21)
Message boards :
Number crunching :
If You Don't Know Where to Put it, Post it here.
(Message 94629)
Posted 16 Apr 2020 by Tom M Post: The few WUs on this host that I see that ran for 12 hours, using the Rosetta v4.15 18 threads with no pauses. I have examined the tasks and they all seem to be "i686" tasks. I thought I had disabled those with the no_alt_platform> but apparently not. ---edit---------- Based on suggestion(s) from other threads I have aborted the "i686" tasks which were all 4.15 versions. Now what? I have gone to NNT while I see what else I can try ---edit--- Tom M |
22)
Message boards :
Number crunching :
If You Don't Know Where to Put it, Post it here.
(Message 94608)
Posted 16 Apr 2020 by Tom M Post: Just looked at the tasks I am crunching this morning. Even though I was set to the "default" (8 hours?) the majority of the tasks are now running near 12 hours. Is this a bubble or a trend? Tom |
23)
Message boards :
Number crunching :
If You Don't Know Where to Put it, Post it here.
(Message 94567)
Posted 15 Apr 2020 by Tom M Post: Of course my comment is only about choosing between Intel and AMD CPUs for CPU only workloads. Point. For the average individual user who probably has a several year old Intel cpu "you dance with who brung ya" For all the new fangled stuff as long as we aren't struggling with some version of the AVX instructions the evidence reported in the reviews does support it. Tom M |
24)
Message boards :
Number crunching :
The most efficient cruncher rig possible
(Message 94566)
Posted 15 Apr 2020 by Tom M Post: Just some added settings. PBO on in bios, memory at 3600mhz. No other changes, most current bios and chipset drivers. :) |
25)
Message boards :
Number crunching :
If You Don't Know Where to Put it, Post it here.
(Message 94523)
Posted 15 Apr 2020 by Tom M Post: As of today AMD CPUs are simply the best for distributed computing, no matter what, be it Ryzen, Threadripper or EPYC. Their total computing power is simply unmatched.If it is CPU only. Sorry, "data" |
26)
Message boards :
Number crunching :
The most efficient cruncher rig possible
(Message 94518)
Posted 15 Apr 2020 by Tom M Post: I am using a 3800x AMD with a 1080 video card, 16gb memory and a couple ssd's. I run Boinc at 100% with 8 core/16 threads but to keep it cool and use less power I am going into power and sleep, then under the power plan settings I set it to power saver, then under advanced settings I set maximum processor state to 90%. Temps change from 78c(its throttling at 80c)to 55-57c and the cpu power use drops from 80 watts to 32 watts with Boinc and firefox running. This is according to Ryzen Master. +1 |
27)
Message boards :
Number crunching :
If You Don't Know Where to Put it, Post it here.
(Message 94517)
Posted 15 Apr 2020 by Tom M Post: As of today AMD CPUs are simply the best for distributed computing, no matter what, be it Ryzen, Threadripper or EPYC. Their total computing power is simply unmatched.If it is CPU only. It also depends on if you can run all the date being processed in the gpu in question. If not then "he" is right. If I understand the way Rosetti is approaching the processing using larger databases instead of discrete data packages then that design for data-analysis will not show the true benefits of the massive parallel processing offered by gpus. Tom M |
28)
Message boards :
Number crunching :
Threadripper and Ryzen (and EYPC)
(Message 94514)
Posted 15 Apr 2020 by Tom M Post: The other CPU? More like the main one these days! That is a great question. I wish I had a robust answer. On the grounds of theory alone more memory channels should allow a "high cpu/thread count" system to access and run "faster" on the same speed memory chips. The brand/model of memory chips I am running are leftover from my Amd 2990wx experiment and so do not appear to be "slowing" my machine down visibly BUT I probably can't tell because Rosetta is a "time invariant" task project and it usually takes upwards to 6 weeks to have your RAC stabilize. I have been crunching Rosetti at this high level maybe 3 weeks. I have no clue what it will stabilize at. I am hopeful to get up well past 50,000 but I am not running all my CPU threads on Rosetta. My cpus run a mix of 4 projects with Rosetta running 18 threads. I also have all the turbo/turbo-like functions of the CPU disabled in the bios. So it is munching along at about 3.2 Ghz. Tom M |
29)
Message boards :
Number crunching :
The most efficient cruncher rig possible
(Message 94512)
Posted 15 Apr 2020 by Tom M Post: Yes, I agree that the Ryzen 7 and 9 are one of the top contenders. I promise to NEVER forget that issue. I bought an Amd Threadripper 2990WX (32c/64t) and discovered to my horror that 26 threads was the most productive I could get. It was due to the memory model which required 2 hops to access memory for half the cpus.
