Digital Video Interactive (DVI) was the firstmultimedia desktop video standard for IBM-compatiblepersonal computers. It enabled full-screen, full motion video, as well asstereo audio, still images, and graphics to be presented on a DOS-based desktop computer using a special compression chipset. The scope of Digital Video Interactive encompasses a file format, including a digital container format, a number of video and audio compression formats, as well as hardware associated with the file format.[1]
Development of DVI was started around 1984 by Section 17 of TheDavid Sarnoff Research Center Labs (DSRC) then responsible for the research and development activities ofRCA.[2] WhenGeneral Electric purchased RCA in 1986, GE considered the DSRC redundant with its own labs, and sought a buyer. In 1988, GE sold the DSRC toSRI International, but sold the DVI technology separately toIntel corporation.[2]
DVI technology allowed full-screen, full motion digital video, as well asstereo audio, still images, and graphics to be presented on a DOS-based desktop computer. DVI content was created using the Authology Multimedia authoring system developed by CEIT Systems and usually distributed onCD-ROM discs, which in turn was decoded and displayed via specialized add-in card hardware installed in the computer. Audio and video files for DVI were among the first to usedata compression, with audio content usingADPCM. DVI was the first technology of its kind for the desktop PC, and ushered in the multimedia revolution for PCs.
DVI was announced at the second annual Microsoft CD-ROM conference in Seattle to a standing ovation in 1987. The excitement at the time stemmed from the fact that a CD-ROM drive of the era had a maximum data playback rate of ~1.2 Mbit/s, thought to be insufficient for good quality motion video. However, the DSRC team was able to extract motion video, stereo audio and still images from this relatively low data rate with good quality.
The first implementation of DVI developed in the mid-80s relied on three 16-bitISA cards installed inside the computer, one for audio processing, another for video processing, and the last as an interface to a Sony CDU-100 CD-ROM drive. The DVI video card used a custom chipset (later known as the i80750 ori750 chipset) for decompression, one device was known as the pixel processor & the display device was called the VDP (video display processor).
Later DVI implementations used one, more highly integrated card, such as Intel's ActionMedia series (omitting the CD-ROM interface). The ActionMedia (and the later ActionMedia II) were available in both ISA andMCA-bus cards, the latter for use in MCA-bus PCs likeIBM'sPS/2 series.
Intel utilized the i750 technology in driving creation of theMMX instruction set.
The DVI format specified two video compression schemes – Presentation Level Video, or Production Level Video, and Real-Time Video – and two audio compression schemes – ADPCM and PCM8.[3][1]
The original video compression scheme, called Presentation Level Video, was asymmetric in that aDigitalVAX-11/750 minicomputer was used to compress the video in non-real time to 30 frames per second with a resolution of 320 × 240. Encoding was performed by Intel at its facilities or at licensed encoding facilities set up by Intel.[4] Video compression involved coding both still frames and motion-compensated residuals usingvector quantization in dimensions 1, 2, and 4. The resulting file (in the.AVS format) was displayed in real-time on an IBM PC-AT (i286) with the add-in boards providing decompression and display functions at NTSC (30 frame per second) resolutions. The IBM PC-AT equipped with the DVI add-in boards hence had 2 monitors – the original monochrome control monitor and a second Sony CDP1302 monitor for the color video. Stereo audio at nearFM broadcasting quality was also available from the system.
The Real-Time Video format was introduced in March 1988, then called Edit-Level Video (ELV).[5] In fall 1992, version 2.1 of the Real-Time Video format was introduced by Intel asIndeo 2.[5][6]
 | This sectionmay containoriginal research. Pleaseimprove it byverifying the claims made and addinginline citations. Statements consisting only of original research should be removed.(April 2016) (Learn how and when to remove this message) |
The original team from DSRC (David Sarnoff Research Center) set up an Intel operation NJ1 as the Princeton Operation. The team occupied new quarters after moving out of the DSRC in Plainsboro, New Jersey. From the original 35 researchers the Princeton Operation grew to over 200 people at its height.Andy Grove was a great[clarification needed] supporter of the Princeton Team during its term of operation. However in 1992Ken Fine (a vice president of Intel) decided to shutter the operation and transfer those employees willing to move to other Intel sites in Arizona and Oregon. Fine left the company shortly after he implemented this decision. Final site closure occurred almost a year later in September 1993.