1 Testing methodology: Using 'make clean;sync;time make' 3 times in a row. Disregarding first
2 attempt. Record "real" time.
6 dual (4 virtual cpus) P4 Xeon 3Ghz (no -j): 46.5s 46.1s
7 dual (4 virtual cpus) P4 Xeon 3Ghz (-j4): 19.9s 19.8s
8 P4 1800: 1m10.5s 1m10.8s
9 dual PPC 970 2Ghz OSX: 1m11.7s 1m11.5s
10 dual PPC 970 2Ghz OSX (-j2): 41.3s 41.3s
12 using gcc 3.3.3 (all using linux, unless otherwise specified)
14 athlon 1700 smp (no -j): 1m8s 1m5s
15 athlon 1700 smp (-j2): 34.4s
16 athlon64 3000 64-bit: 36.9s 35.3s
19 itanium 733 debian: 2m41s
20 athlon 1700 smp (no -j): 1m1s
21 athlon 1700 smp (-j2): 32.5s
24 Sparc IIe 500 (solaris 9): 4m45s
26 P4 2400 HT (-j2): 40.5s
27 athlon64 3000 64-bit: 36.5s 35.2s