Posts by Joe
log in
1) Message boards : News : Firewall Failure (Message 22818)
Posted 530 days ago by Joe
Or more likely the old firewall is rejecting my ISP?
2) Message boards : News : Firewall Failure (Message 22817)
Posted 530 days ago by Joe
I can't reach the site with my home ISP since the firewall failure. Can with an anonymous proxy and typing this note. Seems something about the firewall my ISP doesn't like...

Tried restarting but no help.
3) Message boards : Number crunching : Computation errors (Message 22742)
Posted 550 days ago by Joe
Are the wu's both 32 and 64 bit. I'm running an XP 32 bit system and it will run for 10-15 hours just fine and then everything errs out for a few hours

Like this

<core_client_version>7.6.9</core_client_version>
<![CDATA[
<message>
Invalid access to memory location.
(0x3e6) - exit code 998 (0x3e6)
</message>
<stderr_txt>
Collatz Conjecture Sieve 1.21 Windows i686 for OpenCL
Written by Slicker (Jon Sonntag) of team SETI.USA
Based on the AMD Brook+ kernels by Gipsel of team Planet 3DNow!
Sieve code and OpenCL optimization provided by Sosiris of team BOINC@Taiwan
Collatz Config Settings:
verbose 1 (yes)
kernels/reduction 32
threads 2^8 (256)
lut_size 17 (1048576 bytes)
sieve_size 2^26 (4149496 bytes)
sleep 1
cache_sieve 1 (yes)
reducecpu 0 (no)
Platform NVIDIA
Device 01995328
Max Dimensions 3
Max Work Items 1024 1024 64
Max Work Groups 1024
Max Kernel Threads 1024
Device Vendor NVIDIA Corporation
Name GeForce GTX 560
Driver Version 358.87
OpenCL Version OpenCL 1.1 CUDA
actual threads 256
The verification of the steps for 2772435020442332670207 failed. The OpenCL device calculated 4278191051 vs 971 steps by the CPU. Processing aborted.
Error: (-998)Unknown at 1708 of RunSteps
Error: GPU steps do not match CPU steps. Workunit processing aborted.
17:29:05 (2452): called boinc_finish

</stderr_txt>
]]>

The wu that validate look like this

<core_client_version>7.6.9</core_client_version>
<![CDATA[
<stderr_txt>
Collatz Conjecture Sieve 1.21 Windows i686 for OpenCL
Written by Slicker (Jon Sonntag) of team SETI.USA
Based on the AMD Brook+ kernels by Gipsel of team Planet 3DNow!
Sieve code and OpenCL optimization provided by Sosiris of team BOINC@Taiwan
Collatz Config Settings:
verbose 1 (yes)
kernels/reduction 32
threads 2^8 (256)
lut_size 17 (1048576 bytes)
sieve_size 2^26 (4149496 bytes)
sleep 1
cache_sieve 1 (yes)
reducecpu 0 (no)
Platform NVIDIA
Device 01995328
Max Dimensions 3
Max Work Items 1024 1024 64
Max Work Groups 1024
Max Kernel Threads 1024
Device Vendor NVIDIA Corporation
Name GeForce GTX 560
Driver Version 358.87
OpenCL Version OpenCL 1.1 CUDA
actual threads 256
Start 2772434215560661696512
Stop 2772434222157731463168
Best 2772434219886451048511
Highest steps 2036
Total steps 59338755199178
Average steps 581
CPU time 0.46875 seconds
Elapsed time 2368.36seconds
21:35:03 (2444): called boinc_finish

</stderr_txt>
]]>

It seems always errs out at 1708 of run step, I'm wondering if its because its downloading 64 bit wu's
4) Message boards : Number crunching : Optimizing Collatz Sieve (Message 21548)
Posted 829 days ago by Joe
Slicker

Got 100 or so results that did not validate last month so reluctantly have moved on to other (less paying) projects that don't seem to be in a constant state of flux.

You do a great job and I understand that constant optimization will prove/disprove the conjecture more quickly but frankly having to fiddle with config files, system setting, etc every few months is not worth the effort in my case.

I have run other projects for 10+ years without modifying a single file.

Possibly, you could have a "Collatz for Dummies" application that would just work for those of us that enjoy DC but don't have the time or inclination to keep up with the latest flavor of the month.

