Traditional computing does not allow software to readily share hardware resources. Virtualization overcomes that limitation and allows two or more virtual machines (VMs) to share the computing resources of a single physical server, and the VM monitor is central to any virtualization deployment. It provides the abstraction that separates software from the underlying hardware and manipulates the virtual workloads so that each workload can share the common processor, memory, I/O and local storage. Early virtualization efforts relied on software emulation to replace hardware functionality. But software emulation is a slow and inefficient process.
Many virtualization tasks were handled through software, so VM behavior and resource control were often poor, resulting in unacceptable VM performance on the server. By early 2005, processors still lacked the internal microcode to handle intensive virtualization tasks in hardware. Both Intel Corp. and Advanced Micro Dynamics Inc. (AMD) addressed this problem by creating a new set of processor extensions (similar to
Requires Free Membership to View
When you register, my team of editors will also send you the latest expert resources covering all areas of server virtualization, such as platforms, architectures and strategies, server hardware, managing virtual environments, application issues and more.
Cathleen A. Gagne, Senior Editorial Director
MMX, 3DNow! and others) that could offload the repetitive and inefficient work from the software. By handling these tasks through processor extensions, traps and emulation of virtualization, tasks through the operating system were essentially eliminated, vastly improving VM performance on the physical server.
AMD Virtualization (AMD-V) technology was first announced in 2004 and added to AMD's Pacifica 64-bit x86 processor designs. By 2006, AMD's Athlon 64 X2 and Athlon 64 FX processors appeared with AMD-V technology, and today, the technology is available on Turion 64 X2, second- and third-generation Opteron, Phenom and Phenom II processors.
AMD performance with hardware-assisted virtualization centers on memory management that builds on prior Direct Connect Architecture. In effect, the physical processor and the guest VM can communicate directly. This direct communication reduces the overhead often encountered with emulation and improves the way that memory space is handled, because the processor works with the virtual instance.
AMD-V introduced Rapid Virtualization Indexing (RVI) features that allow VMs to manage memory directly. RVI reduces the number of processing cycles needed by the hypervisor that would otherwise be wasted handling memory-related operations. In addition, RVI lets the processor switch between numerous virtual guest instances very quickly, further enhancing virtualization performance. Memory handling and switching is also boosted by tagged Translation Lookaside Buffer (TLB) capabilities that map memory space to the individual VM. This reduces memory management and speeds up the switching process between VMs.
AMD-V-supported processors also provide extended migration capabilities that allow virtualization platforms to easily and quickly migrate VMs across servers that run these processors. It's important to note that extended migration is not necessarily compatible with Intel processors, and this may cause performance issues (or outright failures) when migrating VMs between servers with different processor makers.
This was first published in October 2009