Posts by BennyRop

21) Message boards : Number crunching : turn off hyperthreading? (Message 37893)
Posted 16 Mar 2007 by BennyRop
Post:
One of the Athlon 64 motherboards I've used has DEL being the key to get into the BIOS, and I've seen F10 as the BIOS entry key in the past as well. It'd be nice if everyone standardized on F1..

22) Message boards : Number crunching : Sony plugs PS3 into folding@home (Message 37890)
Posted 16 Mar 2007 by BennyRop
Post:
If you look back in the neighborhood of a month prior to the release of the PS3, there were several mentions of the PS3 being used for F@H; along with mention of the fact that F@H is running the GPU client on ATI x1900+ cards.
The threads usually included a request for why Rosetta and/or Boinc wasn't setup to use GPU clients or ready to use the PS3. :)

Pretty neat to see DC apps moving onto new platforms. :)
23) Message boards : Number crunching : Catagorizing, firewall problems (Message 37865)
Posted 15 Mar 2007 by BennyRop
Post:
Can you locally use the ip# instead of the url?
24) Message boards : Number crunching : Job Sizes (Message 37600)
Posted 7 Mar 2007 by BennyRop
Post:
Anyone else observe this behavior?

I've had my runtime preference set at 24 hours ever since that was possible, and noticed my last few jobs saying they were 6 or 8 hours long. My settings here still say 24 hours. I'm running Boinc 5.4.9 still; if the new server software requires a newer Boinc client to work right.

25) Message boards : Rosetta@home Science : MS-Disease (Multiple Sclerosis) (Message 37511)
Posted 6 Mar 2007 by BennyRop
Post:
One of the things that led me here was the fact that the improvements we're making to the Rosetta code (or helping to make by testing it out) will enable others to come along and setup DC projects for those diseases that aren't represented in the Boinc community.
Recently, projects that various Rosetta team members have been working on progressed to the point where we're starting to do medical research, not just working on improving the Rosetta code.

I don't remember the statements that came up on previous MS requests - but devoting cpu time here while waiting for a dedicated MS Boinc project makes sense to me, if Rosetta doesn't have MS researches onboard.

26) Message boards : Number crunching : Did you see any AMD grandfather 4x4 on BOINC? see this lie (Message 37246)
Posted 28 Feb 2007 by BennyRop
Post:
Itanic in the news

With the comparisons made about the various flavors of cpu from both companies, it looks like it's a matter of seeing which company can execute better as to who will be on top with the release of the Barcelona. I remember the torture of waiting for the Athlon64 cpus to come out after the Opteron was released; and the fun of having an X2 when they came out. Whatever happens, when it comes time to get a new system in a year or two, there should be a pretty decent boost in production available over the current system - from both camps.
27) Message boards : Number crunching : Did you see any AMD grandfather 4x4 on BOINC? see this lie (Message 37189)
Posted 26 Feb 2007 by BennyRop
Post:
Intel makes comparisons, and AMD makes comparisons. They both show their products in the best light. Then they'll point out what was wrong with the comparison made by the other company (optimized code/drivers for their own machine, but ran the opponent's machine with the default WHQL safe drivers/code.) Then you can view reviews of the systems by 3rd parties and see a much more rounded picture of what the strengths of each system are. And the reviews by people that point out the failings of either side tend to be easier to believe as the truth than those that have only pointed out the weaknesses of one side.

It'd be interesting to see independent RAC/$ ratings of top systems and RAC/watt ratings of the same systems to give us a better idea of whether it's worth it to upgrade or when it'll be worthwhile to upgrade. :)
28) Message boards : Number crunching : Trojan boinc installation by rogue member (Message 37091)
Posted 22 Feb 2007 by BennyRop
Post:
During the Nimda outbreak, most ISPs in the 24.x.x.x range seemed interested in Nimda infection reports so they could contact their clients and get the problem taken care of. And since I had someone in my area bring their machine in for cleaning that was told of their infection by our ISP, I got the impression that my communications with the security team bore fruit. With my experience, ISPs are willing to pass on information about infected machines to their clients. I wasn't after their email address, contact info, or trying to get them kicked off the ISP.. just get the machines cleaned up.
29) Message boards : Number crunching : Trojan boinc installation by rogue member (Message 37046)
Posted 21 Feb 2007 by BennyRop
Post:
Since projects have the ip#s of the machines that are running the project under his name, they can contact the ISPs and ask the ISPs to forward a message to the owners of those ISP accounts being used. While they won't give out their email addresses to Boinc or the projects, they should be willing to pass on that information to those affected, getting them to remove the client, or sign themselves up and join a team of their own choosing.

