WU processing time getting a tad longer now...

[BespinReactorShaft]

...or are the recent SETI Classic work units taking longer to complete?

They seem to be taking 30mins or so longer, since two days ago. I'm running two threads at once on a P4-3GHz with HT.

 

[OhioDude]

Could be that you've just been getting a lot of lower angle workunits, lately. They seem to come in batches for me sometimes.

 

[BespinReactorShaft]

Yeah I hope so. I've also got a Pentium M on SETI but it's beating my desktop P4 right now. I'm just raring to hit that 5k milestone.

 

[Assimilator1]

I've just checked my SETIQ & I only have 3 out of 136 WUs which are Low Angle Range WUs

Not many slow WUs here

 

[Soggysocks]

I'm having the same problem. My network production seems down about 20%. I'm still trying to figure things out. WU's don't seem any different then normal.

If I come with something...I'll post it on.

 

[Soggys1stBrat]

I don't know about you guys, but mine seem to go a little faster lately. As long as I stay off the PC and just let it crunch I seem to get more done than it used to. Of course like everyone was saying, maybe I'm just getting a bunch af better angles.

 

[Spacehead]

Do you have a program that'll log past WU info? That would be the first thing i'd check. I haven't been getting very many "user friendly" WUs lately, but alot more in the range of 0.394 - 0.428 which is actually nearer the normal angle range. Maybe before you where getting higher AR WUs & it just seems like your times are slowing down.

Here are some of my times for you to compare to. I also have a HT P4 3.0Ghz:
3:09:35 (0.394 AR)
2:35:45 (1.090 AR)
2:10:09 (5.386 AR)

 

[Assimilator1]

I take it those WU times are for 2 WUs at the same time?

 

[Spacehead]

There's no other way to run it, is there?

 

[Soggysocks]

Originally posted by: Spacehead
There's no other way to run it, is there?

 

[BMdoobieW]

Originally posted by: Spacehead
Here are some of my times for you to compare to. I also have a HT P4 3.0Ghz:
3:09:35 (0.394 AR)
2:35:45 (1.090 AR)
2:10:09 (5.386 AR)

I also have a P4 3.0E GHz w/HT running two WU's at the same time. Here are my most recent results:

3:22 (0.011 AR)
2:50 (0.421 AR)
2:50 (0.439 AR)
2:31 (0.827 AR)
2:31 (0.854 AR)
2:07 (2.964 AR)

 

[Soggysocks]

Originally posted by: BMdoobieW

Originally posted by: Spacehead
Here are some of my times for you to compare to. I also have a HT P4 3.0Ghz:
3:09:35 (0.394 AR)
2:35:45 (1.090 AR)
2:10:09 (5.386 AR)

I also have a P4 3.0E GHz w/HT running two WU's at the same time. Here are my most recent results:

3:22 (0.011 AR)
2:50 (0.421 AR)
2:50 (0.439 AR)
2:31 (0.827 AR)
2:31 (0.854 AR)
2:07 (2.964 AR)

This is close to what I was running..........now

add about 45 mins. to those and its what i'm doing. This is really frustrating since I haven't tinkered or changed any bios settings and haven't installed any new software. Mcafee did find a virus W32/Pate.b......
I've cleaned all systems . Nothing new other than that. :disgust:

 

[Wiz]

spyware? check the processes? virii seldom come alone any more - my kids machine got a downloader and bargainsbuddy last night, right through AVG7 & Spybot 1.3 which were both up to date

took me over an hour to get it cleard up :|

then I found it's browser protections were turned off (spybot) - turned them back on and enabled Tea Timer. (protects against system settings changes)

 

[Smoke]

The times reported by the various tools (all created before HT systems) do not accurately report the average CPU times. You need to divide the reported average times reported by SetiQueue, SetiDriver, etc. on Hyper-Threaded Systems by "2" to get an accurate comparative number.

The numbers below are pulled from The Enterprise - TeAm AnandTech SetiQueue:

Avg CPU ?.. RPD ?.. Name
1h52m ?.. 8.66 ?.. GolfBall (This my brother's new P4 2.9GHz system)
1h54m ?.. 11.25 ?.. Smoke - Centrino
2h23m ?.. 20.00 ?.. Vic
2h30m ?.. 9.50 ?.. .Tonto1
2h36m ?.. 17.07 ?.. Crazee
2h37m ?.. 8.30 ?.. Scubado GA-8PE667U 2.80 P4 (Smoke #06)
2h47m ?.. 7.84 ?.. GregD Epox EP-8RDA 2400XP (Smoke #09)
2h47m ?.. 7.53 ?.. ToshaD - 2400+XP
2h50m ?.. 36.84 ?.. Smoke - Server
2h51m ?.. 15.87 ?.. Smoke - Main WS

"Smoke - Server" AND "Smoke - Main WS" are hyper-threaded systems so their AVG CPU should be calculated as 1h25m and 1h25.5m respectively ... easily heading this list as the FASTEST SYSTEMS.

 

[Smoke]

The times reported by the various tools (all created before HT systems) do not accurately report the average CPU times. You need to divide the reported average times reported by SetiQueue, SetiDriver, etc. on Hyper-Threaded Systems by "2" to get an accurate comparative number.

The numbers below are pulled from The Enterprise - TeAm AnandTech SetiQueue:

Avg CPU ?.. RPD ?.. Name
1h52m ?.. 8.66 ?.. GolfBall (This my brother's new P4 2.9GHz system)
1h54m ?.. 11.25 ?.. Smoke - Centrino
2h23m ?.. 20.00 ?.. Vic
2h30m ?.. 9.50 ?.. .Tonto1
2h36m ?.. 17.07 ?.. Crazee
2h37m ?.. 8.30 ?.. Scubado GA-8PE667U 2.80 P4 (Smoke #06)
2h47m ?.. 7.84 ?.. GregD Epox EP-8RDA 2400XP (Smoke #09)
2h47m ?.. 7.53 ?.. ToshaD - 2400+XP
2h50m ?.. 36.84 ?.. Smoke - Server
2h51m ?.. 15.87 ?.. Smoke - Main WS

"Smoke - Server" AND "Smoke - Main WS" are hyper-threaded systems so their AVG CPU should be calculated as 1h25m and 1h25.5m respectively ... easily heading this list as the FASTEST SYSTEMS.

 

[Assimilator1]

Well thay are accurate times but they don't factor in 2 WUs produced togther ,the WU times shown is accurate

 

[Rattledagger]

Uhm, the average cpu-times is correctly displayed as they is, since a dual or HT-machine have 48 cpu-hours per day.

The average crunch-times or run-times on the other hand should be divided by #cpu's.

 

[JC]

Hey Smoke.....I guess you posted that on a HT machine, I see two instances of 'reply' there

 

[BespinReactorShaft]

I'm using SETI Driver to check the times. My desktop P4 is still averaging 2:30 - 3:15 hours (double threaded). The notebook P4 is still happily going at 2:00 - 2:30 per WU (single threaded). However I have a whopping 100 WU cache on the notebook so I guess it'll be some time before it also starts hitting the "slow" WUs.