How much does the new Dynamic Tier cost per GB monthly?

The new Dynamic Tier costs $0.06 per GB monthly for low-latency performance. This single SKU eliminates hidden complexity by serving data directly from persistent disk rather than relying on object-based caching mechanisms.

What maximum throughput does Rapid Bucket deliver for AI training workloads?

Rapid Bucket delivers more than 15 TB of bandwidth within a single zonal bucket. This extreme throughput supports 20 million requests per second with sub-millisecond latency for frequent I/O patterns.

How much does Rapid Bucket reduce GPU blocked time during model training?

Rapid Bucket reduces GPU blocked time by 50% compared to standard regional storage options. This significant improvement increases accelerator utilization for multi-modal training while enabling 2.5x faster data loading speeds.

What aggregate read throughput does Rapid Cache provide for bursty workloads?

Rapid Cache accelerates bandwidth with an aggregate read throughput of 2.5 TB per second for existing buckets. This feature requires no code changes and supports ingest-on-write for immediate cache hits.

How does Managed Lustre throughput compare to other hyperscaler offerings?

Managed Lustre delivers up to 10 TB of throughput, which is 4–20x higher than competitor offerings. This performance stems from new C4NX VMs and Hyperdisk Exapools architecture designed for massive scale.

How much does B2 Neo cost per GB compared to competitors?

B2 Neo operates at just $0.006 per GB for underlying storage technology. Competitors like CoreWeave charge significantly higher rates of $0.03 per GB for their standard object storage tiers.

What throughput speed prevents AI training pipelines from stalling?

The service supports up to 1 Tb of throughput to prevent pipeline starvation during massive data transfers. This high bandwidth ensures rapid checkpoint commits without queuing delays for large artifacts.

How quickly can providers launch storage versus building it internally?

Providers can launch a branded storage service in weeks instead of waiting years. Building a custom minimum viable object storage solution typically consumes eighteen to twenty-four months of engineering time.

Why do hyperscaler egress fees destroy margins for GPU farms?

Hyperscalers charge high egress fees that destroy margins, whereas B2 Neo offers zero transaction fees. Competitors often charge $0.03 per GB, creating a structural cost disadvantage for high-volume AI workloads.

What market growth drives the need for white label storage?

The neocloud market is expanding rapidly from $35.22 billion currently toward $236.53 billion by 2031. This explosive growth forces GPU farms to bypass storage construction distractions to scale faster.

Time on StorageNews

Rapid cloud storage fixes AI training bottlenecks

Wed, 22 Apr 2026 00:00:00 +0000

Google's new Rapid Bucket delivers 15 TB/s bandwidth and cuts GPU blocked time by 50%, directly answering the latency crisis in AI training.

Whitelabel storage saves neoclouds 18 months

Mon, 23 Feb 2026 00:00:00 +0000

Backblaze claims B2 Neo cuts storage build time from 24 months to weeks for GPU farms chasing the $35.22 billion neocloud market.