New addition to F@H :)

Yeggstry

Yeggstry

Yeggstry

Associate
Joined
8 Aug 2005
Posts
1,101
Location
Reading, UK
Hi guys, just thought I'd say hi and say that I've started folding at home on my main machine at home (Sempron 3000+, 1Gb RAM). Its not much, but every little helps :) I am planning on getting one of the new Penryns come February so I will be able to help more but for now this is my contribution :)

From the looks of it will be taking me about 4-5 days for each WU, with this first one at 500 credits. Is this normal for the processor I have and if not what can I do to speed it up? I've followed the guide (which is excellent by the way :)) to the letter, I guess it will give me incentive to upgrade shortly.
 
If you have followed the guide to the letter then everything should be running as it should.

As for the actual time it is taking, 4 or 5 days seems like a lot. Unless the machine is only on part of the time.
 
SiriusB said:
If you have followed the guide to the letter then everything should be running as it should.

As for the actual time it is taking, 4 or 5 days seems like a lot. Unless the machine is only on part of the time.
Click to expand...

Well I got the following on the logs:

[19:27:06] Completed 0 out of 250000 steps (0)
[19:57:08] Timered checkpoint triggered.
[20:27:08] Timered checkpoint triggered.
[20:29:50] Writing local files
[20:29:50] Completed 2500 out of 250000 steps (1)

and since its 100 "chunks" then I would guess its 100 hours which is just over 4 days. It may be more since I will be switching it off some nights. I'll give you a benchmark after its run overnight tonight :)

EDIT: As i was typing this it did the 2nd "chunk" and gave me a PPD of 126.69. I'm definately going to have to get a new machine :P
 
Last edited:
I've got an idea of what it might be anyway, it would also be my main machine and would probably involve a Q9450 and a Geforce 9 series card :) But I definately will, would need some advice on overclocking too:)
 
Yeggstry said:
I've got an idea of what it might be anyway, it would also be my main machine and would probably involve a Q9450 and a Geforce 9 series card :) But I definately will, would need some advice on overclocking too:)
Click to expand...


You cashing in someone's life insurance? That lot's gonna cost more than just an arm and a leg!
 
The Penryns are at about £700 a go, and I bet my left fruit that the 9 series DX10 cards will be around the 450 quid mark.

That's over a grand for two components.
 
SiriusB said:
The Penryns are at about £700 a go, and I bet my left fruit that the 9 series DX10 cards will be around the 450 quid mark.

That's over a grand for two components.
Click to expand...
I think it'll be just like it always is. The new top-end stuff will briefly be crazy expensive, as the QX9650 is at the moment, then after some time the price creeps down to match the current top-end stuff. The now-defunct stuff drops down and either becomes budget stuff or goes out of production. I expect to see prices around the same place they are now. A decent Penryn quad will go for about as much as a decent Kentsfield quad does now.
 
Welcome to the team :)


I'll be looking to upgrade soon too - most likely I'll end up going for value as I can just never quite find the motivation to part with so much more cash for something unless it's REALLY a load better. I guess it depends on what sort of results the Penryn's get over say a Q6600 :o
 
rich99million said:
Welcome to the team :)


I'll be looking to upgrade soon too - most likely I'll end up going for value as I can just never quite find the motivation to part with so much more cash for something unless it's REALLY a load better. I guess it depends on what sort of results the Penryn's get over say a Q6600 :o
Click to expand...

Speaking to someone last night, and they said people are getting up to 5GHz out of the new Penryns. Granted that'll be on phase or something. That's 6.5K PPD right there :p
 
The unlocked Penryns (QX9xxx) will clock very well.

The locked versions (ie the majority) are looking to be FSB limited in current boards. So that, along with the low multipliers (since they'll be 1333mhz) means that they may not be quite the extreme clockers we (including me) have been hoping for.
 
Welcome to the team :)

126ppd sounds fair for that chip. My lappy gets roughly the same with those WUs on one core of a T2600.
 
Thanks for the welcome guys :)

I've completed my first WU last night - 500 points, took about 3 days. However, I have got a slight problem:

[03:19:10] Completed 250000 out of 250000 steps (100)
[03:19:10] Writing final coordinates.
[03:19:12] Past main M.D. loop
[03:20:12]
[03:20:12] Finished Work Unit:
[03:20:12] - Reading up to 2111496 from "work/wudata_01.arc": Read 2111496
[03:20:12] - Reading up to 52520 from "work/wudata_01.xtc": Read 52520
[03:20:12] goefile size: 0
[03:20:12] logfile size: 168464
[03:20:13] Leaving Run
[03:20:13] - Writing 2590196 bytes of core data to disk...
[03:20:13] ... Done.
[03:20:13] - Shutting down core
[03:20:13]
[03:20:13] Folding@home Core Shutdown: FINISHED_UNIT
[03:20:18] CoreStatus = 64 (100)
[03:20:18] Unit 1 finished with 96 percent of time to deadline remaining.
[03:20:18] Updated performance fraction: 0.959309
[03:20:18] Sending work to server


