← All stories
● Covered by 1 source · 1 reportMedium impact1 neutral

Sandisk and SK hynix Release High Bandwidth Flash (HBF) Specification via OCP

🔄 Updated 1d ago
New to BrevFeed? We gather this story from every outlet covering it into one summary — ranked by real-world impact, not just the latest headline — so you never miss what matters. What is BrevFeed? →

Key points

  • Sandisk and SK hynix released the HBF specification via OCP.
  • HBF combines 3D NAND non-volatility with HBM performance.
  • Initial HBF packages offer up to 512GB capacity.
  • Bandwidth ranges from 0.4 TB/s to 3.0 TB/s.

Introduction of High Bandwidth Flash (HBF)

Sandisk and SK hynix have formally introduced the High Bandwidth Flash (HBF) specification, a jointly developed storage technology. This specification was released through the Open Compute Project (OCP), establishing it as an open standard rather than a proprietary interface. HBF aims to merge the non-volatility of 3D NAND with the performance characteristics of High Bandwidth Memory (HBM), targeting applications such as AI inference systems.

Specification Details and Performance

The initial HBF specification defines packages with capacities up to 512GB, utilizing either 8-Hi or 16-Hi NAND die stacks. These are specialized devices with fast interfaces, referred to by Sandisk as HBF core dies. Performance is categorized into three bandwidth grades, ranging from approximately 0.4 TB/s to 3.0 TB/s. The highest performance tier of HBF (3 TB/s) is designed to exceed the memory bandwidth of a single HBM4 stack (2 TB/s), though HBM4 is expected to maintain a latency advantage.

Integration and Roadmap

SK hynix states that HBF uses the Universal Chiplet Interconnect Express (UCIe) standard for integration with heterogeneous computing platforms. Sandisk indicates HBF will adopt the 'xPU-HBF' interface, which may be its interpretation of UCIe. The broad performance range suggests a multi-year roadmap for HBF, encompassing multiple implementations and generations. The specification also covers electrical and interface characteristics, packaging, reliability guidelines, and software I/O requirements for stacked HBF devices.

Technical Implementation

Achieving 400 GB/s of bandwidth from a single 512GB HBF package requires significant engineering. Sandisk previously outlined plans to use 16 HBF core dies, each with numerous arrays for concurrent access via dedicated read/write paths. Alternatively, over 400 GB/s per package can be achieved using a single UCIe interface operating at up to 64 GT/s across 64 lanes, which implies a complex HBF base die. Sandisk and SK hynix position HBF as a new memory tier.

✨ This summary was generated by AI from the outlets' reporting listed below. It is not independently verified and may contain errors — check the original sources. How BrevFeed works →

The daily brief

One email each morning: the day's tech stories, clustered across outlets and summarized. No account needed.

One email a day. Unsubscribe in one click, any time.

Today's brief

Spend a few minutes, get the whole day. Every topic's top stories in one hands-free rundown — listen, watch, or read the transcript.

~7 min · 6 stories · Aug 15

▶ Play today's brief Listen on Spotify

New every morning, and the back catalogue is archived by date.

Reporting from

Sandisk and SK hynix have introduced the High Bandwidth Flash (HBF) specification through the Open Compute Project (OCP), combining 3D NAND non-volatility with HBM-like performance for AI inference systems. This open standard defines HBF packages with capacities up to 512GB and bandwidths ranging from 0.4 TB/s to 3.0 TB/s, aiming to provide a new memory tier for heterogeneous computing platforms.