• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

Low ppd on Laptop

R

Randelic

Weaksauce
Joined
May 12, 2009
Messages
88
Having been folding a couple days now and getting with EvilAlchemist to get everything running smooth and monitored with Fahmon I have a couple questions.

Looking at the numbers that my SMP and my GPU clients are putting up as well as my PS3 I am wondering is it normal for the standard client running on my laptop to be so piss poor. (119ppd according to Fahmon from 2 days of folding) If so, is this due to processor speed as well as RAM? If not is there something or some settings I should be looking at?

Btw, this is far from a high end system. Just a standard Dell Inspiron 9300, I believe it is a 1.6ghz and 1gb of RAM.

Thanks in advance everyone
 
Its normal for the clasic client. I have a 2.4 Ghz northwood desktop gets around 200 ppd depends a lot on the WU. Easy to get spoiled with the SMP and GPU outputs, but if the system is going to be on all output helps in the search for a cure. Fold on.
 
That sounds about right for the classic client and a slow cpu like that, when I was running that client on my Q9550 @ 3.4 Ghz I was getting about double that, or more depending on the WU.
 
the classic client sucks horribly for PPD.. why panda group refuses to fix it none of us will ever know.. i mean hell you give the smp client 1920 points per WU that takes 8-9 hours to complete.. but 100-200 ppd to a cpu running the classic client that takes 9 days to complete.. the point system has become a friggin joke with F@H.. who knows one day maybe panda group will pull their heads out of their asses and fix it.. but doubt thats ever going to happen..
 
sirmonkey1985 said:
the classic client sucks horribly for PPD.. why panda group refuses to fix it none of us will ever know.. i mean hell you give the smp client 1920 points per WU that takes 8-9 hours to complete.. but 100-200 ppd to a cpu running the classic client that takes 9 days to complete.. the point system has become a friggin joke with F@H.. who knows one day maybe panda group will pull their heads out of their asses and fix it.. but doubt thats ever going to happen..
Click to expand...
A few things wrong with your analysis. First of all, the SMP client is much more efficient than the uniprocessor client, and does a lot more work. So that's one reason for the higher point value. Another thing is that a fast CPU will still have very good performance with the uniprocessor client. Several hundred PPD for a multi-core system running one instance per core at least, if not over 1000PPD. The only time that the uniprocessor client takes a long time to complete a unit is when running on a slow CPU, in which case the SMP client would do no better anyhow.
 
There are two opinions given about the uniprocessor client so far and they are both correct in their own ways. The uniprocessor client does suck balls for PPD compared to the GPU and SMP clients. The latter two clients have massive bonuses attached to the value of the work units run on them. This is where the point discrepancies come in. Many of us think the point value differences shouldn't be so great but there's not much if anything we can do about it.

As far as your system is concerned, that sounds about right for the clockspeed. I have a couple uniprocessor clients running on my buddy's Q6600@3Ghz. Currently a single client is doing around 220-250PPD on regular work units and 665PPD on the bonus work units on the uniprocessor clients. The bonus work units get really good PPD for the uniprocessor client which I think isn't too far below the crappy A1 core work units on the WinSMP client.

Still, the bonus work units on the uniprocessor client are going to be hit and miss even though lately I've been getting more of them than the non-bonus work units. It's just the way things go. It was the same way with the QMD work units for the P4 as well as the 364 and 600 pointers that used to get great PPD on anything they ran on until the bonus on those work units was dropped. It's been a while but I want to say the 364 and 600 point work units would do about 900PPD per core on an E6400@3.4Ghz. Outside of the QMDs, that's probably been the best PPD for the uniprocessor client ever.

Conclusion: If you can run the SMP client stably and meet deadlines, it's usually the best choice for points. If you can't run that, then the uniprocessor client is your only option for F@H.

 
I'm still in the process of figuring out what all this stuff is and what it does. I mean 3 days in and I am wondering whats wrong with my laptops ppd. *sigh* Also I was making sure my laptop wasn't half retarded as well. Thanks.
 
It seems the that the points are really geared towards the gpu/smp clients. I have been running the uniprocessor client for a few years on all my machines. I recently redid my workstation and decided to run a notfred vm for kicks to see what it would put out in comparison. I have a dual xeon 5410. each client was averaging around 200-250 ppd per core according to fahmon. The single notfred vm is reporting around 2000 ppd for 2 cores. A 4x increase per core

I was pretty shocked by that but kinda expected it from what I was reading about the smp client
Anyway I'm in the process of moving everything over to the vm's for folding on my workstation so I should see a nice bump in PPD.

It can be discouraging to see a lower PPD then your expecting for something, but that should be more motivation to go borg/ upgrade :)
 
You must log in or register to reply here.
Back
Top