[03:20:18] + Attempting to send results
[03:20:18] - Reading file work/wuresults_01.dat from core
[03:20:18] (Read 2590196 bytes from disk)
[03:20:37] Couldn't send HTTP request to server (wininet)
[03:20:37] + Could not connect to Work Server (results)
[03:20:37] (171.65.103.162:8080)
[03:20:37] - Error: Could not transmit unit 01 (completed December 16) to work server.
[03:20:37] - 1 failed uploads of this unit.
[03:20:37] Keeping unit 01 in queue.
[03:20:37] Trying to send all finished work units


[03:20:37] + Attempting to send results
[03:20:37] - Reading file work/wuresults_01.dat from core
[03:20:37] (Read 2590196 bytes from disk)
[03:20:40] Error: Got status code 503 from server
[03:20:40] + Could not connect to Work Server (results)
[03:20:40] (171.65.103.162:8080)
[03:20:40] - Error: Could not transmit unit 01 (completed December 16) to work server.
[03:20:40] - 2 failed uploads of this unit.


[03:20:40] + Attempting to send results
[03:20:40] - Reading file work/wuresults_01.dat from core
[03:20:40] (Read 2590196 bytes from disk)
[03:20:59] Couldn't send HTTP request to server (wininet)
[03:20:59] + Could not connect to Work Server (results)
[03:20:59] (171.65.103.100:8080)
[03:20:59] Could not transmit unit 01 to Collection server; keeping in queue.
[03:20:59] + Sent 0 of 1 completed units to the server
[03:20:59] - Preparing to get new work unit...
[03:20:59] + Attempting to get work packet
[03:20:59] - Will indicate memory of 1023 MB
[03:20:59] - Connecting to assignment server
[03:20:59] - Successful: assigned to (171.65.103.162).
[03:20:59] + News From Folding@Home: Welcome to Folding@Home
[03:20:59] Loaded queue successfully.
[03:21:00] Error: Got status code 503 from server
[03:21:00] + Could not connect to Work Server
[03:21:00] - Error: Attempt #1 to get work failed, and no other work to do.
Waiting before retry.
[03:21:13] + Attempting to get work packet
[03:21:13] - Will indicate memory of 1023 MB
[03:21:13] - Connecting to assignment server
[03:21:14] Couldn't send HTTP request to server (wininet)
[03:21:14] + Could not connect to Assignment Server
[03:21:14] - Successful: assigned to (171.64.122.136).
[03:21:14] + News From Folding@Home: Welcome to Folding@Home
[03:21:14] Loaded queue successfully.
[03:21:15] - Receiving payload (expected size: 239676)
[03:21:18] - Downloaded at ~78 kB/s
[03:21:18] - Averaged speed for that direction ~153 kB/s
[03:21:18] + Received work.
[03:21:18] Trying to send all finished work units


[03:21:18] + Attempting to send results
[03:21:18] - Reading file work/wuresults_01.dat from core
[03:21:18] (Read 2590196 bytes from disk)
[03:22:18] Couldn't send HTTP request to server (wininet)
[03:22:18] + Could not connect to Work Server (results)
[03:22:18] (171.65.103.162:8080)
[03:22:18] - Error: Could not transmit unit 01 (completed December 16) to work server.
[03:22:18] - 3 failed uploads of this unit.


[03:22:18] + Attempting to send results
[03:22:18] - Reading file work/wuresults_01.dat from core
[03:22:18] (Read 2590196 bytes from disk)
[03:22:31] Couldn't send HTTP request to server (wininet)
[03:22:31] + Could not connect to Work Server (results)
[03:22:31] (171.65.103.100:8080)
[03:22:31] Could not transmit unit 01 to Collection server; keeping in queue.
[03:22:31] + Sent 0 of 1 completed units to the server
[03:22:31] + Closed connections
[03:22:31]
Click to expand...

It appears that it isn't sending the WU back to Standford, but I'm not sure why. I have gone to the web browser for that URL and it works fine, and I didn't have anything like PeerGuardian enabled at the time. Any ideas?

Also, could someone explain the whole stomping/parping thing to me? Obviously since I've just joined I'm going to be gaining over other people but I guess I should be only notifying the ones that are active? :)
 
The servers will go down occasionally. During the week they're fixed pretty quickly, but at w-ends it can take a bit longer. The client will continue to try to upload any finished units every so often, so they will get there eventually.

Stomping is overtaking someone.

Parping is issuing a warning to someone that you are about to overtaken them. It's not really necessary at the lower levels of the table, but as you get higher up it is a polite thing to do.
 
You must log in or register to reply here.
Back
Top Bottom