Which is faster 3.03 or 3.08?

[vaughan]

Anyone know which is faster Seti cmd line v3.03 or v3.08?

P4 or Athlon XP? Why?

[AMDave]

vaughan,

3.03 and 3.08 should be the same.
3.08 simply fixed a possible securityexploit.
Notes can be read here.
http://setiathome2.ssl.berkeley.edu/unix.html

The last significant code optimisation was in 3.03

Found on the web page noted above (just before the first table):

Athlon CPU users please note: we have tried building the client with gcc 3.0.4 with the compile flag -march=athlon and the same optimization flags as the i686 compiled Linux client: -O3 -fno-exceptions -fno-rtti -funroll-loops -fomit-frame-pointer -fno-strict-aliasing -fno-common -pipe and there is no difference in timing between these two versions. There is nothing special in the client code that can take advantage of the Athlon CPU features.

As for hardware platforms, it can depend upon your choice of hardware. The CPU based stats are here:
http://setiathome2.ssl.berkeley.edu/stats/cpus.html
but you really have to take those with an understanding of the erroneous CPU registrations (pinch of salt).

The only way to really tell is to set a target, take a flag to them and start them off...

I don't have either, so I can't help you there.

Did you read the APC Mag article (Feb-05 ed., Pg 36) yet ? They pitted the latest 3.46GHz P4 ExtremeEdition Intel CPU against the Athlon FX-55, and still the Athlon came out in front in all of the benchmarks in spite of the FSB increase on the intel die. :!: Ouch :!: It's a safe bet that there are going to be some project management changes at Intel in the near furture.

[vaughan]

Haven't seen that article. I don't subscribe to APC, just PCUser, PCAuthority and PCWorld.
Trouble with the Intel Extreme Edition and AMD 64 FX series is the damage they cause to the hip pocket nerve. I cannot afford either of them.

[Beerknurd]

^^bump^^

[Ototero]

I read somewhere that an Intel would have to be running at 4.9GHz to match the Athlon FX-55 in benchmarks.

[Empty_5oul]

they do seem really over-priced. i do not see how they will get many sales at £400+.

i see your point ototero as AMD get a lot more per cycle than Intel, i imagine a lot of this could change as the race is on for a dual core chip.

[meckano]

3.03 is way faster. Go here:
http://www.clubplus.net/~dbrown/seti...fo-classic.htm
and search for 3.03, I wrote something short on that.
I've done the testing. Read the whole page, I've done too much testing. lol
- In fact, I just did it again in last month, the testing/proofing.

[meckano]

Here are my results from the 2400+ as a 2800+:


