Threadripper

Discuss all Cyclone Leica Cyclone, Field 360 & Register 360 related issues here.
Post Reply
dhirota
V.I.P Member
V.I.P Member
Posts: 617
Joined: Sun Nov 01, 2009 11:18 pm
Full Name: Dennis Hirota
Company Details: Sam O. Hirota Inc.
Company Position Title: President
Country: USA
Linkedin Profile: Yes
Location: Hawaii, USA
Has thanked: 11 times
Been thanked: 63 times

Re: Threadripper

Post by dhirota »

Phil Marsh wrote:
Sun Mar 29, 2020 9:12 am
If that’s RTC360 data you shouldn’t need to have AutoCloud ticked on import :geek:
Phil

Since I do not pretend to be a Register 360 expert, I just installed the 2020.0.0 version and dropped in Daniel Loney's KingsWay and my only RTC360 B137 data sets and clicked import. Since you mentioned it, I decided to run Daniel's data set with and without the autocloud tick on import on our PC147.

PC147
i9-9980XE (18-cores/36-threads, OVERCLOCKED all Cores to 4.2GHz)) + 128GB RAM + RTX2080Ti

Kings Way with Autocloud tick = 8m:29s
Kings Way w/out Autocloud tick = 8m:29s
B137 with Autocloud tick = 30m:46s

PC166
Ryzen 3990X (64-cores/128-threads, NOT OVERCLOCKED) + 256GB RAM + RTX 2080Ti

Kings Way with Autocloud tick = 7m:46s
B137 with Autocloud tick = 21m:57s

There are probably parameters in the setup that I have not checked on this latest version of Register 360, so I have been consistent on running it for these benchmarks.
Dennis Hirota, PhD, PE, LPLS
[email protected]

mstachoni
I have made 20-30 posts
I have made 20-30 posts
Posts: 20
Joined: Wed Oct 17, 2018 1:49 pm
Full Name: Matt Stachoni
Company Details: Tutor Perini Corporation
Company Position Title: BIM Manager
Country: USA
Linkedin Profile: Yes
Been thanked: 11 times
Contact:

Re: Threadripper

Post by mstachoni »

dhirota wrote:
Thu Mar 26, 2020 10:31 pm
Have decided to change my approach in using the Ryzen 3990X system that I have built. Apparently if you overclock the system, it will draw a substantial amount of power even sitting idle doing nothing as indicated by the task manager.
You should know that the "High Power" columns in Task Manager has nothing to do with the power supply or its ability to adequately feed the system. The High Power column was introduced in Windows 10 version 1809, intended to help users determine which applications are likely to be using more power, based on the application's CPU and GPU load over time. It's perhaps useful for mobile users who are on a battery budget, but for any workstation plugged into a wall, it's meaningless.

Power supplies over 1300W are probably pointless, even for a modern 64 core workstation with multiple GPUs. Figuring a maximum suggested load of 80% of the rating, this gives you a budget of 1040W for your components. This will feed a Threadripper 3990X 64-core CPU on an EATX motherboard with 128GB, 2 x Nvidia GTX 2080TIs, 3 SSDs, and a 7,200RPM hard disk. If you were overclocking, perhaps bump it up a little, but you don't need to go crazy.
- Matt

User avatar
Jason Warren
Administrator
Administrator
Posts: 3449
Joined: Thu Aug 16, 2007 9:21 am
Full Name: Jason Warren
Company Details: Laser Scanning Forum Ltd
Company Position Title: Co-Founder
Country: UK
Skype Name: jason_warren
Linkedin Profile: No
Location: Retford, UK
Has thanked: 89 times
Been thanked: 38 times
Contact:

Re: Threadripper

Post by Jason Warren »

mstachoni wrote:
Mon Mar 30, 2020 11:01 pm
If you were overclocking, perhaps bump it up a little, but you don't need to go crazy.
I know overclocking has its share of risks and wonder if overclocking really makes a difference?
Jason Warren
jason.warren[at]laserscanningforum.com

Image
Please add a LSF Google Review here ;)
https://g.page/laser-scanning-forum-ltd/review?mp

User avatar
smacl
Global Moderator
Global Moderator
Posts: 359
Joined: Tue Jan 25, 2011 5:12 pm
Full Name: Shane MacLaughlin
Company Details: Atlas Computers Ltd
Company Position Title: Managing Director
Country: Ireland
Linkedin Profile: Yes
Location: Ireland
Has thanked: 91 times
Been thanked: 107 times
Contact:

Re: Threadripper

Post by smacl »

Jason Warren wrote:
Fri May 01, 2020 8:43 am
mstachoni wrote:
Mon Mar 30, 2020 11:01 pm
If you were overclocking, perhaps bump it up a little, but you don't need to go crazy.
I know overclocking has its share of risks and wonder if overclocking really makes a difference?
Depends very much on the CPU and load, Intel and single thread stuff benefit the most AFAIK. 20% is typically mooted as the upper limit though I'd say say this is rarely achieved for extended periods and you're also much more likely to crash even if you don't damage the PC. If you're feeling really brave and want to push the envelope, look up delidding. Personally I don't overclock, as in my case as a developer, the rewards don't merit the effort. My goals are to make the code go faster on slower hardware and all the long processing tasks like ongoing regression testing and builds for continuous integration are happening on other PCs.

