The newest version of the Xen hypervisor will support servers that use ARM-based chips, according to a Tuesday announcement from the Linux Foundation. This is a big deal in bringing ARM-based servers into the webscale data center since many large-scale cloud providers are using the Xen hypervisor.
Developers working on the Xen Project collaborated with Calxeda and other vendors to make sure Xen works on ARMv7 and ARMv8 architectures, according to the release. Xen will join KVM in supporting ARM. It’s unclear if and when Microsoft’s Hyper-V or VMware’s ESX hypervisor will support the ARM architecture.
Until a few years ago, almost all enterprise software was written to run only on x86-based chips. VMware has said it wouldn’t support ARM in the past. Getting hypervisor support is important for ARM because without the built-in support companies can’t commit to buying ARM-based servers.
From a blog post on the new Xen 4.3.0 from collaborator George Dunlap:
Hardware is not yet available for 64-bit ARM processors yet, but Xen is running well in 64-bit mode on AEMv8 Real-time System Models by ARM.
ARM-based server vendors have serious interest in enabling virtualization on the boxes. So it’s no surprise to see that AMD contributed 23 percent of all changes to the lines of code comprising the new version of Xen. Other contributors: Fujitsu, Oracle and — surprise! — Intel. And Samsung “is leading the XEN for ARM port,” a Linux Foundation spokeswoman wrote in an email.
Lower-power servers are the way of the future in this age of the parallelized, scale-out cloud world, and ARM is sure to be a big part of that. At least Andrew Feldman, AMD general manager and corporate vice president, said he expected as much at our Structure conference last month. That future might not be so far off.