Thank you for making this a very enjoyable project and good luck
5) Message boards : Number crunching : Computation Errors (Message 20799)
Posted 927 days ago by Joe
Another:

Stderr output

<core_client_version>7.4.42</core_client_version>
<![CDATA[
<message>
(unknown error) - exit code -5 (0xfffffffb)
</message>
<stderr_txt>
Collatz Conjecture v6.08 Windows i686 for OpenCL
Based on the AMD Brook+ kernels by Gipsel
Optimizations provided by Sosirus
verbose=1
threads=9
items_per_kernel=21
kernels_per_reduction=8
sleep=1
lut_size=15
Config: verbose=1 items_per_kernel=2097152 kernels_per_reduction=256 threads=512 sleep=1
Platform: NVIDIA
Device: 003E8228
OpenCL context created
OpenCL command queue created
OpenCL program created
OpenCL program copiled
OpenCL kernel created
Max Work Item Dimensions: 3
Max Work Item Size: 1024 1024 64
Max Work Group Size: 1024
Max Kernel Work Group Size: 1024
Init complete.
Allocate memory complete.
Device Vendor: NVIDIA Corporation
Address Bits: 32
Name: GeForce GTX 560
Driver Version: 327.23
Device Version: OpenCL 1.1 CUDA
Max Clock: 1620
Compute Units: 7
Alignment: 4096
Constant Buffer: 4096
WU Name: solo_collatz_2398022483333544411136_1649267441664
Start: 2398022483333544411136 Stop: 2398022484978516885504
Read checkpoint complete.
Using: verbose=1 items_per_kernel=2097152 kernels_per_reduction=256 threads=512 sleep=1
Error in call to clEnqueueReadBuffer for kernelSteps (-5) at line 1608 Out of resources
17:46:28 (3956): called boinc_finish

</stderr_txt>
]]>

When they were erring out after 2-3 seconds it was no problem but after an hour or two it may make more sense to just dedicate that machine to another project....
6) Message boards : Number crunching : Computation Errors (Message 20771)
Posted 929 days ago by Joe
The new open CL failed after 2 hours

Stderr output

<core_client_version>7.4.42</core_client_version>
<![CDATA[
<message>
(unknown error) - exit code -5 (0xfffffffb)
</message>
<stderr_txt>
Collatz Conjecture v6.08 Windows i686 for OpenCL
Based on the AMD Brook+ kernels by Gipsel
Optimizations provided by Sosirus
Config: verbose=1 items_per_kernel=262144 kernels_per_reduction=8 threads=64 sleep=1
Platform: NVIDIA
Device: 003E8228
OpenCL context created
OpenCL command queue created
OpenCL program created
OpenCL program copiled
OpenCL kernel created
Max Work Item Dimensions: 3
Max Work Item Size: 1024 1024 64
Max Work Group Size: 1024
Max Kernel Work Group Size: 1024
Init complete.
Allocate memory complete.
Device Vendor: NVIDIA Corporation
Address Bits: 32
Name: GeForce GTX 560
Driver Version: 327.23
Device Version: OpenCL 1.1 CUDA
Max Clock: 1620
Compute Units: 7
Alignment: 4096
Constant Buffer: 4096
WU Name: solo_collatz_2397964125955036282880_1649267441664
Start: 2397964125955036282880 Stop: 2397964127604303724544
Read checkpoint complete.
Using: verbose=1 items_per_kernel=262144 kernels_per_reduction=8 threads=64 sleep=1
Error in call to clEnqueueReadBuffer for kernelSteps (-5) at line 1608 Out of resources
00:10:29 (3428): called boinc_finish

</stderr_txt>
]]>
7) Message boards : Number crunching : Errors on CUDA workunit (Message 20722)
Posted 933 days ago by Joe
I'm out of the 6.04 that seemed to have a 3% fail rate.

With the 6.05 its been 100% fail so far.

Output file:

<core_client_version>7.4.42</core_client_version>
<![CDATA[
<message>
(unknown error) - exit code -1 (0xffffffff)
</message>
<stderr_txt>
Collatz Conjecture v6.05 Windows i686 for CUDA 5.5
Based on the AMD Brook+ kernels by Gipsel
Using optimizations provided by Sosirus
Name GeForce GTX 560
Compute 2.1
Parameters --device 0
Start 2397903104588703072256
Checking 1653562408960 numbers
Numbers/Kernel 131072
Kernels/Reduction 64
Numbers/Reduction 8388608
Reductions/WU 197120
Threads 512
Using: verbose=1 items_per_kernel=131072 kernels_per_reduction=64 threads=512 sleep=1
cudaSafeCall() failed at CollatzCudaKernel11.cu:362 : device not ready
21:57:48 (2672): called boinc_finish

</stderr_txt>
]]>
8) Message boards : Number crunching : Windows 10 (Message 19969)
Posted 1189 days ago by Joe
Actually I did find an adaptor after I found out the Asus Evo R2 I got was missing the IDE. It booted into my old XP setup if I don't try a dual boot with only a few glitches so I can continue to have access to my ancient Windows 95 applications that I use for work.

I'm playing with the Windows 10 pre-release and have found it much lass buggy than the Windows 2000 pre-release I tested awhile back...and it runs All the Boinc projects with no problem.
9) Message boards : Number crunching : Windows 10 (Message 19956)
Posted 1192 days ago by Joe
Motherboard finally gave up the ghost on one of my XP machines. Funny thing but it turns out I could not just replace a 2002 vintage board with a 2014ish board..what ever happened to IDE's..

