Security and the Pano

The Pano Device is secure because it cannot run an operating system or any other software. Because there is nothing in a Pano Device that can be infected by a virus or have malicious code installed, it doesn't need to be scanned for vulnerabilities or exploits. And while the virtualized Windows OS is still vulnerable to malware, the ability to easily drop an infected desktop virtual machine and provision a clean copy, even without IT intervention, makes recovering from malware infections simple and foolproof.

Even when a peripheral such as a USB thumb drive is connected, the Pano Device can remain secure. Peripherals work only when the user is authorized via Windows policies enforced by the Pano Manager. If the user isn't authorized, Windows doesn't even see the peripheral that is connected to the Pano Device. If a user is authorized, the peripheral is connected directly to Windows using whatever native drivers are required. This lets you enforces fine-grained access policies based on user group membership, USB device class, and operation.  As an example, a user on a specific Pano Device may be authorized to read from a CD, but not write to it - letting users to copy files or load software onto their virtual desktop, but preventing any confidential data from leaking out. Even when users are authorized, Pano Manager can record USB operations so that the business can keep track of all its information assets.

Finally, because the Pano Device has no internal storage device or even any memory, it can never hold any confidential or private data.  This means security breaches and data losses due to lost or stolen endpoints (like laptops) are a thing of the past.  All data is kept in the data center where it can best be secured, managed, and backed up.  And beyond security benefits this centralization shortens the path between applications and the data they process letting IT optimize the storage infrastructure, resulting in performance improvements over desktop PC hard drives and old client-server computing models.