Many GPU BOINC projects now use OpenCL instead of CUDA.

I thought it tended to be aswell as, since Cuda is more efficient.

OpenCL will run on Nvidia GPUs, AMD GPUs, the GPUs built into some Intel CPUs, and maybe a few more types I haven't read about.

Doesn't it also run on CPUs? Look at the start of your Boinc event log, it says OpenCL against the CPU.

Python is for the CPU portions of the tasks only, and does not affect which GPUs will run the GPU portions of the tasks.

You may need to install a rather new version of VirtualBox to get the vbox64-mt that some of the projects using Python require.

I need version 5, not 6, or it screws up some projects, like Cosmology. Anyway, I have no Nvidias.

It was all above the board. Nothing negative. Just inquiring.
Couple of other "senior crunchers" added their input.

Doesn't matter, if you imply they're not working hard enough you're a naughty boy.

**Then they have self esteem issues and need to go see the mental health department**

It really is looking dim for BOINC type projects in health.
I checked on TI and they still have the same issues as about a year or 6 months ago.

What is TI?

***Hmmmm...it's midnight so I don't know what I was typing a hour ago. Not important I guess**

WCG has bugs up to wazoo so they are offline.
I think I am going to take that off the list of projects until they get things fixed, no reason to keep it.

I've got 90 CPU cores running it 24/7 and have done for the last week.

**I don't know, unless they black listed my system....I got a few tasks from them a week or more ago and since then I don't get squat. I'll just wait until they figure things out, it not like I don't have other options***

QuChem I can't to work on my system for some reason.

Is that a VB project? I find all VB projects work better if you use VB 5 instead of 6. Cosmology requires limiting to 6 cores per task in the app_config.xml or they stick. Rosetta needs an AVX capable CPU.

**I tried VB5 with QuChem. Still wouldn't run and then I get messages from other projects that you need 6. So screw it. **

Sidock I already run, thats one that keeps chugging along really good.

Agreed. So you're not one of those oooh they're Russian let's boycott them people :-)
***Country of origin for science has nothing to do with politics. I don't care these guys are Russian. If they are doing something good, great. If their leader isn't well that's politics. It just another power grabbing old man running the country into the ground.***

GPU Grid looks like they got their bugs figured out and are sending out tasks again, but they are massive. 2.67GB and written in Python.

I've got that left connected and it has never given me a ingle task. Is it still Nvidia GPU only? I have AMD GPU, Intel GPU, CPU, Android connected. All I can see on their server status is "Python apps for GPU hosts". Since it's a weird page and looks nothing like normal Boinc project pages, I can't find their applications page to see what they should be running on.



Minimum requirements

GPU: A recent NVIDIA GPU
A recent Nvidia driver
CPU: At least one core
RAM: 4GB (2GB for Vista, W7, Ubuntu)
HDD: 10GB free space, no more than 80% used (very important under windows)

Supported OS

Linux 64-bit
Windows 64-bit

https://www.gpugrid.net/join.php

I do FAH and yeah these guys have their act together. Always work to be had. And interesting stuff as well.
Never read all the info on the task, but what bit I do read is interesting.

Agreed, although the scheduler is worse than Boinc and they get very angry when you ask them to fix something, once you get it working it's ok. It just has a habit of downloading things you didn't ask for and there's no ability to abort.

The rest of my projects are science related and Prime Grid to work my 1080 to its maximum potential.

I sometimes do that for fun, but I think Physics and Biology are more important than finding yet another trillion digit long prime number. That's not going to cure sick people and not going to find another planet to live on and not going to discover a new law of physics.

**But GPU's are not used that much in medical with exception of GPU grid and FAH. So in order to run something different I run math. If GPU was easier to use for project data, more projects would be using them [with the exception of this one that wants to keep only to CPU].

Einstein when they do gravitational wave searches is nice, because LIGO is located close to where I grew up.

Not sure what happened to that. They were in a hurry to finish gravity and had a supercomputer running it in it's spare time. There seems to be no sign of new work for it. There's two new gamma sources to analyze when the current ones finish, and a new way of analyzing radio waves which is in the process of being ported to GPU. Ask Bernd for more information here: https://einsteinathome.org/content/em-searches-brp-raidiopulsar-and-fgrp-gamma-ray-pulsar (Look tomorrow when I have time)

For now I see no need invest any more money in updating my hardware. I have my system tweaked out to grind fast enough to keep up with the work.

