Messages in this thread | | | Date | Thu, 23 Mar 2000 22:07:03 -0600 (CST) | From | Mike Field <> | Subject | Athlon: Graphics and proc-top |
| |
A few comments about the Athlon under linux 2.2.12 and 2.3.51.
Under 2.3.51, this was the result of a time trial:
[1] Done iter_comp -B2000 hex_trig1 > trash 205.360u 0.060s 1:42.94 199.5% 0+0k 0+0io 715pf+0w
The actual time was approximately 103s. I have no idea why top thinks the machine is SMP.
Under either 2.2.12 or 2.3.51 the floating point performance was about the same (approximately 1.5 times faster than a PII 400MHz SMP machine I use). Performance on compilation was better - about 1.75 times faster.
However, graphics is another matter. Under linux 2.2.12, X is *very* slow - in fact a total disaster (using AcceleratedX and a G400 MAX card). We are talking a factor of 10 here. Performance under 2.3.51, though not a disater, is not great - in fact it is worse than under the PII 400MHz machine (by about 20%). (AGP 2X). Although I haven't compared 2.2.12 to 2.3.51 under Xfree3.3.5, I did compare 2.3.51 under Xfree3.3.5 and AcceleratedX and Xfree3.3.5 was about 20% slower. The graphics performance uising the PIII on the tests I was running was about the same using either a G400 MAX or a G400 - suggesting that under the PIII the limit was either floating point performance (what I expected) or saturation of the memory bus (possible, but unlikely).
Any suggestions/comments as to what might be going (wr)on(g) here?
[Asus motherboards, ample SDRAM, details if needed. 1600x1200 resolution, 24 bit color - similar results with lower resolution or Hi-color]
Please cc to mf@uh.edu as I am not presently on the Kernel mailing list.
Mike Field
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/
| |