Posts by Zxian

1) Message boards : Number crunching : CPU Comparison question (Message 47130)
Posted 28 Sep 2007 by Profile Zxian
Post:
Perhaps they are attached to multiple projects.

Could be... although I wonder how many Q6600 systems there are out there, and how many are Rosetta only.
2) Message boards : Number crunching : CPU Comparison question (Message 47128)
Posted 28 Sep 2007 by Profile Zxian
Post:
And it's still going up... I easily broke 1200 with this machine even without overclocking. I'm not sure why other Q6600 systems aren't doing better. Most of the ones I see are below 1000.
3) Message boards : Cafe Rosetta : Personal Milestones (Message 47047)
Posted 26 Sep 2007 by Profile Zxian
Post:
My personal computer, Eclipse, just broke the 1700RAC marker, and just passed 50K total credit. =D
4) Questions and Answers : Unix/Linux : Lower Credit/time for Linux than Windows? (Message 46944)
Posted 24 Sep 2007 by Profile Zxian
Post:
Yup - that line was there. I had checked that last week when I first realized this problem.
5) Questions and Answers : Unix/Linux : Lower Credit/time for Linux than Windows? (Message 46939)
Posted 24 Sep 2007 by Profile Zxian
Post:
The CPU speed of the FC7 machine is at 1.8GHz now. I was digging through various configuration files, and found out that dispite everything being "as it should", the system was still only sitting at 1.2GHz when Rosetta was running. The GNOME CPU frequency monitor showed one speed, while /proc/cpuinfo showed another - very strange.

It's running at 1.8GHz now, so we'll see if the credit scores go up accordingly. *crosses fingers*
6) Questions and Answers : Unix/Linux : Lower Credit/time for Linux than Windows? (Message 46894)
Posted 24 Sep 2007 by Profile Zxian
Post:
That could have to do with the lower measured benchmarks. If they're reading lower than normal, and the WU comes out to an average amount, the system probably thinks that it's done a "good" job, and will report a higher estimated credit (since it did well).
7) Questions and Answers : Unix/Linux : Lower Credit/time for Linux than Windows? (Message 46882)
Posted 24 Sep 2007 by Profile Zxian
Post:
I can say with certainty that the CPU on the Linux machine is indeed working at full capacity. That isn't the issue at hand.

It's just strange for me to see the granted credit be this low for this system. For my windows system, it seems as though 3 hours of work gives about 35-40 credit on average, while the Linux system is given 20-30.

Does anyone know if there are significant compiler optimizations that are present in the Windows executable as opposed to the Linux binary?
8) Questions and Answers : Unix/Linux : Lower Credit/time for Linux than Windows? (Message 46761)
Posted 22 Sep 2007 by Profile Zxian
Post:
The Windows machine has the same amount of cache - it's the exact same CPU. I built both systems in the past couple of months, and used the E2160 for both. Perhaps the Windows machine is indicating the cache "per CPU", even though the L2 cache on the E2160 is shared.

I've disabled SpeedStep on both systems, since they're running R@H anyways, and never get a chance to slow down. Therefore, they're both running at 1.8GHz. This is confirmed by CPU-Z on the Windows machine, and from the information stored in /proc/cpuinfo on the Linux machine.

Both top and Process Monitor indicate that there are two processes (Rosetta_something) that each use 100% of a core.
9) Questions and Answers : Unix/Linux : Lower Credit/time for Linux than Windows? (Message 46743)
Posted 21 Sep 2007 by Profile Zxian
Post:
If you have a look at my systems, I've got two computers that are pretty much the same - both based on the 965 chipset, E2160 CPUs. The main difference is that the system running Server 2003 is using 2GB DDR2-800 and the other (which runs FedoraCore7) is 1GB DDR2-677.

Windows Machine
FC7 Machine

I realize that the RAC of the FC7 computer has yet to reach it's maximum, but if you look at the results for the two computers, the Windows system seems to be getting a higher credit per time than the Linux system.

Has anyone else had this kind of scenario? Both machines are running 24/7, and I find it hard to believe that the difference in RAM speeds would be the cause of this (I doubt R@H is pushing 6GB/s of RAM access).
10) Message boards : Number crunching : Pentium D 925 3.0 GHz for Rosetta? (Message 46667)
Posted 20 Sep 2007 by Profile Zxian
Post:
I had a P4D 3.2 system a while back that ran R@H 24/7. I remember that it got up to about 530 RAC before I had to return the system. It had only been running R@H for a few months, so I'm not sure if the RAC had leveled off yet.

