Artificial Intelligence / Deep Learning
The ways in which technology is being applied to real world challenges has grown increasingly more automated and autonomous as computing technology and data algorithms have advanced over the years. Artificial Intelligence (AI), fueled by Deep Learning (DL), is transforming the computing landscape across a broad spectrum of industries that leverage image recognition technology to fuel their needs, including financial technology, healthcare, manufacturing and law enforcement.
NVIDIA GPU processors are the overwhelming choice to address these use cases which need a high level of parallel computing power. Coupled with E8 Storage, customers have a high-performance low latency architecture to enable parallel processing of data in real-time across the AI compute cluster.
- Shared storage performance on par with local NVMe SSD
- Scalable to 1000s of GPUs in a high performance cluster
- Easily integrated into existing high performance GPU server clusters
Understanding the human genome is one of the foundations for research into cures for disease and saving lives. Researchers across the globe are mapping individual genomes, as well as comparing large groups of genomes as part of their primary and secondary research. Analyzing these datasets can require 500TB or more of storage and take days to complete without access to high performance compute and storage.
E8 Storage tackles this performance challenge head on, seamlessly integrating into high performance clusters built on GPU or FPGA processors and 100Gb Ethernet (RoCE) or Infiniband networks. With proven performance, E8 Storage has demonstrated up to 100X improvement in genome processing time over competing solutions.
“The E8 Storage appliance behaved flawlessly and the integration with InfiniBand was simpler than expected” – Oxford Big Data Institute
- Massive acceleration in genome processing time
- Scalable to 1000s of GPUs / FPGAs in a high performance cluster
- Management APIs for easy integration into automated job scheduling
Accelerating IBM Spectrum Scale
The pace of business is transforming today’s data centers, with data- and transaction-intensive applications driving the need to extract value from data quickly and efficiently. These applications need high performance and low latency in their compute and storage architectures, and IT departments look to the latest in storage technologies to deliver on those needs.
E8 Storage systems were created with these applications in mind, incorporating NVMe SSDs into an architecture designed specifically to extract the full performance of flash. When paired with a high performance clustered file system such as IBM Spectrum Scale, the result is a file storage architecture that delivers both high performance and massive scalability for data hungry applications.
- Best performance and lowest latency Spectrum Scale storage
- Easily add E8 Storage to existing Spectrum Scale infrastructure
- Shared read / write volumes for parallel processing
Financial firms need to read and process data fast in order to make the right business decisions that lead to higher profits. However, data is only relevant for a limited period of time and there is a tight time window for making the right decision. This poses a major problem in the world of high speed analytics and big data management. Not enough data, and/or not enough time to process it can result in a decision that is less accurate and that might even cost money. Investing to speed up networks, servers and storage can pay off handsomely in the form of smarter and quicker decisions.
E8 Storage delivers a highly reliable and scalable storage solution that enables customers to process data at the speed of local SSD. Customers can reduce their storage costs without compromising on their performance requirements, while ensuring that their data analytics needs are met in real-time.
- Consolidate storage and eliminate data replication with shared storage
- Scalability – easily add storage capacity as the dataset grows
- Maintain high performance with very low latency
- High reliability with RAID 6 data protection and no single point of failure
Web and Hosted Services
Service providers have increasingly turned to web based services as the delivery platform for their mission critical customer applications. Many of these hosted services need very high performance and low latency in order to meet service level requirements to end users, in particular travel services, payment processing and the like. Fast response times are essential and not meeting SLA’s directly affects the bottom line.
E8 Storage delivers consistently low latency and high performance to enable service providers to guarantee their services to end customers. With a high availability architecture, E8 Storage delivers shared storage at the speed of local SSDs at a fraction of the cost of competing solutions.
- Consistently strong performance and low latency, as low as 100µs
- Supports shared storage access for up to 126 clustered servers
- Easily scalable to keep up with business growth
- High reliability with RAID 6 data protection and no single point of failure
Time sensitive data queries and data analytics for large data warehouse stores often require faster processing than even large server clusters can provide when they need to work together on a large dataset. Accelerating these large scale data queries means moving to a parallel processing architecture at the server level, such as clustered databases like Oracle RAC. These clustered applications need shared read / write access to data in order to fully realize the performance and resiliency benefits that scale-out application clusters provide.
E8 Storage is the first NVMe based storage system to work natively with clustered databases and filesystems to deliver high performance read / write access to shared data. E8 Storage’s unique architecture provides scalable, high performance storage without sacrificing on reliability or data protection.
- Consolidate storage and eliminate data replication for shared data
- Improved resiliency for both application servers and shared accelerated storage
- Reduce overall processing time for workloads such as batch analytics
- Ensure high performance and low latency for all shared workloads
Top investment banks and hedge funds have complex and data-intensive applications that require very high performance and low latency in order to keep abreast of real-time market trends. Every second spent waiting on data prevents these financial firms from responding to rapid market shifts, costing revenue. Financial firms generate billions and trillions of data records over time, resulting in an explosive escalation in data volumes. These customers need a high performance, scalable storage solution for their mission critical trading applications.
E8 Storage is the shared storage solution of choice for customers needing a scalable, high performance storage. It’s unique architecture built on NVMe leverages the latest in high performance storage without sacrificing on reliability, availability and scalability.
- Latency as low as 100µs for real-time performance
- Deploy one or many appliances for petabyte scalability
- Supports up to 126 servers in a database cluster
High Frequency Trading Platforms
The ultimate test of storage performance comes from high frequency trading platforms such as regional stock exchanges. Millions of trades daily need to be tracked and reported in sub-millisecond end to end speed, requiring in-memory database cache levels of performance.
Read how E8 Storage is working with Intel and the Tel Aviv Stock Exchange to test out the latest in flash storage innovations.
- Native support for Storage Class Memory SSDs
- Ultra low latency, well under 100µs for high frequency trading
- Massive TCO improvement over storage resident memory cache