grenley@nsc.nsc.com (George Grenley) (10/26/88)
Numbers, Numbers - but no proof... In article <6410@daver.UUCP> dlr@daver.UUCP (Dave Rand) writes: >In article <7275@nsc.nsc.com> grenley@nsc.nsc.com.UUCP (George Grenley) writes: >> [ dhrystone benchmarking stuff] >>This is about right, depending on which test conditions you specify... >>Our 30 meg part is around 17K... >I get 18,335... >Dave Rand Dave, That's nice, but how? Specifically, please publish the code you ran (all of it, including library calls, esp strcpy and strcmp). Also please document exactly what hardware you ran on, the wait-state environment, and the compiler used. Then we'll see if the number can be reproduced. These are supposed to be deterministic systems, you know. Regards, George Grenley
dlr@daver.UUCP (Dave Rand) (10/26/88)
In article <7320@nsc.nsc.com> grenley@nsc.nsc.com.UUCP (George Grenley) writes: >Numbers, Numbers - but no proof... > >In article <6410@daver.UUCP> dlr@daver.UUCP (Dave Rand) writes: >>>Our 30 meg part is around 17K... >>I get 18,335... > >exactly what hardware you ran on, the wait-state environment, and the >compiler used. 532 DK board. 0 waits, 30 Mhz. CTP 3.3 (National's compiler). Check with Jeff Baer for details - apparently, all of the FAE's from National have this demo! Drop me mail if you want a copy of the tar. Dave Rand {pyramid|hoptoad|sun|vsi1}!daver!dlr -- Dave Rand {pyramid|hoptoad|sun|vsi1}!daver!dlr