Page 1 of 2
Scale Computing on Tuesday said it closed a new round of funding for $12 million to help it continue developing hyperconverged infrastructure, or "datacenter-in-a-box," offerings that integrate server, storage, networking and hypervisor into a single managed system.
With the new Series D round of funding, total funding in Indianapolis, Ind.-based Scale Computing comes to $43 million, said CEO Jeff Ready. The new round of funding will be used to develop its hyperconverged solution technology and launch a new partner program later this year.
Scale Computing, which started as a builder of clustered storage solutions, is one of a handful of vendors developing hyperconverged IT solutions that tightly integrate the server, storage, networking and virtualization technologies into a single appliance. Adding additional nodes expands both the performance and the storage capacity of the infrastructure.
Ready said hyperconverged solutions differ from more hardware-based converged infrastructure offerings such as the VCE Vblock, the joint EMC-Cisco VSPEX solutions, or the joint NetApp-Cisco FlexPod or ExpressPod solutions in the integration level.
"In the early hardware-based converged infrastructures, when you order a kit, you get the components and a nice manual for putting them together," he said. "With hyperconvergence, there is no storage layer to manage, or no virtualization layer to manage. Customers get one system to manager. There's no buying or managing a SAN. The SAN doesn't exit. It all happens behind the scene."
Scale Computing's hypervisor, which is based on Red Hat's KVM, is buried deep within the company's HC3 appliances so no license is needed for the virtualization layer. "There's no more need for a human to sit there worrying about the virtualization than there is for a human to sit worrying about the processor communicating with the RAM."
The Scale Computing hyperconverged solutions start with a minimum of three HC3 nodes, each of which is a 1U rack-mount appliance, Ready said. A four-node solution can serve about 40 to 50 virtual machines with all the needed storage, networking and hypervisor, all of which runs as a single package. "If you need more compute power, or more storage, plug in more nodes," he said.