Welcome new [H] members to DC

pututu

pututu

[H]ard DC'er of the Year 2021
Joined
Dec 27, 2015
Messages
3,093
Welcome Holdolin to our [H] DC (distributed computing) family. Please drop a note to say hi to the rest of your team mates.

If you don't mind a quick note about yourself, how you get started into DC or anything that you like to share. The sky is the limit ;)

Feel free to ask any questions in the DC forum.
 
A quick note? how about a low B :p Ok for real though I'm one of those techno geeks. My wife always says our house looks more like Microcenter than a house. I been workin and playin with then since about 1986 when I learned BASIC on an (now) old Apple IIe. Since then I just be messin with a lot of the new tech and figuring out what I like and don't.

Professionally I've been an engineer my whole life. Ever since my first oil change with my dad when I was a kid. We won't even discuss when that actually was hehe. Since then though I worked with bigger and bigger stuff. Hey, I'm a typical guy. I love big toys. There's something to be said about brining engines to life that power generators larger then most small towns or propulsion engines that are so big when you do an inspection it includes opening the inspection door and walking in.

I've since retired but my love of big toys never went away. Now I build big computers to donate to the causes I believe in. A life's worth of work and learning and I still find myself learning nearly every day, mostly about the computers I work with but hey, I'm still learning. Speaking of learning, if any of you have hints for dealing with moody teenage girls lemmie know. All my wife says is "it's normal", and while that may be true it still makes me wanna pull my hair out at times.

I'm also a huge sci-fi fan. A peek at my rig names would prolly give that away though. Frankly, if I were less than half a century into life I'd prolly take my engineering love/expertise to the aerospace field, but I don't feel like competing with teenagers for jobs like that. Besides, a life of making the best choices I could has allowed me to not bother with another job after the Navy so I guess I'll just stick to watching SG-1, TNG and VOY when I'm not working in my dungeon (I found the basement is the best place to keep all that stuff as it's away from the kids/cats and I can run any cables I need without worry.

Ok, seen enough? Bet you don't say the sky is the limit again haha.

Oh, I also play bass in my basement, hence the low B joke at the beginning.
 
Woot, woot! Glad to see your first post in DC forum. Welcome!

We are a bunch of nerdy, I mean a mix of nerdy and non-nerdy, old and young [H]ordes doing a lot of distributed computing and sometimes boost about the milestones reached in a project.

And if you like a bit of competition (watch your blood pressure), we sometimes run both BOINC and non-BOINC projects helping our team in Formula BOINC, BOINC pentathlon and sometimes friendly competition between teams like TAAT vs [H] in folding at home schedule to start on Jan 23rd. Just don't pull your hairs out (if you still have some :)) during the challenges.

Glad to see you running MilkyWay to keep us in a good position in Formula BOINC marathon. (y)
 
Welcome Holdolin ! I'm from another team (TeAm Anandtech) but I stick around here too since there's tons of useful info, everybody seems happy to share info and nobody has told me to leave (yet? :ROFLMAO: )
I run a lot of Milkyway@home and was looking through various computers on there to compare task times - do you have a Radeon VII Pro? If so if you have any info on average task times/PPD or anything like that for Milkway@home or any other project I'd definitely be interested in hearing about it. I've been wanting one of those cards since before they released, and haven't been able to find any distributed computing benchmarks for them anywhere. They probably do super well at Milkyway@home with the 1:2 Double Precision ratio.
 
From anandtech article directly, in theory should almost double, well about 6.5/3.5 = 1.86x exactly the output in MilkyWay. Holdolin, if you have one on this beast, please share some result. Thx.

1610814714406.png


Icecold, you are always welcome here. We are a bunch of friendly people ;)
 
Icecold said:
Welcome Holdolin ! I'm from another team (TeAm Anandtech) but I stick around here too since there's tons of useful info, everybody seems happy to share info and nobody has told me to leave (yet? :ROFLMAO: )
I run a lot of Milkyway@home and was looking through various computers on there to compare task times - do you have a Radeon VII Pro? If so if you have any info on average task times/PPD or anything like that for Milkway@home or any other project I'd definitely be interested in hearing about it. I've been wanting one of those cards since before they released, and haven't been able to find any distributed computing benchmarks for them anywhere. They probably do super well at Milkyway@home with the 1:2 Double Precision ratio.
Click to expand...
Yes, I have the Pro VII's. They do amazing. I run .25 GPU (4 instances) and my time per WU is ~30 seconds, so 30/4 is...well I don't math, I have these computers for all that :p I've taken a run at some Einstein work, but am currently draining my que to get back to MW. The only real problem I've ran into is the <radio edit> drivers. The only Ubuntu driver says it's for 16.04.4, but the way Ubuntu runs things they keep updating the kernel, so the dkms install blows up cause the kernel is newer than expected. This can be dealt with with the command: sudo-apt mark hold linux-generic linux-image-generic linux-headers-generic. Once I nail down which kernel actually works I will be happy to post such if anybody would like to know. That is the other reason I'm letting my work cache run dry. Already had a catastrophic failure that resulted in 800 lost WUs from Einstein (thankfully their system picked up that there appeared to be lost WU's so sent them to me agian) so when I start testing which kernel works for the 18.04 release It will be with no WUs outstanding.
 
Thank you for the information - that definitely is a great card(or cards) for MW! For comparison, a Firepro S9150 does 4 tasks at a time in around 1 minute 30 seconds, so the VII Pro is 3 times as fast. I think the normal VII is around twice as fast as a Firepro S9150, but I don't have a Radeon VII(pro or otherwise). The VII Pro sounds like a great card.
 
Icecold said:
Thank you for the information - that definitely is a great card(or cards) for MW! For comparison, a Firepro S9150 does 4 tasks at a time in around 1 minute 30 seconds, so the VII Pro is 3 times as fast. I think the normal VII is around twice as fast as a Firepro S9150, but I don't have a Radeon VII(pro or otherwise). The VII Pro sounds like a great card.
Click to expand...
Glad to provide the info =)
 
Holdolin, you got a special request from TAAT forum to see if you can run one MW task (credit 227.5x) per gpu instead of four concurrently to provide the average run time per task accurately? Stat collected is to be added here. I presume you run at default clock speed?
You got a special recognition ;)
 
You must log in or register to reply here.
Back
Top