 10% of the total capacity of most pre-built notebook and PC factory storage solutions.
The same goes for virtually every AI application planned for the coming months, making AI PCs extremely dependent on increasingly larger SSDs. In other words, the popularization of AI immediately creates an almost mandatory upgrade window for storage solutions for those who adopt the new technology, and current and future readjustments are clearly taking advantage of the AI hype to drive sales at higher prices .
Continues after advertising
Source: TrendForce