I keep on adding more chips. I'm now consuming more than a UK outlet.

***Over 3.2kW, that wire was getting very hot. 3.2kw and your wires are getting hot? Your going to burn up your house if you keep that up. If your that crazy, then you should get some proper heavy gauge wire installed or a separate 3phase line with its own distribution box. If you burning up your connection to the wall, that's really dangerous. And with the power rates here in BE, I can't afford to be burning that much power. It's already to the point where we pay them money at the end of the year after the average monthly payments. And that is running this thing16 hours a day plus normal household use.**

I used to be jealous of Xenon units

You mean Xeon? I have two dual xeons. Old things, I got the CPUs for £7 each, times 4. 48 cores altogether. But the hassle of getting the server boards to work which won't boot without the correct riser boards and USB cables attached, and need weirdly wired power supplies.

and I was looking at a AMD threadripper, but didn't have enough money to get that.

Spend your money on GPUs, they do a lot more work for the price. I do that unless I run out of things to connect them to. A lot of them become quite fussy in their old age and won't co-exist with certain other GPUs. I have to shuffle things round a lot.

But being this is a volunteer system, I do not see the need to waste more money on building a higher end system. And the way things are going on a lot of projects, it just reinforces that thought process.

I'm wondering if the total Boinc throughput is dropping due to rising electricity costs. This seems to indicate I'm wrong: https://www.boincstats.com/stats/61/project/detail/credit - look at the graph that goes back several years. It doesn't look like there's a load of people turning their machines off.

BOINC doesn't offer anything new on its page and several projects are outdated or dead.
Then there is stuff that isn't even on their page that is out there.
So if they don't update their info, then the stats won't show where people are going.
16hrs is enough for me. I ran 2 older systems into the ground with 24/7 running.
But again, electricity costs are to much.

**I don't know, unless they black listed my system....I got a few tasks from them a week or more ago and since then I don't get squat. I'll just wait until they figure things out, it not like I don't have other options***

I engaged a 5 minute tickle, that might have something to do with it. They let you talk to the server every 2 minutes, so I got Boinc to check for new work every 5 minutes.

**I tried VB5 with QuChem. Still wouldn't run and then I get messages from other projects that you need 6. So screw it. **

I've never had a project require 6, which ones?

***Country of origin for science has nothing to do with politics. I don't care these guys are Russian. If they are doing something good, great.

Indeed. I started an argument about that on their forum when a few people said they were boycotting it, and for once I wasn't banned. The admins just deleted all the posts to do with politics. They seem to be getting through a lot of work, so I think the boycotters were a small minority.

If their leader isn't well that's politics. It just another power grabbing old man running the country into the ground.***

He's actually doing a very good job of running every other country into the ground. Either we buy his gas and give him military money, or we don't buy his gas and our prices go through the roof.

Minimum requirements

GPU: A recent NVIDIA GPU
A recent Nvidia driver
CPU: At least one core
RAM: 4GB (2GB for Vista, W7, Ubuntu)
HDD: 10GB free space, no more than 80% used (very important under windows)

Supported OS

Linux 64-bit
Windows 64-bit

https://www.gpugrid.net/join.php

But there was something I heard recently about them producing CPU and AMD GPU work.

**But GPU's are not used that much in medical with exception of GPU grid and FAH. So in order to run something different I run math. If GPU was easier to use for project data, more projects would be using them [with the exception of this one that wants to keep only to CPU].

Without Folding, I use Milkyway and Einstein. They seem much more useful than raw maths.

***Over 3.2kW, that wire was getting very hot. 3.2kw and your wires are getting hot? Your going to burn up your house if you keep that up.

Nah, just the extension cord all the computers (on a bookshelf) were running through. I added a second 25A 240V feed with two 13A outlets.

If your that crazy, then you should get some proper heavy gauge wire installed or a separate 3phase line with its own distribution box.

The main incoming to the house is 100A 240V. Plentiful. I'm not going over about a third of that for the whole house consumption, including computers, AC/heating, lights, domestic appliances. There are fuses and breakers where applicable anyway. There weren't when I moved in, a 100A main fuse protecting an 80A meter and a 60A fusebox, WTF? Got three fuseboxes now. Two in the house and one in the garage.

If you burning up your connection to the wall, that's really dangerous.

I trust my finger. If a cable goes over body temperature, I upgrade it. If a heatsink on a CPU/GPU is too hot to keep my finger on, I use a bigger fan or better transfer compound etc.

