architectureToday, Bigstream announces the general availability of the industry’s only hyperacceleration platform for big data. After years of research and nearly two years of stealth mode development, the Bigstream Hyperacceleration Layer delivers 2X to 5X performance acceleration of Apache Spark workloads while requiring zero application code changes. Subsequent Bigstream releases will incorporate seamless acceleration using Field Programmable Gate Arrays (FPGAs) and Graphics Processing Units (GPUs) that can achieve up to 30X performance acceleration of Spark SQL, Machine Learning and other big data platforms

Frictionless Experience on Amazon AWS

Providing seamless hyperacceleration as a service in the cloud, Bigstream is available today on the Amazon AWS Marketplace. Once you subscribe to the Bigstream service, it takes just minutes to configure. To engage Bigstream hyperacceleration, it is simply a matter of creating an Amazon EMR cluster and adding a single Bigstream bootstrap URL as you create your Amazon EMR cluster. And that’s it: the Bigstream service is ready to use.

Bigstream offers a 14-day free trial of the service and can work with you to achieve the ideal acceleration based on your specific workloads and requirements.

Bigstream is also available for use in customer data centers. Customers can register today to get a subscription for use of Bigstream on premises. Bigstream will soon announce hyperacceleration services for Microsoft Azure. 


Bigstream Hyperacceleration Layer


While various hardware and software acceleration technologies are in use today, they tend to have a narrow focus and require very specialized skill sets. The use of hardware accelerators such as GPUs and FPGAs are also of particular interest to those who wish to add computing horsepower to data science and analytics workloads. The great challenge is to apply these accelerators in ideal combinations to get the best possible performance, and to do so with minimum effort.

When orders of magnitude performance acceleration is done seamlessly and dynamically, without application code changes, and without requiring specialized knowledge of FPGAs or GPUs, this is what Bigstream calls hyperacceleration.

The Bigstream Hyperacceleration Layer achieves these goals by combining the most effective software acceleration techniques – native compilation, vectorization, locality optimization – with the ability to dynamically exploit the power of GPUs and FPGAs for maximum effectiveness.

For more detail about how Bigstream achieves this type of hyperacceleration, read the Bigstream whitepaper.

Benchmark Your Spark SQL Project/POC with Bigstream

Bigstream is looking to engage with big data teams with Spark workloads. Bigstream employs some internally developed performance analysis tools for Spark that could provide you with some valuable insights into how your Spark application is performing. Contact the Bigstream development team to inquire about a Bigstream sponsored POC by emailing