Microsoft open sources algorithm, hardware and source code for Azure data compression

Microsoft is open-sourcing and releasing to the Open Compute Project its 'Project Zipline' data-compression algorithm, plus related hardware specs and source code.
Written by Mary Jo Foley, Senior Contributing Editor
Credit: Microsoft

Microsoft is open-sourcing its cloud-compression algorithm and optimized hardware implementation for cloud storage. Microsoft is contributing that algorithm, known as "Project Zipline," plus the associated hardware design specs and related source code to the Open Compute Project, the company announced on March 14.

"Project Zipline is a cutting-edge compression technology optimized for a large variety of datasets, and our release of RTL (Register Transfer Language) allows hardware vendors to use the reference design to produce hardware chips to allow the highest compression, lowest cost, and lowest power out of the algorithm," Microsoft officials said in a blog post.

Microsoft is hoping the Open Compute Project community will contribute to the algorithm and specs. Microsoft officials said they anticipate the Zipline compression technology could make its way into various markets in areas like network data processing, smart solid state drives, archival systems, cloud appliances, and general-purpose microprocessor, and edge devices.

Microsoft made the announcement today to coincide with the kick off of The Open Compute Project Global Summit 2019. At the Open Compute Project U.S. Summit last year, Microsoft introduced Project Denali, its spec for standardizing SSD firmware interfaces.

Before that, Microsoft introduced and contributed Project Cerberus, a standard for a cryptographic microcontroller. Microsoft officials described Cerberus as the next phase of Project Olympus, its datacenter server design which the company also contributed to the Open Compute Project. Microsoft itself deployed Project Olympus hardware in Azure with its Fv2 virtual machine family

Microsoft joined the Open Compute Project (OCP) in 2014 and is a founding member of and contributor to the organization's Switch Abstraction Interface (SAI) project. The OCP publishes open hardware designs intended to be used to build data centers relatively cheaply. The OCP has already released specifications for motherboards, chipsets, cabling, and common sockets, connectors, and open networking and switches.

Editorial standards