And with the power rates here in BE, I can't afford to be burning that much power. It's already to the point where we pay them money at the end of the year after the average monthly payments. And that is running this thing16 hours a day plus normal household use.**

I just got £820 back from my electric company as they took it without asking. My bank got a little bit annoyed when I told them.

BOINC doesn't offer anything new on its page and several projects are outdated or dead.
Then there is stuff that isn't even on their page that is out there.
So if they don't update their info, then the stats won't show where people are going.

A lot of projects don't report to Boinc they exist. Sometimes because they don't want lots of people on a small server. Wikipedia has a page with everything listed, Boinc or not: https://en.wikipedia.org/wiki/List_of_distributed_computing_projects

16hrs is enough for me. I ran 2 older systems into the ground with 24/7 running.

I like running stuff into the ground. Get every last breath out of them. Just sold a couple of GPUs on Ebay that were completely fucked, people seem to be able to fix things I cannot. Nobody seems to want the one I blew a transistor clean off the board though.

But again, electricity costs are to much.

Get some solar panels.

I have attached everything that was on boincstats. And set all projects that i don't want to crunch to no new tasks

I have attached everything that was on boincstats. And set all projects that i don't want to crunch to no new tasks

They don't have everything. Private GFN server (an offshoot of Primegrid) for example wasn't on there until I requested it: http://boincvm.proxyma.ru:30080/test4vm/ but I can't find the bloody invitation code (I'm already joined, but I can't post it here as I can't find it now and didn't note it down).

[snip]

[quote]***Country of origin for science has nothing to do with politics. I don't care these guys are Russian. If they are doing something good, great.

Indeed. I started an argument about that on their forum when a few people said they were boycotting it, and for once I wasn't banned. The admins just deleted all the posts to do with politics. They seem to be getting through a lot of work, so I think the boycotters were a small minority.

SiDock in in Slovenia, not Russia. Slovenia used to be part of the Soviet Union, before it broke up.

https://www.gpugrid.net/join.php

But there was something I heard recently about them producing CPU and AMD GPU work.

They tried producing OpenCL work that could run on AMD GPUs, Then they decided that such work would run too slowly to be worthwhile, and stopped generating any more.

Some CPUs from Intel can also do OpenCL work. But, few BOINC projects mention whether their OpenCL work will run on those CPUs.

They have generated some CPU work, using programs that don't use GPUs. It seems to be unavailable for any computers that have ever run some of their GPU work, though.

But again, electricity costs are to much.

Get some solar panels.

If you want your computers to run only during daylight, and only if the clouds aren't too heavy.

I think invitation code is PrimeGrid

I have attached everything that was on boincstats. And set all projects that i don't want to crunch to no new tasks

They don't have everything. Private GFN server (an offshoot of Primegrid) for example wasn't on there until I requested it: http://boincvm.proxyma.ru:30080/test4vm/ but I can't find the bloody invitation code (I'm already joined, but I can't post it here as I can't find it now and didn't note it down).

The projects I've seen that use invitation codes put it on their home pages, but only left it there until they thought they had about as many users as their servers could handle.

SiDock in in Slovenia, not Russia. Slovenia used to be part of the Soviet Union, before it broke up.[quote]Two of the four biologists are Russian. That annoyed some of the silly users.

[quote]They have generated some CPU work, using programs that don't use GPUs. It seems to be unavailable for any computers that have ever run some of their GPU work, though.

I have never done any of their work and have a variety of CPUs up to a Ryzen 9 3900XT. I've never got any CPU work.

If you want your computers to run only during daylight, and only if the clouds aren't too heavy.

But you can store the power in batteries or sell and buy to the grid.

I was amazed to see a blatant lie from a solar panel installer in the UK which claimed it didn't matter if the sun was out or not, it just had to be light! ROTFPMSL!

I have run out of tasks as well.