Seti was recently credited as being able to help track down a stolen laptop; so DC projects can be used to help identify systems being used on those projects.
30) Message boards : Number crunching : Odd graphics quirk, possibly (Message 36264)
Posted 7 Feb 2007 by BennyRop
Post:
Stephen:
In case it helps pinpoint the problem - could you track down the brand and model of your video card, chipset on the video card, amount of ram on the card, version of the device driver for the video card, resolution being used, (800x600, 1280x1024, etc), version of Direct X on the system?
31) Message boards : Number crunching : V8 / AMD64 X2 5200 WU comparison (Message 35698)
Posted 29 Jan 2007 by BennyRop
Post:
It shows that you'd need about 50% more AMD X2 5200 cores to match the performance of the unavailable Intel part being shown. We can then decide if hordes of cheaper X2 5200s are better for us than waiting for the release of the new Intel part.


I don't know how you did your math ... but you are wrong.
those big map are useless, an XLS files will be better. You ll need more than 4 times for sure, based on few addition of 20 lines I did ...
50% is in your dreams ...

I don t see the 24hours work load in this bitmap eather, so, back to the counting ... many parts are missing.

you are comparing apple and banana , what matter is granded credit, and only this!
As i expected, it is useless.[/quote]

The v8 used around 200 sec/decoy; the X2 5200 used around 300 sec/decoy for a similar set of WUs. (listed with 1.42 and 1.43 credits/decoy). Other WUs showed a higher ratio, while others showed a lesser ratio that ranged between 1 and 2. It seems to hover around the 1.5 ratio - or a 50% increase. Thus 3 x2 5200 cores should equal 2 of your v8 cores under Rosetta. Or 12 x2 5200 cores to roughly equal the 8 cores.
32) Message boards : Number crunching : V8 / AMD64 X2 5200 WU comparison (Message 35689)
Posted 29 Jan 2007 by BennyRop
Post:
It shows that you'd need about 50% more AMD X2 5200 cores to match the performance of the unavailable Intel part being shown. We can then decide if hordes of cheaper X2 5200s are better for us than waiting for the release of the new Intel part.
33) Message boards : Number crunching : HEY, nVidia just released a new science software language for distributed GPUs! (Message 35594)
Posted 27 Jan 2007 by BennyRop
Post:
It's not about frequency or the crunching power of the cpu.


Incorrect! For linear computation, it is. And that's what's going to drive consumer CPUs. I just don't see the performance of the family computer improving by 32x on a 32 core CPU. And despite the valiant and noble efforts of you dual-core DC record-breakers, DC is done mainly on common family PCs, not multicore servers. I can't see the gamily PC ever having 128 cores, but this evening you can drive to CompUSA and buy a video card with 128 general-purpose devices capable of running fluid dynamics equations independently.

-----
There were various discussions about Moore's observations on The Register and The Inquirer for the past few years on how "Moore's Law" was not being quoted right.

http://download.intel.com/research/silicon/moorespaper.pdf
Here's a pdf of Gordon's article that was linked to after Intel paid $10,000 for a mint copy of the issue of Electronics magazine that he'd written for.

Gordon's article deals with the size, cost, and number of transistors on a die; not the "horsepower" of the cpu.

http://www.vnunet.com/vnunet/news/2127129/gordon-moore-calls-law
http://www.theregister.co.uk/2005/04/13/moores_law_forty/
From an hour long phone interview with various reporters, House gets blamed for the cpu "horsepower" doubling every 18 months observation.

Regardless, things are improving at a wonderful rate for both cpus and gpus - and pretty soon, we'll have gpus providing the house's hot water if ATI & nVidia don't manage to lessen the power usage on their upcoming chips.
34) Message boards : Number crunching : 8-core cruncher? (Message 35585)
Posted 26 Jan 2007 by BennyRop
Post:
Astro:
When this topic was brought up in the past, people that watched a WU on their system pointed out that each decoy was taking different amounts of time. So you can run a WU for 24 hours and give an average time for a WU with short running decoys - but they don't run in a set amount of time.

35) Message boards : Number crunching : HEY, nVidia just released a new science software language for distributed GPUs! (Message 35584)
Posted 26 Jan 2007 by BennyRop
Post:
Correct me if I'm wrong - but Moore's observation was that the number of transisters composing a cpu was doubling every so often. The original time frame has shortened a few times since his observation was first published. 2 years, 18 months, and perhaps we've hit the 12 month mark by now. It's not about frequency.. or the crunching power of the cpu. It wasn't about size - although all 3 of those are also improving and helping keep the observation in effect.

