Register here to join Brent Gorda, GM of Intel Corporations High Performance Data division as he presents on the topic of how the Intel® Enterprise Edition for Lustre* software makes parallel storage simpler to manage and more productive for data intensive applications. Big Data has been synonymous with high performance computing for decades, and has become the primary driver fueling new types of applications within many HPC sites. Today the majority of HPC workloads still rely on proven, traditional simulation and modeling techniques. However, the technical and business forces shaping Big Data will lead users to consider and deploy new forms of HPC configurations to unlock insights housed in unimaginably large stores of data. As the boundaries between HPC and big data continue to blur, it’s clear that HPC and enterprise alike are demanding solutions that help solve important problems using new software tools. Intel is the leading innovator, primary developer and global technical support provider for Lustre, the most widely used file system for HPC workloads worldwide.
Note that Lustre clients run on Intel Xeon Phi!
THOUGHTS FROM BRENT GORDA, “LUSTRE PAST AND FUTURE”
Lustre has come a long way since 2010 when Whamcloud, OpenSFS and EOFS were all founded by an impacted and concerned community. Today, the technology is in a position of strength. The technology has undergone 6 major and over 20 maintenance releases in this time-frame, introducing many new features; including long sought after ones like scaling metadata servers (DNE) and hierarchical storage management (HSM). Many customers have, in fact, returned to Lustre, regaining their trust in the name/technology and the open community that shepherds it. Intel is proud to have played a key role in the revitalization of this critical technology. By acquiring Whamcloud, Intel put solid ground under the leading contributor to the tree. Intel has invested and continues to invest in ensuring Lustre will evolve and thrive. We are committed to a healthy and vibrant open source community around Lustre – anchored with OpenSFS and EOFS.
The latest slides from the 2014 China Lustre Users Group meeting
- Keynote: How Lustre* Enables Science – Brent Gorda (Intel)
- Lustre Releases – Peter Jones (Intel)
- Intel Enterpreise Edition for Lustre software – Zhang HongChao (Intel)
- Huawei Lustre – Lou Weidong (Huawei)
- Lustre Performance Evaluation Based on Intel SSD – He Wanqing, Michael Habenstreit (Intel)
- Practice Lustre in HPC Solution and Optimization – Wang Weiwei (Inspur)
- Lustre QoS Based on NRS Policy of Token Bucket Filter – Shuichi Ihara, Li Xi (DataDirect Networks)
- Hadoop on Lustre – Liu Ying (Intel)
- HPC in Cloud with Lustre – Yang Qui (Bingsoft)
- Lustre Network Failure – Liang Zhen (Intel)
- ParaPlus Online Operating Solution for HPC Datacenter – Xinping Huang (Paratera)
- Intel Enterprise Edition of Lustre HSM – Robert Mollard (SGI)
- Lustre Consistency Verification – Fan Yong (Intel)
- Practices on Lustre File-Level RAID – Qi Chen (Jiangnan Institute of Computing Technology)
Here is Brent at the 2013 Lustre Users Group meeting:
Note that both SGI and Cray have Lustre-based hierarchical storage systems

Leave a Reply