22.07.2022 22:28:12 |  | cc_config.xml not found - using defaults
22.07.2022 22:28:12 |  | Starting BOINC client version 7.16.20 for windows_x86_64
22.07.2022 22:28:12 |  | log flags: file_xfer, sched_ops, task
22.07.2022 22:28:12 |  | Libraries: libcurl/7.47.1 OpenSSL/1.0.2s zlib/1.2.8
22.07.2022 22:28:12 |  | Data directory: E:BOINC
22.07.2022 22:28:12 |  | Running under account LigH
22.07.2022 22:28:13 |  | CUDA: NVIDIA GPU 0: GeForce GTX 1660 SUPER (driver version 456.71, CUDA version 11.1, compute capability 7.5, 4096MB, 3553MB available, 5153 GFLOPS peak)
22.07.2022 22:28:13 |  | OpenCL: NVIDIA GPU 0: GeForce GTX 1660 SUPER (driver version 456.71, device version OpenCL 1.2 CUDA, 6144MB, 3553MB available, 5153 GFLOPS peak)
22.07.2022 22:28:13 |  | Version change (7.16.11 -> 7.16.20)
22.07.2022 22:28:13 |  | Windows processor group 0: 16 processors
22.07.2022 22:28:13 |  | Host name: LIGH-RZ7
22.07.2022 22:28:13 |  | Processor: 16 AuthenticAMD AMD Ryzen 7 3700X 8-Core Processor [Family 23 Model 113 Stepping 0]
22.07.2022 22:28:13 |  | Processor features: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 htt pni ssse3 fma cx16 sse4_1 sse4_2 movebe popcnt aes f16c rdrandsyscall nx lm avx avx2 svm sse4a osvw ibs skinit wdt tce topx page1gb rdtscp fsgsbase bmi1 smep bmi2
22.07.2022 22:28:13 |  | OS: Microsoft Windows 10: Professional x64 Edition, (10.00.19044.00)
22.07.2022 22:28:13 |  | Memory: 31.95 GB physical, 63.95 GB virtual
22.07.2022 22:28:13 |  | Disk: 1.96 TB total, 753.54 GB free
22.07.2022 22:28:13 |  | Local time is UTC +2 hours
22.07.2022 22:28:13 |  | No WSL found.
22.07.2022 22:28:13 |  | VirtualBox version: 6.0.14
22.07.2022 22:28:13 | Rosetta@home | General prefs: from Rosetta@home (last modified 07-Sep-2009 14:27:59)
22.07.2022 22:28:13 | Rosetta@home | Host location: none
22.07.2022 22:28:13 | Rosetta@home | General prefs: using your defaults
22.07.2022 22:28:13 |  | Reading preferences override file
22.07.2022 22:28:13 |  | Preferences:
22.07.2022 22:28:13 |  | max memory usage when active: 16357.12 MB
22.07.2022 22:28:13 |  | max memory usage when idle: 29442.81 MB
22.07.2022 22:28:13 |  | max disk usage: 32.00 GB
22.07.2022 22:28:13 |  | max CPUs used: 4
22.07.2022 22:28:13 |  | suspend work if non-BOINC CPU load exceeds 25%
22.07.2022 22:28:13 |  | (to change preferences, visit a project web site or select Preferences in the Manager)
22.07.2022 22:28:13 |  | Setting up project and slot directories
22.07.2022 22:28:13 |  | Checking active tasks
22.07.2022 22:28:13 | Rosetta@home | URL https://boinc.bakerlab.org/rosetta/; Computer ID 4848956; resource share 100
22.07.2022 22:28:13 |  | Setting up GUI RPC socket
22.07.2022 22:28:13 |  | Checking presence of 8 project files
22.07.2022 22:28:13 |  | Running CPU benchmarks
22.07.2022 22:28:13 |  | Suspending computation - CPU benchmarks in progress
22.07.2022 22:28:44 |  | Benchmark results:
22.07.2022 22:28:44 |  | Number of CPUs: 4
22.07.2022 22:28:44 |  | 5464 floating point MIPS (Whetstone) per CPU
22.07.2022 22:28:44 |  | 19012 integer MIPS (Dhrystone) per CPU
22.07.2022 22:28:59 | Rosetta@home | work fetch suspended by user
22.07.2022 22:29:04 | Rosetta@home | work fetch resumed by user
22.07.2022 22:29:09 | Rosetta@home | Sending scheduler request: To fetch work.
22.07.2022 22:29:09 | Rosetta@home | Requesting new tasks for CPU
22.07.2022 22:29:11 | Rosetta@home | Scheduler request completed: got 0 new tasks
22.07.2022 22:29:11 | Rosetta@home | No tasks sent
22.07.2022 22:29:11 | Rosetta@home | Project requested delay of 31 seconds
22.07.2022 22:30:23 |  | Running CPU benchmarks
22.07.2022 22:30:23 |  | Suspending computation - CPU benchmarks in progress
22.07.2022 22:30:54 |  | Benchmark results:
22.07.2022 22:30:54 |  | Number of CPUs: 4
22.07.2022 22:30:54 |  | 5480 floating point MIPS (Whetstone) per CPU
22.07.2022 22:30:54 |  | 18613 integer MIPS (Dhrystone) per CPU
22.07.2022 22:30:56 |  | Resuming computation
22.07.2022 22:35:50 | Rosetta@home | Sending scheduler request: To fetch work.
22.07.2022 22:35:50 | Rosetta@home | Requesting new tasks for CPU
22.07.2022 22:35:53 | Rosetta@home | Scheduler request completed: got 0 new tasks
22.07.2022 22:35:53 | Rosetta@home | No tasks sent
22.07.2022 22:35:53 | Rosetta@home | Project requested delay of 31 seconds