Unless inflation/stagflation were to kick in the "cost of money over time" (eg. interest) is amazingly low. Tom M |
30)
Message boards :
Number crunching :
Threadripper and Ryzen (and EYPC)
(Message 94446)
Posted 14 Apr 2020 by Tom M Post: This project is clearly cpu oriented and seems to be utilizing a lot of "external" to the application data sources instead of downloading a stand alone app and a packet of data to process and then uploading the results. Assuming my characterization of the project is fairly accurate would you say a AMD 2700 (8c/16t)[or 2700x] be a better buy over a 2600 (6c/12t)? There have been significant price declines in the 2600/2700 new cpu market. I do understand the Ryzen/Threadripper 3000 series is significantly better in terms of power efficiency (I run a 3950x). But if I was building a "new" system and were sensitive (but not completely rigid) to my capital costs which would be "better"? Tom M |
31)
Message boards :
Number crunching :
Threadripper and Ryzen (and EYPC)
(Message 94444)
Posted 14 Apr 2020 by Tom M Post: I thought I would start up a thread to talk about the "other" CPU. :) The topic has been popular over at Seti@Home maybe it will be popular here. Tom M |
32)
Message boards :
Number crunching :
If You Don't Know Where to Put it, Post it here.
(Message 94443)
Posted 14 Apr 2020 by Tom M Post: About 1/3 of the top 100 crunchers here are AMD systems. Many are Threadripper or EPYC server systems. I even found an AMD 3950x (16c/32t) in the top 100 listing which does give me hope for mine :) Tom M |
33)
Message boards :
Number crunching :
The most efficient cruncher rig possible
(Message 94442)
Posted 14 Apr 2020 by Tom M Post:
I have had good luck with B350/B450/X470 chip motherboards. I run several different brands including an Asus ROG Crosshair VII Hero (Amd 3900x/upgraded to a 3950x) which has been VERY reliable but is limited to 6 gpus. And a MSI B350 chip MB. Take a look at the reviews looking for the highest end VRM (power) stuff for your price range. If you are planning on exceeding 6 gpus you will need to avoid Asus. I have a mid-grade Asus Prime (B450) MB out on loan but while I was using it, it seemed to be reliable and a good fit for 24/7 crunching. HTH, Tom M |
34)
Message boards :
Number crunching :
If You Don't Know Where to Put it, Post it here.
(Message 94441)
Posted 14 Apr 2020 by Tom M Post: [quote]This topic has reminded me of things I wanted to say over the last few weeks. Sid, Thank you for a courteous and thoughtful response. +1 Tom M |
35)
Message boards :
Number crunching :
Problems and Technical Issues with Rosetta@home
(Message 94410)
Posted 13 Apr 2020 by Tom M Post: I had nothing but errors on both the i686 applications on my Ryzen. Gave up on Rosetta and moved to Einstein. Discovered later that you can set a flag in cc_config.xml to ignore alternate platforms. Huh, I didn't notice mine apparently. Tom M |
36)
Message boards :
Number crunching :
Minirosetta 3.73-3.78
(Message 94407)
Posted 13 Apr 2020 by Tom M Post:
It is hidden but it is a last resort browser if/when the original EDGE was broken or a website doesn't tolerate anything else (yes, incredible but possible). Tom |
37)
Message boards :
Number crunching :
What are your tips for new Rosetta@home users?
(Message 94335)
Posted 13 Apr 2020 by Tom M Post: Due to how power saving and turbo boost is implemented in processors, it is much more efficient to run 50% of the time over every core than to use 50% of the cores. How about disabling the Turbo like I do? |
38)
Message boards :
Number crunching :
Discussion on increasing the default run time
(Message 94237)
Posted 12 Apr 2020 by Tom M Post:
+1 Thank you Moderator for a clear exposition on the "nuts and bolts". Tom |
39)
Message boards :
Number crunching :
Task start optimisation suggestion
(Message 94236)
Posted 12 Apr 2020 by Tom M Post: I had a quick look at some slot directories after getting the 4.15 Rosetta app on a Raspberry Pi today. Hi, I would rename this thread to something about Raspberry Pi. You have two good points and this kind of i/o optimization may also be good for cellphone versions too. Tom |
40)
Message boards :
Number crunching :
If You Don't Know Where to Put it, Post it here.
(Message 94235)
Posted 12 Apr 2020 by Tom M Post: I just took a look at the current list of "Mega Crunchers" systems on the Rosetti@Home. In the top 20 I don't recognize a single "type" of system that was/is common at Seti@Home. Instead I see a huge number of "very high thread count" systems (upwards to 192 cores/threads). Another thing to note is the top producers are not unanimous in running the default 8 hours setting. Instead a number of them are running 3 hours or less tasks. If 8 hour tasks really do stress the the server(s) less and I have every reason to believe they do. Perhaps we should start "nagging" (in private or setup a public shame list?) for every top producer (say the top 100) not running at least the 8 hour default to start running the default. Presumably they would get the same credit scores but access the server(s) less often? And given the volume they may be running it should lower the load on the server(s) significantly. Respectfully, Tom M |
©2024 University of Washington
https://www.bakerlab.org