I find the posts on hardware and software performance on the forum really interesting and am very much of the opinion that software improvements will enhance future performance going forward more than hardware. Still a lot of bottle-necking code that needs to be rewritten to be efficiently multi-threaded and to make better use of GPU compute. Algorithm improvements offer even greater improved performance.

ericguizzetti
V.I.P Member
V.I.P Member
Posts: 172
Joined: Tue Mar 06, 2018 6:56 pm
Full Name: Eric Guizzetti
Company Details: Construction and Engineering
Company Position Title: RealityCapture
Country: USA
Linkedin Profile: Yes
Has thanked: 9 times
Been thanked: 6 times

Re: Threadripper

Post by ericguizzetti »

dhirota wrote:
Mon Mar 25, 2019 1:29 am
jedfrechette wrote:
Sun Mar 24, 2019 4:50 pm
I wouldn't underestimate the value of throwing lots of cores at a problem, even ones that require significant IO. This discussion reminded me of some simple test results I shared on this forum a few years ago.

viewtopic.php?f=88&t=10474&p=55862#p55810

The short version was that a multithreaded application accessing a very inefficient data store (text csv) was much faster than a single threaded application accessing an efficient data store (laz)
Jed and Shane

I did a quick test with a recent 400 acre scanned Hawaiian jungle area using our VZ400i, stop and go, generating a 76GB LAS file (2.3B pts). I took one of the three scans (516M pts, 17GB LAS) and loaded it into our dual Xeon (E5-2637v4) workstation (256GB RAM, 16 threads @3.5GHz) and the same file into our i9-9980XE (128GB RAM, 36 [email protected]) workstation. I used Sequoia 1.163 in both cases to import and create the SPRT cache file and view the imported file on the screen at 4K. Both systems were running all threads at 100% most of the time(sorting and writing) except at the beginning during the read phase.

dual Xeon workstation took 6min:09sec to finish loading, creating the cache and displaying view.

i9-9980XE workstation took 4min:09sec to finish loading, creating the cache and displaying view.

I would expect more threads to be significantly faster, but it might be dividing the file into more smaller pieces which may be increasing the write time. As mentioned earlier, this would be dependent on the application, function, programmer approach.

I processed the same file to load and display it with another program and the application used 5% CPU of the dual Xeon and 100% of the GTX 1080. I quit at 175M pts displayed after 14min:19sec.

____

Am I confused?

Dual Xeons
Performance
# of Cores
4
# of Threads
8
Processor Base Frequency
3.50 GHz
Max Turbo Frequency
3.70 GHz

-vs-

I9
Performance
# of Cores
18
# of Threads
36
Processor Base Frequency
3.00 GHz
Max Turbo Frequency
4.40 GHz

The i9 is faster, more cores and more cache.
Eric Guizzetti
San Diego, CA
Insta: ericguizzetti
Linkedin - https://www.linkedin.com/in/ericguizzetti/

dhirota
V.I.P Member
V.I.P Member
Posts: 617
Joined: Sun Nov 01, 2009 11:18 pm
Full Name: Dennis Hirota
Company Details: Sam O. Hirota Inc.
Company Position Title: President
Country: USA
Linkedin Profile: Yes
Location: Hawaii, USA
Has thanked: 11 times
Been thanked: 63 times

Re: Threadripper

Post by dhirota »

ericguizzetti wrote:
Mon May 18, 2020 10:41 pm
dhirota wrote:
Mon Mar 25, 2019 1:29 am
jedfrechette wrote:
Sun Mar 24, 2019 4:50 pm
I wouldn't underestimate the value of throwing lots of cores at a problem, even ones that require significant IO. This discussion reminded me of some simple test results I shared on this forum a few years ago.

viewtopic.php?f=88&t=10474&p=55862#p55810

The short version was that a multithreaded application accessing a very inefficient data store (text csv) was much faster than a single threaded application accessing an efficient data store (laz)
Jed and Shane

I did a quick test with a recent 400 acre scanned Hawaiian jungle area using our VZ400i, stop and go, generating a 76GB LAS file (2.3B pts). I took one of the three scans (516M pts, 17GB LAS) and loaded it into our dual Xeon (E5-2637v4) workstation (256GB RAM, 16 threads @3.5GHz) and the same file into our i9-9980XE (128GB RAM, 36 [email protected]) workstation. I used Sequoia 1.163 in both cases to import and create the SPRT cache file and view the imported file on the screen at 4K. Both systems were running all threads at 100% most of the time(sorting and writing) except at the beginning during the read phase.

dual Xeon workstation took 6min:09sec to finish loading, creating the cache and displaying view.

i9-9980XE workstation took 4min:09sec to finish loading, creating the cache and displaying view.

I would expect more threads to be significantly faster, but it might be dividing the file into more smaller pieces which may be increasing the write time. As mentioned earlier, this would be dependent on the application, function, programmer approach.

I processed the same file to load and display it with another program and the application used 5% CPU of the dual Xeon and 100% of the GTX 1080. I quit at 175M pts displayed after 14min:19sec.

____

Am I confused?

Dual Xeons
Performance
# of Cores
4
# of Threads
8
Processor Base Frequency
3.50 GHz
Max Turbo Frequency
3.70 GHz

-vs-

I9
Performance
# of Cores
18
# of Threads
36
Processor Base Frequency
3.00 GHz
Max Turbo Frequency
4.40 GHz

The i9 is faster, more cores and more cache.
Eric:
I am not sure of the question?

SUMMARY: I used
dual core Xeon(E5-2637v4)=16 threads, not over-clocked @3.5 GHz turbo @3.7 GHz TIME=6min:09sec.

i9-9980XE=36 threads, over-clocked @4.2 GHz all cores TIME=4min:09sec.

RESULT: i9-9980XE = more, faster cores, less time?
Dennis Hirota, PhD, PE, LPLS
[email protected]

Post Reply

Return to “Leica Cyclone, Field 360 & Register 360”