CAMPBELL, CA – TidalScale releases version 1.0 of the TidalScale HyperKernel. TidalScale offers the unique ability to transparently deploy a software-scaled computer to tackle large, data intensive problems in-memory. Until now, organizations with large data problems have been forced to slice their applications and data to fit into the memory of several smaller physical machines. The TidalScale HyperKernel eliminates that complexity.
The TidalScale HyperKernel will appeal to users who need a large memory space for applications like data mining, machine learning, and simulation.
The HyperKernel is a distributed hypervisor that allows a single operating system (like Linux) to transparently run across multiple commodity servers. The HyperKernel aggregates all the processors, memory, storage, and networks into a single virtual system that automatically and dynamically optimizes for best performance. The TidalScale HyperKernel requires no changes to existing applications or to standard operating systems or databases.
Software-scaled computers open the door to a wider range of new in-memory applications because users can deploy virtual systems much larger than the standard hardware servers typically available in data centers today but using the same equipment.
The TidalScale HyperKernel is currently in use by several customers. Customers are successfully running applications with Revolution R, Python, and AnyLogic. In Q4 2015, TidalScale published a Revolution R benchmark showing TidalScale HyperKernel's performance results.
The product is available for immediate deployment. More information can be found at http://www.tidalscale.com