Artificial Intelligence | News | Insights | AiThority
Bitcoin
$11,522.68
-57.68
(-0.5%)
Ethereum
$394.25
+6.87
(+1.77%)
Ripple
$0.28
-0
(-0.28%)
Litecoin
$54.27
-0.21
(-0.39%)
EOS
$3.05
+0.05
(+1.7%)
Cardano
$0.14
0
(+1.94%)
Stellar
$0.10
-0
(-1.02%)
NEO
$14.57
+0.3
(+2.1%)
NEM
$0.06
0
(+1.55%)
DigitalCash
$89.67
-0.7
(-0.77%)
Tether
$1.00
-0
(-0.1%)
Binance Coin
$21.35
-0.36
(-1.66%)
QTUM
$2.83
+0.09
(+3.37%)
Verge
$0.01
-0
(-5.88%)
Ontology
$0.81
+0.02
(+2.93%)
ZCash
$81.00
-1.13
(-1.38%)
Steem
$0.23
+0.01
(+2.36%)

WekaIO Places in Top Five of The Virtual Institute’s IO-500 10 Node Challenge

0 3

Beating out massive installations, WekaIO proves performance leadership in HPC with significant cost savings and simplified infrastructure

WekaIO, the innovation leader in high-performance, scalable file storage for data intensive applications, announced its achievement of a second-place ranking on the IO-500 10 Node Challenge list, coming within 5% of the world’s fastest Supercomputer. The IO-500 is a ranking list comparing storage systems that work in tandem with the world’s largest supercomputers. Each system is evaluated by the IO-500 benchmark that measures the storage performance using read/write bandwidth for large files and read/write/listing performance for small files. Storage performance in supercomputers is critical for large-scale simulation, big data analysis, and artificial intelligence (AI). The IO-500 list serves as a resource for companies looking to improve their storage performance, which in turn helps to improve processor performance.

Read More:  Fluor Uses IBM Watson to Deliver Predictive Analytics Capability for Megaprojects

WekaIO ran the IO-500 test on eight Supermicro BigTwin™ enclosures, representing less than half of a storage server rack, and scored 67.79, coming within 5% of the Oak Ridge Summit Supercomputer which ran its system on a 40 rack Spectrum Scale storage cluster. The benchmark suite includes a hero-run of both IOR and mdtest configured in whichever manner possible to maximize performance and establish an upper-bound for performance. Finally, it includes an IOR and mdtest run with highly prescribed parameters to determine a lower-bound.

Read More: The AI Gold Rush: How to Make Money off AI and Machine Learning!

“We couldn’t be more pleased with the results of the 10-node challenge,” said Barbara Murphy, Vice President of Marketing at WekaIO. “This benchmark testing demonstrates that we can deliver performance that competes with the largest supercomputing storage systems, but with a much more cost-efficient and simplified solution. The takeaway is that Matrix delivers more I/O density than Spectrum Scale, or other legacy file systems, which makes us the leader in price/performance value. Now, for the first time, HPC-class storage performance is accessible to the enterprise for deep learning and analytics workloads in a greatly simplified configuration with low CapEx and OpEx overhead.”

“Supermicro’s Resource-Saving BigTwin™ is our highest performing free-air cooled 2U multi-node system supporting 205-watt processors, 24 DIMMs, hot-swap NVMe flash storage, flexible I/O and an incredible 10% power savings compared to typical 1U servers,” said Michael McNerney, Vice President of Marketing at Supermicro. “The high ranking on the IO-500 test is quite impressive because WekaIO was able to achieve greater performance in one-half the rack space with better power and cooling efficiency. This performance density, hot-swappable nodes, and cable-less server design are key requirements for the success of new HPC rack deployments.”

Read More: Calabrio Delivers Powerful Enterprise Scalability and Management Capabilities

“The combination of Mellanox’s RDMA and DPDK accelerated InfiniBand and Ethernet solutions and WekaIO software platform provide leading IO performance as can be seen on the IO500 list,” said Gilad Shainer, Vice President of Marketing at Mellanox Technologies. “By leveraging the smart interconnect acceleration engines, WekaIO can utilize off the shelf servers and SSDs to bring world leading performance at lower total cost of ownership. We look forward to seeing even higher performance results with our new 200G gigabit generation.”

The WekaIO Matrix software is an NVMe-native, resilient file system that runs on commodity servers, delivering the highest bandwidth, lowest latency performance to any InfiniBand or Ethernet enabled GPU or CPU based cluster. Integrated tiering automatically and transparently migrates cold data to any S3 compatible cloud for low cost and limitless scale. The POSIX compliant file system runs more efficiently than NFS and leverages an innovative, customized protocol to deliver file-based semantics. As a result, Matrix exceeds NFS performance in a sharable file system solution.

Read More: Interview with Ben Goertzel, CEO at SingularityNET

Leave A Reply

Your email address will not be published.

This site uses Akismet to reduce spam. Learn how your comment data is processed.