So being the adventurous(or cheap SOB) that I am I downloaded the Windows 10 preview edition. It set up fairly easily and I downloaded Boinc and I attached to Collatz. Boinc found the two 750Ti's on the machine and is running them at about 6000 seconds without doing anything on my part.

So the question is ...being as I know diddly about software and don't want to learn much, can I improve on those times?
10) Message boards : Number crunching : Something change again? (Message 19671)
Posted 1309 days ago by Joe
Nothing changed as far as I can tell.

I noticed the wu's I received prior to Jun 21 took about 4000 and 5000 seconds to complete on the 560 and 750. Those received after Jun 22 ate not taking about 5500 and 8500 seconds to complete....
11) Message boards : Number crunching : Something change again? (Message 19660)
Posted 1310 days ago by Joe
I have automatic update turned off. So that wasn't the problem.

Actually it appears Bionic on XP is limited to 8 slots. I cleared the tasks in progress and it loaded another WU and is running as normal...my bad
12) Message boards : Number crunching : Something change again? (Message 19655)
Posted 1311 days ago by Joe
I've been running a 750T and a GTX 560 on a windows XP machine. Tonight the TI stopped being recognized so only one wu is being run. SETI, Einstein, GPU Grid, and Prime Grid work as normal detecting both cards and running wu's on each. I sure didn't change anything but knowing managements tendency to constantly fiddle with the apps I'm wondering if I missed something....
13) Message boards : Number crunching : Large WU Feedback (Message 19313)
Posted 1361 days ago by Joe
I wont tie up a GPU for 2 days 2 hours on one work unit. If it errors out or the machine dies for some reason then I have lost 2 days of credits and from my point of view that is an unacceptable risk..YMMV
14) Message boards : Number crunching : Large WU Feedback (Message 19298)
Posted 1362 days ago by Joe
See large WU' # 5550636 that I aborted after 10 hours 20% completed...do the math

I guess this week I'm supposed to figure out something new to make this project work like it did last week..

It seems to me that you care about this project and are doing everything from your point of view to make it succeed.

However, from my point of view, as someone who has been around for awhile, about 70 projects .. as a suggestion, find something that works and stick to it..

If this project requires more than simply adding it in the Boinc menu then say so...or at least classify it as an alpha
15) Message boards : Number crunching : Large WU Feedback (Message 19295)
Posted 1363 days ago by Joe
I noticed that in put in a couple examples..
16) Message boards : Number crunching : Large WU Feedback (Message 19292)
Posted 1363 days ago by Joe
The mini work units get about 1/2 as much credit as the old solo wu's running about the same amount of time, for example solo 5312294 ran 311 seconds and received 811 credits, micro 5638068 ran 510 seconds and received 743 credits...as I'm only in it for the credit this presents a problem...
17) Message boards : News : Long Run-Times? Optimize. (Message 18706)
Posted 1430 days ago by Joe
That worked thanks.
18) Message boards : News : Long Run-Times? Optimize. (Message 18699)
Posted 1431 days ago by Joe
Have played with your optimized instructions..but you forgot one important variable. XP on an nvida card.I have added countless useless lines of code in god knows where

For many projects XP on a a nvida card is the best PPD I've stuck your code in every file I could locate and nothing but 4800 second WU's that were 1200 seconds before.

I understand I'm not a cool lkinix programer..just a guy looking for points.

Stopped by again because you seem to me to really be trying to make everyone happy. I don't pretend to speak for anyone but myself.

If this project doesnt realize that there are folks out there that have been doing this for 10 plus years over multiple projects, which all are still work today problem free in spite of multiple upgrades/changes over that time, then I can only conclude this is a project that has evolved beyond the scope average user.

I have no problem with the fact you have to make choices and you have seemingly chosen to exclude the average XP user using nivida cards giving maxium PPD on other projects..there are Linix only projects which I obviously don't participate in that have also made this choice.

Again speaking for myself, just label this as a project for advanced users using Win 7 and above or whatever linix version works.
19) Message boards : News : More Updates (Message 18450)
Posted 1442 days ago by Joe
I have neither the knowledge nor the time to edit anything...I tried to copy the files but nothing happened so I aborted the last 2 hour wu. Sorry I'm an idiot but idiot worked before the crash and I didn't acquire the ability to tweak configuration files because of the crash.

Good luck with your project. You are obviously trying very hard to make it work...
20) Message boards : News : More Updates (Message 18429)
Posted 1442 days ago by Joe
Finally got wu's that didn't immediately error out but it takes forever to complete (6000 sec)when compared with previous wu's. So a box that used to crunch 250k/day will be lucky to do 90k..thats a big improvement over zilch/day but not enough to keep a credit whore like me interested long term...hope things gewt sorted in a week or so


Next 20

Main page · Your account · Message boards


Copyright © 2018 Jon Sonntag; All rights reserved.