Virtual-machine-based emulation of future generation high-performance computing systems

Abstract

This paper describes the design of a system to enable research, development, and testing of new software stacks and hardware features for future high-end computing systems. Motivating uses include both small-scale research and development on simulated individual nodes of proposed high-performance computing systems, and large scaling studies that emulate a sizeable fraction of a future supercomputing system. The proposed architecture combines system virtualization, architectural simulation, time dilation, and slack simulation to provide scalable emulation of hypothetical systems. Virtualization-based full-system measurement and monitoring tools are also included to aid in using the proposed system for co-design of high-performance computing system software and architectural features for future systems. Finally, this paper provides a description of the implementation strategy and status of the system.

Using unreliable virtual hardware to inject errors in extreme-scale systems

“…VMMs designed for large-scale systems (e.g., Palacios) allow virtualization to be deployed throughout an entire HPC system with very low overhead [19]. In addition, they are being used to emulate next-generation large-scale systems [4]. As a result, a virtualization-based approach can readily mimic hardware failure regimes in which errors occur not just on a single node but throughout the entire system.…”

Operating Systems and Hypervisors for Network Functions: A Survey of Enabling Technologies and Research Studies

“…In such a so-called whole system VM, the VMM translates the guest operating system ISAs to the host operating system ISA, i.e., the VMM essentially emulates for each guest operating system a host system that understands the ISA of the guest operating system. Such a whole system (emulation) VMM is commonly used during the development and testing of new applications or hardware [30] as the whole system (emulation) VMM completely decouples the guest operating systems and applications from the host operating system and platform hardware.…”

High-performance emulation of heterogeneous systems using adaptive time dilation

“…High-performance virtual machines . The work in Bridges et al (2012) provides a high-level framework in terms of using virtualization in high-performance computing. The work in Cui et al (2012) closes the gap between the real and emulated systems by using receiver-side optimizations (optimistic interrupts and cut-through forwarding) in real time.…”

“…The approach in Grau et al (2009) uses the concept of epoch-based virtual time (Grau et al, 2008) where the experiment is divided in epochs, each of which has a constant TDF. Another paper (Bridges et al, 2012) also recognizes dynamic time dilation as a key component of future high-performance computing systems.…”