Just as the increasing awareness level of the stereoscopic cinema, so the perception of limitations while watching movies with 3D glasses has been emerged as well. It is not only that the additional glasses are uncomfortable and annoying; there are some tangible arguments for avoiding 3D glasses. These “stereoscopic deficits” are caused by the 3D glasses itself. In contrast to natural viewing with naked eyes, the artificial 3D viewing with 3D glasses introduces specific “unnatural” side effects. The most of the moviegoers has experienced unspecific discomfort in 3D cinema, which they may have associated with insufficient image quality. Obviously, quality problems with 3D glasses can be solved by technical improvement. But this simple answer can -and already has- mislead some decision makers to relax on the existing 3D glasses solution. It needs to be underlined, that there are inherent difficulties with the glasses, which can never be solved with modest advancement; as the 3D glasses initiate them. To overcome the limitations of stereoscopy in display applications, several technologies has been proposed to create a 3D impression without the need of 3D glasses, known as autostereoscopy. But even todays autostereoscopic displays cannot solve all viewing problems and still show limitations. A hyperview display could be a suitable candidate, if it would be possible to create an affordable device and generate the necessary content in an acceptable time frame. All autostereoscopic displays, based on the idea of lightfield, integral photography or super-multiview could be unified within the concept of hyperview. It is essential for functionality that every of these display technologies uses numerous of different perspective images to create the 3D impression. Such a calculation of a very high number of views will require much more computing time as for the formation of a simple stereoscopic image pair. The hyperview concept allows to describe the screen image of any 3D technology just with a simple equation. This formula can be utilized to create a specific hyperview matrix for a certain 3D display – independent of the technology used. A hyperview matrix may contain the references to loads of images and act as an instruction for a subsequent rendering process of particular pixels. Naturally, a single pixel will deliver an image with no resolution and does not provide any idea of the rendered scene. However, by implementing the method of pixel recycling, a 3D image can be perceived, even if all source images are different. It will be proven that several millions of perspectives can be rendered with the support of GPU rendering and benefit from the hyperview matrix. In result, a conventional autostereoscopic display, which is designed to represent only a few perspectives can be used to show a hyperview image by using a suitable hyperview matrix. It will be shown that a millions-of-views-hyperview-image can be presented on a conventional autostereoscopic display. For such an hyperview image it is required that all pixels of the displays are allocated by different source images. Controlled by the hyperview matrix, an adapted renderer can render a full hyperview image in real-time.