The AI cloud storage tiering debate: flash is not the only answer
Understand this story
An analysis article argues that in AI training, only the active training phase truly requires flash-level speed, while data set storage, checkpoint writes, and model archiving are better suited to high-capacity, low-cost storage. The article says that if neoclouds hand these workloads to the object storage of hyperscale cloud providers, they will be gradually locked in by data gravity and high egress fees, ultimately losing the entire block of customers beyond GPU contracts. The article advocates a tiered storage architecture: flash for active training, and the rest handled by capacity-oriented storage such as B2 Neo, and mentions that SSD prices rose 257% in less than a year. These are the article author's claims, not verified industry conclusions.
AI overview based on source reports · Updated 10 hr ago
Coverage timeline
Follow the coverage to explore different perspectives.
- Hacker News AIWhy the Best AI Clouds Don't Run on Flash Alone
Neoclouds built around GPUs rely on flash storage for active training, but other lifecycle stages—data ingestion, checkpointing, and storing finished models—need cheaper high-capacity storage, which neoclouds often outsource to hyperscalers whose egress fees create data gravity. The article argues a tiered storage architecture, with flash only where speed is required and high-capacity storage elsewhere, lets neoclouds keep the whole customer account, citing Backblaze's B2 Neo as a white-label option. It notes SSD prices rose 257% in under a year and that Google, Meta, and Microsoft have run tiered storage for years.
Activity for this story
- Current comparable activity
- 8
- Peak comparable activity
- 10Nov 8 04:00
- Change in the past 24 hours
- –
Trends compare the same fully observed participants, which may cover fewer participants than current activity. Move the pointer or tap the chart to inspect hourly activity; use the left and right arrow keys with a keyboard.