Why there is no single answer
The title of this article asks for a machine, but virtual reality did not begin with one. There was never a computer designed from the start as “the VR computer,” and no model that can be pointed to as the first. VR grew out of other work: flight training, real-time graphics and display research. What exists instead is a series of machines and installations, each of which supplied part of what an immersive system needs.
The term immersive simulation is a useful way to describe that shared goal. A simulation tries to reproduce how something behaves, and an immersive one surrounds the user with it. The question is which pieces of hardware first made that practical, and the answer is different for each piece.
Before digital computers: the Link trainer
The Link Trainer, developed by Edwin Link around the late 1920s, was a mechanical flight trainer with a cockpit that tilted and turned in response to the pilot’s controls. It had no computer or even a screen showing the outside world. It was used widely for instrument training, and its significance is conceptual: it showed that simulation could substitute for dangerous or costly real practice, which is still the main commercial argument for immersive systems.
Later simulators added electrical and analog computing to model how aircraft behaved. These were not VR in the modern sense, but they set up the reasoning that trainees could learn in an artificial environment that responded to their actions.
Digital simulators and generated imagery
As digital computers became fast enough in the 1960s and 1970s, simulator makers began to generate the outside view with a computer instead of filming a model landscape. Early visual systems were limited, drawing simple shapes such as runway lights and basic ground features, but they could be updated in real time as the trainee moved. Specialized graphics hardware from companies such as Evans & Sutherland was used for this purpose.
This is probably the strongest candidate for something that deserves a place in the story of an early VR computer. These were not general-purpose machines, though. They were purpose-built systems combining a host computer, graphics processors and displays, designed to run a single kind of task. Their cost meant that only air forces, airlines and large manufacturers could afford them.
Display and tracking research
In parallel, researchers were working on the viewer’s side of the system. In the 1960s, head-mounted displays were built with head tracking, which let a user look around a computer-generated scene. A later generation of research, particularly in the 1980s, added helmet-mounted displays for military pilots. One well-known project, the Super Cockpit led by Thomas Furness at the US Air Force, explored giving a pilot a three-dimensional view of information around the aircraft.
Another strand did not involve headsets at all. In the 1970s, Myron Krueger developed interactive installations in which cameras captured a user’s silhouette and a computer placed it in a projected world. It was a sign that immersion could be achieved by several routes.
The workstation era
By the late 1980s, graphics workstations made by firms such as Silicon Graphics could draw shaded three-dimensional scenes quickly enough for interactive use, at prices that a university or company could afford, though not an individual. Many early VR labs used one of these as the central computer, connected to a tracker and a display. NASA researchers, for example, assembled VR demonstration systems from this kind of equipment.
For a few years, therefore, the closest thing to a standard “VR computer” was a high-end graphics workstation, and the experience was still limited by the speed at which it could redraw two views. Delay between head movement and the updated image caused discomfort, and engineers spent considerable effort on tracking and prediction.
From specialized machines to ordinary hardware
Over the following decades, the processing that once required a room of equipment moved into smaller packages. The relevant ingredients are covered elsewhere: the growth of microprocessors made general computers faster, and the spread of personal computers put them in homes. Eventually, consumer graphics cards could do the work of the earlier workstations. Modern headsets either connect to such a computer or contain a small computer of their own.
So the lineage is not a line from one machine to the next. It is a handing-over of capability, from custom simulator hardware, to workstations, to commodity graphics hardware. Each stage made the previous stage’s expensive features available to a wider group.
A careful way to summarize
It is reasonable to say that:
- no single machine was the first VR computer;
- real-time computer-generated imagery for training devices was an early and important precursor, though the details of which system came first are hard to pin down and sources differ;
- laboratory systems of the 1960s and 1980s supplied the tracked, head-mounted display concept;
- general-purpose hardware only became capable of delivering a comfortable experience much later.
Anyone claiming one exact machine and one exact date is probably choosing a definition that suits the claim.