Open Source Hazelcast Data Grid Optimized on Intel Optane for AI, IoT Workloads

The recently announced collaboration between in-memory computing platform provider Hazelcast and processor maker Intel to accelerate Hazelcast technologies on Intel chips has borne fruit. The latest release of Hazelcast's namesake Java-based in-memory data grid (IMDG 4.0) has been optimized for Intel Optane DC Persistent Memory. The result, the two companies say, is increased data density and more cost-efficient access to in-memory speeds.

The strategic co-engineering effort, dubbed Project Veyron, is focused on accelerating the performance of real-time applications, artificial intelligence (AI), and internet of things (IoT) solutions for the enterprise. Intel Optane is workload-optimized technology that comes with all new 2nd Generation Intel Xeon Scalable processors.

The Hazelcast IMDG is an open source product distributed under an Apache 2 license that allows developers to include the grid in their applications. The combination of Hazelcast in-memory solutions with Intel scalable processors with memory persistence amps up speed, reliability, and scalability, the companies said, because organizations can run more workloads in-memory.

Intel bills its Optane DC Persistent Memory as a more scalable and cost-effective alternative to RAM. When deployed in persistence mode, they said, Intel Optane DC Persistent Memory and Hazelcast IMDG provide resiliency by enabling fast recovery from temporary node outages.

The embrace of ultra-low latency computing is quickly becoming "a defining element" of leading-edge enterprises, said Hazelcast CEO Kelly Herrell, in a prepared statement. With support for Intel Optane, Hazelcast continues to provide "advancements for large enterprises to leverage in-memory computing to power business-critical applications at the lowest possible latency and accelerate their businesses," Herrell said.

Hazelcast IMDG 4.0 comes with a number of other updates, including a re-engineered inter-client protocol designed to reduce network hops, which increases write throughput by about 30 percent. With this release, Hazelcast is also speeding up client-server connectivity and transport layer security (TLS) communications by 50 percent and 200 percent, respectively.

Earlier this year, Hazelcast introduced the industry's only CP capability in an in-memory data grid, which delivers the highest data correctness. This capability, which refers to the consistency and partition tolerance guarantees in the CAP Theorem for distributed systems, ensures a continuously consistent view of data, even in the event of a computer or network failure. In the latest release, Hazelcast enhances the CP Subsystem by adding robust data persistence, so that CP values will be available on restart.

Palo Alto, Calif.-based Hazelcast has roots planted deeply in the Java language and platform (its in-memory data grid is written in Java), but the company has spread its support over the years, thanks in part to the efforts of its open source community, to include several clients and programming languages.

About the Author

John K. Waters is the editor in chief of a number of sites, with a focus on high-end development, AI and future tech. He's been writing about cutting-edge technologies and culture of Silicon Valley for more than two decades, and he's written more than a dozen books. He also co-scripted the documentary film Silicon Valley: A 100 Year Renaissance, which aired on PBS.  He can be reached at [email protected].