TUIDs are unique identifiers designed for high performance computing. They are realized using native integer datatypes which allows efficient use as key fields in key/value stores and SQL databases. They provide guaranteed uniqueness across threads/processes/machines without IPC or synchronization. TUIDs can be monotonically increasing so their natural sort order is the order of creation.

tuid64_s *myctx = tuid64_init(NULL);
uint64_t mytuid;
 
do {
    mytuid = tuid64_r(myctx);
    ...
    
} while (! is_done());

See tuid.h for usage.

Copyright 2013 David Sparks. See the file LICENSE.txt included with the TUID distribution for details.