Software-Defined Storage (SDS) Modernization for HPC Service Infrastructure

Use Case  1

Customer Profile :

One of ASIA's leading high-performance computing (HPC) service providers

A large-scale HPC service provider in ASIA, focusing on gene sequence analysis, faced the challenge of managing and processing vast amounts of unstructured genomic and scientific data.

Solution :
Deployed AVSTOR mSCALE SDS

> AVSTOR mSCALE SDS solution addresses the challenge of managing an ever-growing volume of unstructured data exceeding 10 petabytes (PiB) in raw capacity, ensuring scalability, performance, and operational efficiency.

> S3 tiering feature is automatically to categorize and move the data between different storage tiers—Tier1 (hot, frequently accessed) and Tier2 (warm, rarely accessed)—based on usage patterns.

> Data Passage technology between Files (CIFS/NFS) and S3 enables the user to export S3 bucket as NFS folder.

Software-Defined Storage (SDS) Modernization for Electronic Medical Record (EMR)

Use Case  2

Customer Profile :

One of Taiwan Medical Centers

Large-scale healthcare provider managing over 300 million EMR records.

Solution :
Migrated to AVSTOR mSCALE SDS Object Storage to store EMR data

Original pain point: native NAS limitation. Huge amount of unstructured data is tough challenges for storage scalability, performance, and cost.

> S3 multi-region with single name space

> Data replication across multiple hospital sites. No service interrupt even during site failures.

> Doctor access time reduced from 6 seconds to 2 seconds for full EMR records.

Software-Defined Storage (SDS) Modernization for Healthcare Provider's PACS

Use Case  3

Customer Profile :

One of Taiwan's leading Healthcare Management providers

Large-scale healthcare management provider managing over 100 million DICOM files.

Solution :
Migrated to AVSTOR mSCALE SDS with Data Passage technology

> Data passage technology between File (CIFS/NFS) and S3 enables files in standard NAS (CIFS/NFS) folders to be seamlessly and transparently stored in a backend S3 bucket.

> In the future, for AI and big data analytics, these files stored in the S3 bucket can be directly accessed, searched, and categorized using S3 object tagging.

> Additionally, the files originally uploaded via the NAS (CIFS/NFS) folder into the backend S3 bucket can also be directly accessed using the S3 protocol. This is especially beneficial in scenarios involving a large number of files, as accessing through the S3 protocol can improve speed and efficiency.