Putting 2,4,8,16, or 32 cores on a cpu is just as valid a way of increasing the ability of the cpu (for multithreaded apps - or those of us using all those extra cores for DC projects) as increasing the frequency and power usage (NetBurst!!!), and has doubled, quadrupled, and octupled the number of transisters on the cpu.

GPUs have had an even shorter doubling period - and now you can have video cards eating up 4 times or more the power usage of the high end cpu they're being driven by.. :) But when they're 1000 times faster than they are today, that still only means that they can do what they do today.. at 1000 times the speed. Everything they haven't been optimized for will run at much reduced speeds.

36) Message boards : Number crunching : What benchmarking program is most similar to Rosetta? (Message 35475)
Posted 25 Jan 2007 by BennyRop
Post:
Back in the 2002 range, when we had 200/266Mhz buses (100/133 DDR) - there was a lot more talk about sync/async issues - along with discussions over which slots on a nVidia chipset motherboard offered the best performance with various memory options. There were noticeable changes on various memory benchmarks back then - between sync and async. So you might want to look at older reviews.

Perhaps taking a look at the DDR ram articles on the tech site (drawing a blank at the moment) that has wonderful articles on CPU makeup. i.e. showing the difference between the PIII, the P4, the Athlon XP, and the Athlon 64 cpus as each appeared on the market.

37) Message boards : Number crunching : HEY, nVidia just released a new science software language for distributed GPUs! (Message 35470)
Posted 25 Jan 2007 by BennyRop
Post:
The folks over at Folding at Home were working with nVidia on getting the client to work on nVidia GPUs but ran into so many problems and poor performance that they ended up switching to the ATI GPUs to get it to work at a decent speed.

We were quoted that the new ATI GPU clients were 20-40 times faster than the old client - but most of us missed the fact that it is 20-40 times faster at what it does; however, it's a subset of the normal cpu client. There was code in the cpu client that couldn't take advantage of the gpu abilities and they didn't use that in the gpu client. It doesn't speed up everything.

It'd be nice to see what optimization is available with the nVidia API - and whether that will prolong the life of some of our older AGP capable crunchers on this project. But it won't be 100 times faster (with the latest nVidia card) than the current Rosetta client while performing everything the current client does.



38) Message boards : Number crunching : Problems with Rosetta version 5.43 (Message 35359)
Posted 23 Jan 2007 by BennyRop
Post:
Are you all using Windows or Linux? I've had no failures on 3 Windows XP Pro systems I am running, but I finally had to detach from Rosetta on *ALL* of my Linux systems (I have two running Ubuntu 6.06 and one running Ubuntu 6.10) as I was seeing nearly 90% failure and these systems were offering rosetta 50% of their time. I didn't mind the tasks that ran 10-15 seconds before failing, but one of these systems is a new P4 dual-core and I had WU's hogging 6 to 8 hours (or 4 days for one WU!!!!) before they failed. Better that I donate that CPU power to another project that produces something useful.

Here is the dual-core's result page: 398561

- Lynn

Lynn:
While I don't use Linux for crunching - do these machines run Rosetta fine if they're running Rosetta 100% of the time? We've had better luck with the windows machines if we set the keep in memory setting to yes. Perhaps someone can confirm or deny if that's been a problem with the Linux systems as well. (trying to eliminate the keep in memory problem and possible problems with interactions between Rosetta and whatever other Boinc app or apps take up the other 50%). Since you've got 3 linux machines with similar problems - have you tried a different version of Linux like Red Hat? (trying to rule out versions not able to identify your hardware properly) Or verified that the 3 haven't managed to get infected?
I'd ask if you'd tested the memory - but that wouldn't appear on 3 machines at once unless you recycled the ram from dead machines..
39) Message boards : Number crunching : How much has your RAC Dropped Since 12/6/06 (Message 35262)
Posted 22 Jan 2007 by BennyRop
Post:
Stu asked a question and was directed here, so I'll respond here.
522,024.70 1602.03 1502.33 604800
His system shows 522k seconds of work done when if it's running 24*7, it should be showing roughly 604,800 seconds. Something's eathing close to 24 hours worth of seconds of processing time away from Rosetta. He's requesting 1600 credits and getting 1500 credits - which may or may not be typical for a Sempron. It looks like something's running in the background.


40) Message boards : Rosetta@home Science : How can we bring more users to the Rosetta project? (Message 34690)
Posted 14 Jan 2007 by BennyRop
Post:
Which of the 4 letter channels has the building on the tram line?


Previous 20 · Next 20



©2022 University of Washington
https://www.bakerlab.org