HPC & Machine learning Experts

AWS

Pacific Teck worked together with the AWS in Australia to conduct a POC for BeeGFS on AWS.  BeeGFS is one of the most flexible parallel file systems on the market and together we proved that it would perform well in the cloud as well as on premise.  Our results show that BeeGFS can get the maximum theoretical performance on AWS.  See below for more details on configuration:

Pacific Teck worked together with the AWS in Australia to conduct a POC for BeeGFS on AWS. BeeGFS is one of the most flexible parallel file systems on the market and together we proved that it would perform well in the cloud as well as on premise. Our results show that BeeGFS can get the maximum theoretical performance on AWS.

See below for more details on configuration:
3 Client Benchmark

 

Background

BeeGFS was provisioned in AWS cloud for a specifc Amazon cusstomer. This provisioning method is now replicable for future AWS users.

 

Results

Performance exceeded expectations with 8GB/s throughput. We engineered a shared filesystem capable of over 8 GB/s throughput for the XVA/CVA layer using a shared filesystem called BeeGFS.

 

Pacific Teck’s Role

Pacific Teck worked with ThinkParQ to prepare necessary information regarding provisioning on AWS, and then used the knowledge to implement on-site at the customer’s location. Pacific Teck’s engineer successfully provisioned the file system and worked with AWS to optimize performance. We have benchmarks that are availabe upon request.

 
 

BeeGFS

Learn More about BeeGFS