Speed and streamline hybrid and on-premises big data ecosystems with the world’s first 1U heterogeneous compute accelerator
The Ryft ONE heterogeneous compute accelerator eliminates data latency bottlenecks throughout any big data ecosystem. It’s the first and only accelerator to make data analytics fast and simple by combining heterogeneous FPGA/x86 compute, patented SSD-based storage, a library of analytics algorithms and an open API in an easy-to-use, 1U form factor.
The Ryft ONE Open API and Linux front-end allow the Ryft ONE to seamlessly integrate into any hybrid or on-premises environment to streamline infrastructures and speed analytics by 100X or more.
For the first time, you’ll have the speed, scalability, simplicity and security you need for real-time intelligence applications like genomic research, fraud prevention, cybersecurity and financial compliance.
Competitive benchmarks prove the Ryft ONE outperforms the fastest, traditional CPU-based clusters to speed insights into massive quantities of any type or format of data with no indexing, transformation or curation required. Users also achieve 70% lower TCO (total cost of ownership) by reducing maintenance and operational costs such as energy use, rack space, top-of-rack routers, and cabling.
1 TB Reddit Content (2012-2016) Hamming Distance 2
Time to insights using 100 executors in hours
*Benchmark comparisons of Ryft ONE against Apache Spark running on a cluster of AWS EC2 – c3.8xlarge “Compute Optimized” 2U servers that require 1100 Watts each.
Get the answer
1 TB JSON Reddit Content (2012-2016) Hamming Distance 2
Time to insights using 15 executors in hours
Find the answer
1 TB Genome Set For 100 Humans Edit Distance 4
Time to insights using 200 executors in hours
*Benchmark tests using server instances on Amazon Web Services. Floor and electric estimates determined using a colocation data center pricing calculator: https://www.thecloudcalculator.com. Average price per U.S. KW/hour in 2012 was $0.12.
Reduce Floor Space by
Reduce Power & Cooling Costs by
The era of painful tradeoffs in big data is over. Ryft’s no-compromises solution doesn’t force buyers to choose between speed and simplicity or other critical objectives.
World's only heterogeneous FPGA/x86 accelerator in a 1U device
Ubuntu Linux front end and Open API
Flexible deployment in the cloud, datacenter and at the edge
Built-in Library of Data Analytics Algorithms
Up to 48 TB of all-Flash, secure self-encrypting SSD storage
1/10/40 GbE and InfiniBand network interfaces
Real-time analytics performance with without indexing, transformation or curation
Flexible data mining interfaces to explore complex data
Encryption and decompression with no performance loss
Native integration with Apache Spark, ODBC/JDBC and more
NAS protocols Including NFSv4
Easy-to-manage 100+ node Ryft CLUSTER delivers 200+ GB/second throughput
The Ryft ONE enables you to power a range of real-time intelligence applications and complex searches while taking advantage of both batch and streaming data, lower operating costs, increased revenue opportunities and better service quality.
Today’s data centers rely on commodity CPU clusters, but high-performance data analysis often requires a mix of CPUs with more specialized GPU and FPGA-based platforms. The Ryft ONE accelerator leverages the ideal mix of compute architectures, which along with high-level software abstractions, deliver data analytics at speeds of 200X or more than current infrastructures and with a 70% lower TCO.
Ryft offers an ever-expanding library of algorithms—search, edit distance, numerics and more—that make it simple and fast to analyze data.
Search structured and unstructured data together at speeds of 15+ GB/second with no data preparation needed
Access the most comprehensive and accurate results instantly by broadening search distance limits beyond conventional search tool capabilities
Increase edit distance values up to 10 or more to eliminate “no results found” errors—all without data transformation or indexing
Narrow search results by a specific date or date range—no matter the format—to get fast, accurate insights without time-consuming data preparation
Thin data to find only results matching a specific or range of time codes, all without extensive data indexing
Find specific—or ranges of—IP addresses for real-time insight on who is accessing what parts of your network
Find any specific or range of numbers using arbitrary formats, including integers, decimals, scientific notation, and more
Ryft’s growing library of data analytics algorithms enables you to get fast, actionable insights from all of your data
Download the IDC Technology Spotlight paper to learn more about the explosive growth in the HPDA market and how any organization can now afford HPC resources to gain valuable insights from their big data.
Download our API to seamlessly integrate Ryft into existing big data environments for faster data analysis and real-time intelligence.
See a live demonstration or learn how to transform high performance analytics in the AWS cloud for fast, data-driven decisions.