I was in hospital for the last 3 weeks; but the lack of new tasks happened already before.

---

Fun and success!

Jobs: holzon + 12angebote
Hobbies: doom9/Gleitz + PlaneShift

I have run out of tasks as well.

Check the details for your computer on the Rosetta webpage. It may have banned your computer from doing Python Virtualbox tasks. At the bottom, if "allow" is next to VirtualBox tasks, click it. If "skip" is displayed, then I'm wrong. It's supposed to ban computers unable to run them, but mine that can run them get banned just for the odd failure (their programming error)

Have you guys that run GPU grid...have you been crashing all your tasks lately?
I found a few of mine that showed as buggy, but others just crashed.

Is a 750 to old to run this program?

This workunit failed with gtx 780.
https://www.gpugrid.net/result.php?resultid=32965184
I think gtx 750 is definitely not supported

This workunit failed with gtx 780.
https://www.gpugrid.net/result.php?resultid=32965184
I think gtx 750 is definitely not supported

Then they shouldn't send tasks to it.

This workunit failed with gtx 780.
https://www.gpugrid.net/result.php?resultid=32965184
I think gtx 750 is definitely not supported

Then they shouldn't send tasks to it.

But they were. To both cards. 1080 and 750.

But they were. To both cards. 1080 and 750.

These projects never do things properly. I'm getting Africa Rainfall WCG tasks with less time limit than the CPU can get done in time. I've been aborting batches from them. Not sure why they have anything from 2 to 6 days time limit.

I have run out of tasks as well.

Check the details for your computer on the Rosetta webpage. It may have banned your computer from doing Python Virtualbox tasks. At the bottom, if "allow" is next to VirtualBox tasks, click it. If "skip" is displayed, then I'm wrong. It's supposed to ban computers unable to run them, but mine that can run them get banned just for the odd failure (their programming error)

I do not even have VirtualBox installed at the moment. Is it suddenly required for Rosetta?
_

OK, I installed VirtualBox, now I received some "rosetta python projects".

---

Fun and success!

Jobs: holzon + 12angebote
Hobbies: doom9/Gleitz + PlaneShift

I do not even have VirtualBox installed at the moment.

Yes, you do.
From one of you errored Python tasks,

2022-07-23 23:32:11 (18972): Detected: vboxwrapper 26202
2022-07-23 23:32:11 (18972): Detected: BOINC client v7.16.20
2022-07-23 23:32:12 (18972): Detected: VirtualBox VboxManage Interface (Version: 6.1.36)

VBoxManage.exe: error: Not in a hypervisor partition (HVP=0) (VERR_NEM_NOT_AVAILABLE).
VBoxManage.exe: error: AMD-V is disabled in the BIOS (or by the host OS) (VERR_SVM_DISABLED)
VBoxManage.exe: error: Details: code E_FAIL (0x80004005), component ConsoleWrap, interface IConsole

But you need to make sure HyperV is not enabled in Widows & that Virtualisation is enabled in the BIOS if you want things to work and not just produce errors.

Is it suddenly required for Rosetta?

Only for a year or more.
The Rosetta 4.20 Tasks are pretty much non existent, with only the very occasional small batch being released.
The only readily available Tasks are Python ones which require VirtualBox.

---

Grant
Darwin NT

But you need to make sure HyperV is not enabled in Windows

How does that get enabled? I've checked all mine and all 7 are off.

Run “bcdedit /set hypervisorlaunchtype off” to disable hyper-v
You can use https://leomoon.com/downloads/desktop-apps/leomoon-cpu-v/ to check whether Harware Virtualization is Enabled in BIOS