The COVID-19 pandemic and the sudden adoption of work-from-home schemes for many knowledge economy workers has generated considerable interest in creating a long-lasting remote workforce. Hiring remotely can decrease wage costs by hiring outside expensive cities, consequently distributing wealth more evenly geographically, and bring down rent costs as offices could become smaller. This is incentivizing companies to seriously investigate the merits and drawbacks of cloud-based IT architectures to determine whether moving away from physical offices, where processing power is redundantly repeated at every desk, is feasible.
Virtual desktop infrastructure (VDI) and its mobile-device counterpart, virtual mobile infrastructure (VMI), have been growing in popularity with the general trend towards cloud-based systems. The COVID-19 situation has catapulted them to the forefront of potential future structures as the style and culture around work undergoes rapid transformations due to the virus.
This article briefly explains what VDI and VMI are, how they work, benefits and disadvantages, and then compare them. Finally, we conclude with a short section on the trend to Unified Endpoint Management (UEM), which brings desktop, mobile, and IoT devices into a single overarching framework.
Let’s compare VMI vs. VDI.
What are VMI and VDI?
As a virtual mobility company, we have separate articles detailing what VMI is, how it works, security benefits, and the costs. VDI is very similar except it applies to desktop environments (Windows, Linux, and macOS) instead of mobile environments (mostly Android and iOS).
The foundational idea of this virtual infrastructure scheme is to physically locate the data storage and processing on a centralized, remote server. This can be accomplished through containerization or hypervisor virtual machine creation. The client device – here the employee’s laptop, desktop, smartphone, or tablet – connects to the centralized server, is assigned a virtual machine on that server, and can graphically view that machine through a screen-painting program on the local client device.
This setup ensures full files are never downloaded locally to the employee device, greatly improving security. Moreover, backups are easily handled by the centralized server, so data is generally much easier to recover. Since all the processing is also handled by the powerful centralized server, very low-spec “thin client” hardware can still initiate computing tasks that would quickly overwhelm the local machine.
Persistent environments – those that are saved and accessible again by specific users – make it easier for employees to move to other devices, such as desktop terminals in other buildings or between various mobile devices. Non-persistent versions do not allow any saving on their particular instance, which makes it easier for companies that need exactly the same environment on every device, regardless of which employees access it.
The shared, centralizing conceptual idea of remote processing and data storage implies VMI and VDI share quite a few traits. These include benefits for security, mobility, and management plus the network problem drawback. Adoption rates and methods also face similar challenges.
As mentioned above, files are never downloaded in their native form to the local machine. This means security is heightened, preventing the theft of files from a cloned hard drive or a RAM dump. The only information on the local persistent memory or in RAM is an image data of whatever was viewed during the session – and in RAM it is mostly just be what is on-screen at the time of the attack.
Containerization and virtualization both also provide tools to track exactly who is doing what in a centralized manner and block access to certain suspicious client devices or IP addresses. IT teams can efficiently and thoroughly patch security vulnerabilities for both VMI and VDI by configuring instances on the central server, restarting existing sessions if necessary. This is much harder if every employee has his or her own local environment that only connects at will. Furthermore, updates need only be focused on the installations on the central server, not on every single employee device, which could contain widely varying hardware.
Both types also permit easy migration to other terminals or locations. Employees simply need to authenticate themselves and access a session. Any “dumb terminal” can be used, from leftover computers in the office storage room to the latest tablets and smartphones.
Finally, both systems can be presented to employees for adoption in similar ways. Bring-your-own-device (BYOD) is quite popular for encouraging VMI adoption. The scheme allows workers to use their own smartphones to access all their work data and programs with the security benefits and the ease of mobility readily available. BYOD is also possible for VDI, where a user can access an instance of a desktop on a company server with the same programs, security, and mobility benefits as for mobile environments.
The adoption rates for mobile BYOD can be promising, though resistance can be high from users that are wary of Mobile Device Management (MDM), especially when VMI is not implemented. A similar privacy concern can lead to resistance for desktop BYOD schemes.
VMI vs. VDI
The architecture and many of the benefits and even adoption methods are similar between VMI and VDI. However, there are key differences that an organization must evaluate when considering to adopt either scheme.
First and foremost are the target device environments. VMI is designed for mobile devices, which tend to have lower hardware specs, are more conscious of power usage, and move freely between networks. Mobile devices employ over-the-air (OTA) cellular connections and thus are more conscious of data consumption. Conversely, stationary desktops and plugged-in laptops – which have replaced desktops at many firms – consume more power and data, usually on unmetered connections.
In relation, though invisible to users, CPU architecture is important, since operating systems are not designed to run on all CPUs. Stationary devices (including laptops) tend to utilize x86 or amd64 processors while mobile devices usually rely on ARM. The difference is a non-issue for users – most have probably never even heard of ARM/x86/amd64 architectures – but CTOs and IT professionals must be cognizant of their organization’s hardware.
Very visible to users, however, is display scaling. VMI solutions are designed for mobile displays, which are smaller than desktop displays. No one wants to zoom in to every point of action on a VDI program because they’re using a mobile device to access it, and scaling of VMI viewer programs to cover entire desktop spaces can result in unsightly stretching. Both issues can lead to frustration.
The use cases for desktops and mobile platforms are different, too. The former are employed in more complex tasks, such as manipulating and building spreadsheets, while the latter are used more for information consumption and simpler tasks. It is possible to build spreadsheets or code on a mobile touchscreen, but it is certainly not the most effective or smooth approach. A mobile device may be useful for setting off a very complex calculation on the central server without locally processing anything, but it is not very good for setting up the calculation in the first place.
Perhaps surprisingly, the culture surrounding the two platforms delineates other differences. Mobile devices are newer and the ecosystems more malleable, so users are more accustomed to changes. Desktops are more entrenched in traditional business practices and users may be more resistant to changes or new approaches. Furthermore, true desktops (not laptops) provide a physical distinction between work and personal life that manifests at the office’s front door, while mobile BYOD schemes blur that line by tying work and personal life to the same device. With mobile BYOD already normalized, the workforce may more willingly adopt mobility solutions, but VDI solutions installed on personal machines may continue to face stiff resistance over fears of further encroachment of work-life into the 24 hours of the day.
Unfortunately, the difference in spatial presentation and the kinds of action that can be efficiently performed makes it difficult to unify the two schemes. But technology always evolves.
A Unified Future?
There is some hope in a Unified Endpoint Management (UEM) system that can unify these two schemes and implement them on all devices so mobile, desktop, and IoT management can be handled more centrally. At least for now, UEM may turn out to be a solution to “dashboard” device management without actually being a single suite of software that adapts to mobile, desktop, and IoT screens, hardware, and use cases. But managing everything centrally lifts some of the burdens from thinly-stretched IT teams and reduces headaches in management.
A major appeal of this framework is the empowerment of employees to seamlessly transition between desktop and mobile platforms as they move in and out of spaces. A backend system may save the state of an office desktop environment and recreate that state for a mobile environment when the employee loads their workspace at the client site.
Even if UEM does not bring these approaches together, they both may shift towards “as a Service” products, namely Mobility as a Service (MaaS) and Desktops as a Service (DaaS), wherein a specialized company organizes and manages much of the technical background. From corporate management’s standpoint, this at least reduces the contact points to third-party companies and diminishes the headaches of managing everything internally.
Hypori offers services for mobility management through its virtual mobility solutions, which can serve as one of these reduction vectors.