An insiders' exuberant, albeit candid, account of how a backwater outpost in IBM's empire managed to overcome its shortcomings and produce a new family of computers that not only proved a bestseller but also earned its creators a prestigious Malcolm Baldridge National Quality Award. Bauer and his associates were all involved in the Silverlake Project, a crash program undertaken in the mid-1980's by Big Blue's Rochester, Minn., unit to retrieve its flagging fortunes in mid-range systems--a protean category encompassing machines whose price tags run from $15,000 to $1 million apiece. At the outset, according to the authors, the outlook was bleak, mainly because IBM Rochester had lost touch with its prospects as well as customers and was turning out hardware designed to please in-house engineers. By applying a ten-point set of operating principles, however, the facility was able to build and introduce the AS/400 series in just 28 months (against a normal cycle of five or more years). Among other valuable insights gained from the project, the authors cite visionary leadership, empowerment of employees, and close working relationships with buyers. By their account, error-free design, market research, and realistic priorities are also musts if schedules are to be met and goals achieved. IBM Rochester clearly did a lot of things right, since AS/400 revenues alone would make it the world's second largest computer enterprise (trailing only the parent organization). Whether the lessons to be learned from the Silverlake Project are applicable to other concerns lacking Big Blue's researchers is an open question. There's real interest, though, in the possibility that Rochester could serve as a paradigm for the restructuring to which IBM publicly committed itself last November.