The P4D's crunch pretty well - if you can get one for cheap, go for it.
11) Message boards : Number crunching : More compute errors than usual... (Message 46559)
Posted 19 Sep 2007 by Profile Zxian
Post:
I've been noticing that I've been getting more WUs finishing with a Compute Error status than usual. This is happening on multiple machines, so it's not limited to one of my systems.

http://boinc.bakerlab.org/rosetta/result.php?resultid=106089082
http://boinc.bakerlab.org/rosetta/result.php?resultid=105906754
http://boinc.bakerlab.org/rosetta/result.php?resultid=106491513

They all seem to be coming from the CAPRI14 WUs. Any thoughts on this?
12) Message boards : Number crunching : CPU Comparison question (Message 46548)
Posted 18 Sep 2007 by Profile Zxian
Post:
Here's the stats from my Q6600 system. I broke the 1000 credit barrier on stock speeds, and I use the system for at least 5 hours each day (not necessarily at full potential mind you). I only built that system about a month ago, so there's no way my RAC has leveled off yet.

Before the system outage, it had pushed upwards of 1200. During that time, I overclocked it to 3.0GHz (a 25% overclock). We'll see where it level's off at, but it's currently within the top 40 overall computers on R@H. I'm not sure why, but my Q6600 seems to be doing better than other comparable systems.

http://boinc.bakerlab.org/rosetta/show_host_detail.php?hostid=586752
13) Message boards : Number crunching : loss of credit post crash (Message 46025)
Posted 11 Sep 2007 by Profile Zxian
Post:
Since the system outage, I'm getting far, far more WUs with the 20-credit "thank you" than before. I actually think that I never saw this before the outage. I've tried to "fix" this by making my machines run for only 3 hours per WU, but this isn't an ideal solution.
14) Message boards : Number crunching : Problems with Rosetta version 5.78 (Message 45967)
Posted 11 Sep 2007 by Profile Zxian
Post:
I've also had several WU's come out with only 20 granted credit, regardless of how long the WU actually ran for.

This is on several different computers with different versions of Windows (XP, 2003).
15) Message boards : Number crunching : Naive Quad Core to be released in August. (Message 45694)
Posted 2 Sep 2007 by Profile Zxian
Post:
Hello folks, sorry to butt in, but I found this thread very interesting.

Please advise on the following:-

Server built...

Case, 600w PSU, 6x Samsung 250gb, LG DVD RW, FD, 4gb RAM, 4x Samsung 250gb NAS, Windows Home Server O/S.

I have been using AMD for past 16 years but I am not biased. would just like a good recommendation for a quad - server will be in heavy use, both internally & externally.

Recommendations/comments appreciated.

Thanks


http://www.legitreviews.com/article/521/1/

Honestly - Barcelona is still going to be behind the mark when released. I'd stick with Intel, since it seems as though AMD is going to be the next 3dfx.

AMD systems generally have higher memory bandwidth, but Intel systems have faster computational speed. If you know what your limitations are going to be, then you have your answer.

On that note - I do a lot of simulations for my research (molecular dynamics, DFT, etc), and I have yet to see our dual quad-core Xeon X5355 systems be limited by memory requirements (at least not that we've been able to notice). Performance on our machines scales as expected with such a setup, so we haven't seen any additional bottlenecks thus far.
16) Questions and Answers : Windows : 64bit application for AMD64 processors? (Message 44347)
Posted 26 Jul 2007 by Profile Zxian
Post:
@The_batman - What's preventing you from running R@H on your XP-X64 machines? I've got 20 machines, all running 64-bit linux, and they're crunching away at Rosetta. Rosetta is a much more beneficial project than SETI IMO.
17) Questions and Answers : Windows : Credit Claimed vs Granted (Message 44344)
Posted 26 Jul 2007 by Profile Zxian
Post:
Sorry to say, but a 2GHz AMD system (which equates to a 3200+ or 3400+) isn't really much to write home about in this day and age. Is the AMD system dual core as well, or is it an "old" single core CPU?

The claimed credit vs granted credit is a little confusing for me as well. I've got several server machines that run R@H when they're not being used by my research group. Those machines often get more credit than they claim, but my lower end machines around the lab or at home typically get less credit granted than claimed. This is probably due to the fact that the cluster machines are using Xeon X5355 CPUs, while the other systems are PIII and AthlonXP CPUs (Xeons have plenty of cache).

The AMD 3200+ CPU has 512K of cache, while your new dual core Intel (I'm guessing an E6400 or E4400) has 2MB or 4MB of cache. That alone will contribute to better simulation runs.
18) Questions and Answers : Windows : System with Deepfreeze - How to save work units? (Message 44273)
Posted 26 Jul 2007 by Profile Zxian
Post:
I've got a couple of terminals here that we use mainly for surfing the web and basic document writing. I was hoping to run R@H on them, but the only problem is that the systems are locked by a program called DeepFreeze, which essentially restores the system to an "original state". Along with this original state is an empty work queue in BOINC.

Does anyone know of a workaround for this? Would installing BOINC to an "unfrozen" partition of the system work?






©2024 University of Washington
https://www.bakerlab.org