The von Neumann computer, developed in the 1940s, refers to a type of computer architecture in which instructions and data are stored together in a common memory. Computers using this design revolutionized “life as we know it” by providing an efficient appliance for compute-centric tasks. That’s why von Neumann computers—most commonly seen in x86 systems—still rule the world of laptops, supercomputers and workstations.

To their credit, big data and high performance computing (HPC) pioneers have been squeezing faster and faster performance from their available von Neumann systems for years. Even today, despite the crushing volume of data stemming from mobile, social media, cloud computing and the Internet of Things, they continue to try to make conventional clusters perform well enough to meet the insatiable desire for faster business insights. They’ve parsed the data across hundreds of servers, layered on the coolest new open source software and tuned algorithms to streamline data feeds. In the process, they’ve pushed x86 architecture to its absolute limit.

It’s Time for a New Architecture
At Ryft, we understand the challenges of those trying to get faster results from ever-increasing, ever-changing data sets. For years, we’ve designed high performance data analysis hardware that eliminates performance bottlenecks inherent in von Neumann computers. Working with our customers—some of the most sophisticated and largest users of high performance data analysis applications in the world—we built products that delivered unprecedented analytics performance in compact packages. Our tablets can score and match photos at 2.2 million records per second, while our chassis cards can search all of Twitter—without indexing—at the same high performance levels.

Today, at long last, we are pleased to share the news that we are making available our latest technology breakthrough. The underlying hardware technology is a hybrid FPGA/x86 compute architecture, and it has enterprise-friendly features to help a broad range of companies unlock faster analysis performance. In addition to the easily integrated hardware, we’ve created a portfolio of algorithm components, called the Ryft Algorithm Primitives (RAP) library. The result of combining the two is called the Ryft ONE. It’s the world’s first 1U appliance purpose-built to enable enterprises to accelerate big data analysis. And it’s really fast. The Ryft ONE lets you analyze data at an amazing 10 GB/second or faster.

With the Ryft ONE, getting answers from your business data is faster, easier and more cost-effective than any existing solution. Ryft ONE allows the simultaneous analysis of both historical and streaming data at speeds up to 100X those of conventional hardware. The speed and performance bottlenecks that hampered von Neumann architectures disappear. Using primitives, such as search, fuzzy search and term frequency, big data users can unlock actionable insight from all data, without time-consuming indexing or expensive extract, transform and load (ETL) operations.

Experience More Data and Less Center
With the hybrid FPGA/x86 hardware and RAP working seamlessly together, the Ryft ONE looks like a Linux server but performs like a high performance computer. Compliant with open standards, it works with a wide range of visualization, scheduling, performance monitoring and systems management tools. In addition, the Ryft ONE supports popular programming languages, such as C/C++, Java, R, Python, Scala and others. And incredibly, the Ryft ONE does everything using the power of a hair dryer. That means not only will you experience incredible performance but you can also reduce operational costs by 70%.

Want to see it in action? Check out a demo and schedule a free trial to get started today.

Leave a Reply

Your email address will not be published. Required fields are marked *