OBVIOUS LOWER TIME HERE WITH 3.03, go lower down for 3.08

2005 Jan 17 06:26pm 41.44,17.62 0.428 TAKE1 2:44
2005 Jan 17 03:41pm 37.976,19.33 0.423 TAKE1 2:44
2005 Jan 17 12:57pm 53.25,9.42 0.443 TAKE1 2:44
2005 Jan 17 10:13am 44.663,12.41 0.558 TAKE1 2:43
2005 Jan 17 07:31am 33.808,9.12 1.141 TAKE1 2:25
2005 Jan 17 05:05am 41.477,25.21 0.652 TAKE1 2:43
2005 Jan 17 02:22am 11.837,8.09 3.39 TAKE1 2:24
2005 Jan 16 11:58pm 9.771,8.35 1.129 TAKE1 2:25
2005 Jan 16 09:33pm 13.16,19.81 0.438 TAKE1 2:45
2005 Jan 16 07:45pm 36.145,28.07 8.343 TAKE1 2:23
2005 Jan 16 07:44pm 35.629,22.73 0.526 TAKE1 2:45
2005 Jan 16 07:44pm 35.207,21.07 0.454 TAKE1 3:53
2005 Jan 16 07:44pm 9.78,28.59 0.394 TAKE1 4:03
2005 Jan 16 07:54am 20.259,8.14 0.444 TAKE1 2:44
2005 Jan 16 05:11am 32.697,28.59 0.394 TAKE1 2:56
2005 Jan 16 02:15am 52.553,9.42 0.443 TAKE1 2:45
2005 Jan 15 09:23pm 32.533,28.59 0.394 TAKE1 3:00
2005 Jan 15 06:24pm 8.287,13.91 1.194 TAKE1 2:28
2005 Jan 15 03:55pm 8.261,14.02 0.528 TAKE1 2:47
2005 Jan 15 01:08pm 54.147,21.18 2.026 TAKE1 2:27
2005 Jan 15 10:42am 38.29,23.35 0.548 TAKE1 2:46
2005 Jan 15 07:56am 38.29,23.35 0.548 TAKE1 2:46
2005 Jan 15 05:10am 38.262,23.22 0.539 TAKE1 2:46
2005 Jan 15 02:24am 38.262,23.22 0.539 TAKE1 2:46
2005 Jan 14 11:39pm 38.103,22.48 0.499 TAKE1 2:47
2005 Jan 14 08:52pm 38.129,22.59 0.504 TAKE1 2:50
2005 Jan 14 06:01pm 37.973,21.75 0.417 TAKE1 2:54
2005 Jan 14 03:05pm 38.04,21.75 0.7 TAKE1 2:52



OBVIOUS HIGHER TIME WITH 3.08



2005 Jan 14 09:42am 42.308,17.62 0.428 TAKE1 3:13
2005 Jan 14 06:29am 42.263,17.62 0.428 TAKE1 3:13
2005 Jan 14 03:16am 5.003,18.08 0.426 TAKE1 3:13
2005 Jan 14 12:04am 6.071,20.55 0.442 TAKE1 3:21
2005 Jan 13 05:14pm 6.255,20.11 0.436 TAKE1 3:30
2005 Jan 13 01:09pm 35.667,11.71 0.621 TAKE1 3:27
2005 Jan 13 09:42am 35.213,13.31 0.502 TAKE1 3:30
2005 Jan 13 06:12am 20.9,8.38 0.625 TAKE1 3:27
2005 Jan 13 02:45am 20.676,8.19 0.638 TAKE1 3:28
2005 Jan 12 11:17pm 43.942,8.09 0.646 TAKE1 3:25
2005 Jan 12 06:07pm 19.314,27.06 0.69 TAKE1 3:19
2005 Jan 12 02:48pm 41.778,27.29 0.727 TAKE1 3:22
2005 Jan 12 11:26am 23.604,9.38 0.873 TAKE1 3:11
2005 Jan 12 08:15am 35.173,24.35 0.55 TAKE1 3:23
2005 Jan 12 04:52am 52.957,9.42 0.443 TAKE1 3:29
2005 Jan 12 01:23am 1.682,28.57 0.887 TAKE1 3:12
2005 Jan 11 09:07pm 1.244,27.82 0.825 TAKE1 3:22
2005 Jan 11 05:45pm 1.114,27.4 0.786 TAKE1 3:27
2005 Jan 11 02:10pm 0.912,26.63 0.718 TAKE1 3:22
2005 Jan 11 10:48am 0.734,25.85 0.655 TAKE1 3:23
2005 Jan 11 07:25am 6.559,10.16 1.322 TAKE1 2:47
2005 Jan 11 04:38am 0.694,26.01 0.403 TAKE1 3:41
2005 Jan 11 12:57am 0.47,26.01 0.403 TAKE1 3:41
2005 Jan 10 09:15pm 0.357,26.01 0.403 TAKE1 4:14

Stats courtesy of SetiQueue

[vaughan]

Where do you download 3.03 from? I could only find 3.08 for windows command line version at Berkeley.

[meckano]

http://www.clubplus.net/~dbrown/seti/helpinfo.htm

Hit first small link at top of page then
read first large section. CLI section