Discrete　manufacturing　workshops　are　confronted　with　problems　of　processing　diverse　products　and　strict　realtime　requirements　for　data　service　calculation　and　manufacturing　equipment,　which　makes　it　difficult　to　provide　realtime　feedback　and　compensation.　In　this　study,　a　high-availability,　high-performance,　and　high-concurrency　digital　twin　reference　model　was　constructed　to　satisfy　a　large　number　of　manufacturing　requirements.　A　multiterminal　real-time　interaction　model　and　information　aging　classification　rules　for　virtual　and　physical　models　were　established.　Moreover,　a　multiterminal　virtual　interaction　model　was　proposed,　and　a　generalized　distributed　computing　service　digital　twinning　system　was　developed.　This　digital　twin　system　was　considered　a　machine　tool　box　processing　line　as　an　actual　case.　Consequently,　a　full　closed-loop　manufacturing　process　digital　twin　platform　for　physical　request　service,　real-time　response,　and　quality　information　feedback　from　iterations,　which　provides　case　guidance　for　subsequent　factory　digital　twin　systems,　was　realized.　The　proposed　system　can　satisfy　the　requirements　of　multidevice　big　data　computing　services　in　modern　manufacturing　plants,　as　well　as　multiplatform,　low-latency,　and　high-fidelity　information　visualization　requirements　for　managers.　Thus,　this　system　is　expected　to　play　an　important　role　in　information　factories.