Intel® vPro™ Technology

Enabling Dynamic Virtual Client Computing with Intel® vPro™ Technology

Virtual Client Computing Vision and Futures

Our vision entails a combined software and hardware solution architecture for client platform virtualization in which end-user empowerment is enabled yet balanced with the delivery of robust IT security and manageability. By enabling a Dynamic Virtual Client computing framework, data types, computing execution models, software delivery strategies, and end-user roaming options can be balanced across alternative client computing scenarios, contexts, and administrative domains. Figure 3 depicts a platform model in which we envision hardware and software solution integration strategies.

System Overview

Figure 3: Client virtualization platform model
click image for larger view

The requirements for DVC computing are consistent with the core CIO needs for compliance, security, reliability, performance, and lower costs—all integrated with enterprise computing infrastructure and service delivery. Alternatively, DVC users will expect flexibility in terms of computing and image customization, mobility, and personalization, as well as assurance that their privacy needs will be met. As reflected in Figure 3, we envision multiple end-user run-time scenarios securely launched and executing on a VMM or software-based virtualization architecture with hardware-based virtualization underpinnings. In supporting an evolutionary client virtualization strategy, we recognize the following transitional vectors for DVC computing:

1. Hardware-assisted software isolation and delivery. The requirement for hardware separation enables alternative software packaging, data management, and delivery strategies, and thus, broader solution choices for IT administrators. Providing an unconstrained software innovation platform, while ensuring operational robustness through hardware, is essential to enabling a flexible computing experience for the user.
2. Usage-driven virtual machine platform architectures. As client virtualization matures and gains broad adoption by the IT community, we envision greater convergence on deploying alternative VM models. More specifically, we view client provisioning and deployment technologies facilitating alternative platform customization and software delivery strategies. These strategies will be based on functional or business contexts, user usage models, and varying client device acquisition and IT support scenarios.
3. Seamless physical and virtual computing manageability and security. A challenging but necessary shift that must take place to support virtual computing environments. First, management and security solutions must be redirected to comprehend the robust merits of hardware for secure manageability. Second, software must enable flexibility to adapt to contexts and the complexities of managing mixed physical and virtual computing environments.

Software Container and Delivery Models
Supporting the requirement for mixed computing models is an emerging usage scenario. For example, personal-corporate environment or dual OS environments for platform migration will soon be deployed seamlessly and coexist robustly within the same physical machine. Within this context, we expect to see alternative scenarios for image containment or software delivery over existing physical or virtual machine paradigms. Lightweight delivery models can support granular browsers, applications, or service container deployments, and heavier-weighted containers can support local or remote image OS (for example, streaming) deployments. Moreover, we envision a client virtualization framework that comprehends a true dynamic, distributed architecture that adjusts to IT and user constraints across the end-to-end spectrum. Thus, end-user requirements are met for roaming, optimal user experience and computing environment flexibility. At the same time, policies relating to critical IT runtime services and data requirements will be managed dynamically by centralized or decentralized instantiations as required by the IT organization.

Client Virtual Machine Architectures
Within a virtual client computing environment, either Type 1 or Type 2 virtualization is feasible and supported. We will look to integrate common hardware capabilities into both virtualization models, enabling robust virtual container and software delivery models. Matching traditional single-image models for client computing in terms of security, isolation, power and QOS, and reliability in the client virtualization space are challenges that we are tackling jointly via hardware- or software-based solutions. Intel® Virtualization Technology (Intel® VT) enables robust execution of unmodified guest operating systems to run on VMMs enhanced with Intel VT, as discussed in the earlier sections. Enhancements to Intel VT will continue to improve effective (QOS, security) IO device sharing, and will optimize energy and performance to support operating in a variety of DVC computing usage scenarios.

In addition, software delivery or image management systems will need to discover and customize images, based on alternative VMM platform and user execution profiles. Customers may employ a heterogeneous hardware and image deployment strategy that can utilize the inherent advantages of Type 1 or Type 2 virtualization environments and alternative image customization strategies that are based on specific usage scenarios. It is likely that there will be usage scenarios in which both models are supported. Providing a strategy that enables usage-based IT choice for Type 1 or Type 2 models, and more specifically, a strategy that is integrated into a common framework for platform deployment, software delivery, and manageability is key to ensuring broad client virtualization solution opportunities. In such cases, prescriptive guidance on how best to choose or manage the co-existence of mixed virtualization environments will be needed to accelerate deployments and unlock greater business value for the IT industry.

Management and Security
In supporting a client virtualization framework, systems management tools and Intel AMT should be integrated in a fashion that comprehends different virtualization architectures, software delivery models, and platform categories. Supporting a user's computing flexibility and roaming needs, synchronizing and securing data, and provisioning and managing of image (OS, application) granularities and delivery strategies will require hardware-based mechanisms to harden the virtualized platform. Software-based solutions will be needed that support a user's computing flexibility. We envision systems management tool capabilities and platform mechanisms evolving to support dynamic virtual client computing, based on the following requirements:

• Supporting existing in-band and out-of-band physical platform management capability and manageability features including alternative software installation or software delivery models, user profile management, and traditional platform management functions.
• Transitioning similar platform capability and feature mechanisms across locally-hosted virtual machine and application virtualization environments, including enabling more robust boot procedures and streaming delivery strategies that are fully integrated into the virtualization environment.
• Enabling novel management and security solution capabilities, specific to virtualized environments that are partitioned or fully integrated into end-user and corporate constructs. Additionally, dynamic, policy-based provisioning models that is, support for roaming, image and data delivery) for virtual machine environments and software container models should all be flexible enough to serve centralized or decentralized management models.

In partnership with key ISV partners, we will investigate the three major DVC vectors through advancement of hardware and software solution capabilities that support integration of traditional IT provisioning and administrative frameworks into underlying physical and virtual machine manageability constructs.

Back to Top