The runaway success of Sun Microsystems and Apollo Computer in the technical workstation market has energised a string of major companies from IBM through Hewlett-Packard to Tektronix to attempt to win a piece of the action, but the total failure of IBM’s RT to make an impact on the US market underlines the up hill struggle faced by companies hoping to break the stranglehold of the Sun-Apollo duopoly. Bloodied but unbowed by the failure of its first attempt, NEC next spring will come back with a redesigned product, have lighted on the workstation market as its white hope for making its first big impact on the US computer market. Computer Systems News reports that after having had its first generation 68020 workstation on the US market for a year to very little effect, NEC Information Systems Inc has gone back to the drawing board, redesigning the thing with two 16.67MHz 68020s and a native VMEbus in place of the Q-bus with VMEbus attached that was used in the orig inal product. The new workstation will use one 68020 as the central processor and a second to manage graphics and windows, supported by a custom graphics chip. NEC is hedging its bets by also designing an 80386-based station with 80387 maths co-processor and optional Weitek floating point co processor as well. Unix System V.3 will host MS-DOS as a task within a window and the station will come with X Window, Ethernet and TCP/IP commun ications. The biggest hurdle that NEC will have to overcome is lack of applications: the company has just 10 for its first generation station, but claims that it will have products from 50 to 75 soft ware developers for the new ones – but that compares with over 1,000 each for Sun and Apollo. NEC will also have to compete fiercely on price, and margins are already ex ceedingly slim at the low end of the workstation market, where the cheapest Suns and Apollos are pit ched below 80386-based personals.