VGTU project@Home our Team Crunching@EVGA
About VGTU project@Home 
The aim of this project is to provide a powerful distributed computing platform for scientists of Vilnius Gediminas Technical University (VGTU) as well as others Lithuanian academic institutions.
 
Our goal is to persuade scientists to use this kind of computations in their research and to assist them in preparing BOINC applications. We hope that scientific applications from various fields will be attractive to many current and future BOINC community volunteers and this will allow us to accumulate enough computing resources to achieve significant (unachievable otherwise) results.
 
VGTU project@Home preferences
"VGTU @ Home application for grillage optimization" is what I have selected and is what has Tasks ATM.
This application was developed in collaboration with our colleagues from the VGTU Department of Information Technologies (Ph.D. student Mikalojus Ramanauskas and Dr. Dmitrij Šešok).
They are working on of Civil Engineering problem - grillage optimization.
More information on their research (including references to the already published results) you can find in the following forum topic.
Check out the video example!              
This distributed computing application presents an attempt to employ more significant computational resources in their scientific research.              
Note that the current version of application is not using the replication.
Application-specific validation is done for the single job results. 
 
Odd to post that my AMD Ryzen 7 1800X Eight-Core Processor is two times faster than any of my Intel CPUs.
Yep! it seems that AMD CPU is what want here, with HT off it made no difference on the Intel CPU.
https://boinc.vgtu.lt/vtuathome/results.php?hostid=19664&offset=0&show_names=0&state=4&appid
Intel
https://boinc.vgtu.lt/vtuathome/results.php?hostid=12168&offset=0&show_names=0&state=4&appid
https://boinc.vgtu.lt/vtuathome/results.php?hostid=25385&offset=0&show_names=0&state=4&appid
 
Remember that their are No Requirements for Team Members to run POTM.
 
This Project is still an open Sprint over on  This May Be Next Team!

HK-Steve
Awesome, Am loaded up with VGTU now...
bcavnaugh, let's get's get some good points..

Indeed, though I started all my rigs on Leiden Classical after then of the end of the last Sprint.
So I should be back on tonight.

Opolis
I will fire up some VGTU.

+1  Transitioning over to VGTU from WCG ASAP

bill1024
Bill, is that true with the AMD server too? (Being faster than the intel)

I will fire up my 2P and see...View System
 
They are showing 13:15 Hours when starting up will wait 30 Minutes and report back.
https://boinc.vgtu.lt/vtuathome/results.php?hostid=11452
@ 4% time jumped up to +1 Day. 
 
I got my Flu Vaccine yesterday and no big deal.
But today I got my Shingles Vaccine and my arm is killing me.

bcavnaugh
 
Odd to post that my AMD Ryzen 7 1800X Eight-Core Processor is two times faster than any of my Intel CPUs.
 

 8 real cores vs 4 real cores and similar clock rates/IPU efficiencies.
 
 Or are you saying it's twice as fast per work unit?
 

My i7-4930k and 3930k at 4.2ghz does them in 3 hours and a couple min..
Why is your 5960X taking so long Bill?

I have HT off on the two I mentioned.
 
My E5-2670 Sandybridge 8 core xeons with HT on are taking 24 hours.
I believe they run around 3ghz or so and run stock, no OC.
I will turn HT off and see what they do. 24 hours seems long when the others are done in 3 hours.
Could be the memory speed had something to do with it. The xeons mem is 1333 while the other is 1866-2133mhz.

My E5-2670 Sandybridge 8 core xeons with HT on are taking 24 hours.
I believe they run around 3ghz or so and run stock, no OC.
I will turn HT off and see what they do. 24 hours seems long when the others are done in 3 hours.
Could be the memory speed had something to do with it. The xeons mem is 1333 while the other is 1866-2133mhz.
 
 
I turned off HT, the times were cut way down to around 4-4.5 hours
So for sure I would not recommend HT on for this project.
 
Finishing up tasks now, then over to the primenet challenge.

I should be clear, my time was not cut by 4-4.5 hours.
They were cut down TO 4- 4.5 hours.
From around 24 hours down to 4- 4.5 is nearly 20 hours
That is a huge difference.
I have never seen HT slow down tasks that much.
Rule is if you can do more than 2x tasks in lest that 2x time per task, HT is good then
 
 
And heck yeah, under 3 hours is good for the AMD Ryzen, glad it is good at some BOINC projects.
I haven't heard much good about them (BOINC wise) up until now.
 

Bcavnaugh, just curious what CPU clock and memory clock do you have on your 1800x? Your results seem favorable compared to mine and some other Ryzens I could find in the list of top computers.

Mine is a 1700 at 3.7GHz with memory at 2400MHz. I tried with SMT on and off. With off (8 cores/threads) I'm getting times close to yours, but about half the points. With SMT enabled (16 threads) it's similar points but with 2x the runtime. Just wondering if maybe VGTU likes the faster memory?

https://boinc.vgtu.lt/vtu...sults.php?hostid=19237

My PC is a dual boot Win10 and Lubuntu. So I booted into Windows and ran more tasks with SMT off, and I'm seeing the same thing as I did with Lubuntu.

https://boinc.vgtu.lt/vtu...sults.php?hostid=11499

I noticed you have 4 dimms of RAM where I'm only using 2 (I verified I'm using the correct motherboard slots). The number of dimms shouldn't matter, right?

bcavnaugh
As far as I know, but some MB it matters what Slot they are installed.
Maybe it is the CPU 1800X v 1700 (without the X)
This set looks ok https://boinc.vgtu.lt/vtuathome/results.php?hostid=11499&offset=20&show_names=0&state=4&appid
The first set looks https://boinc.vgtu.lt/vtuathome/results.php?hostid=11499&offset=0&show_names=0&state=4&appid 
like the time is as fast but lower Credit.
 Device peak FLOPS 4.54 GFLOPS 1800X 5.28 GFLOPS Link
This was a 300+ But the time was 4+ Hours
https://boinc.vgtu.lt/vtuathome/result.php?resultid=13457628
 
Compare our Computers 1800X and 1700
Measured floating point speed  5278.34 million ops/sec <> 4544.52 million ops/sec
Measured integer speed 18638.85 million ops/sec <> 16827.89 million ops/sec
 
Average processing rate 15.65 GFLOPS <> 7.46 GFLOPS
Listed under Application details Show in the Links 1800X and 1700
 
Have you posted anything on their Forum?
 

No, I haven't posted on their forums as my results are in line with other Ryzen 1700/1700x I could find under Top Computers for the project. Not a big deal, just curious if I could squeeze out a bit more. Thanks for looking into the numbers!