LHC@home

This suggests LHC don't use GPUs, and don't mention ever doing so in the future:
http://lhcathome.web.cern.ch/test4theory/does-lhchome-support-gpus

This BOINC list shows the AMD logo on LHC as though it's supported:
https://boinc.berkeley.edu/projects.php

The LHC project settings have a tick for AMD:
https://lhcathome.cern.ch/lhcathome/prefs.php?subset=project

Are there any plans for GPU work? Currently the server gives me none.

---

See here...

https://boinc.berkeley.edu/forum_thread.php?id=13171

Wow, "native" triggering "ati". REALLY bad programming! I can think of many many words containing the letters ati.

---

Hello Cruncher and forum manager!
At first i want to speek from the computer market of GPU
The most selled GPU are NVIDIA GTX... with Always greater power for computing and with a regular upgrade of the whole models of GPU drivers
If i compare the computing speed of 2 similar tasks as sample by SETI then you have for the same task with GPU a computing time of only 11 minutes to 20 minutes depending of GPU model and if you compute only with CPU then you are needing 1h30 minutes to 2 h 30 minutes

If this is applied Inside of the CERN which is actually doing a great upgrade, then i whish for the future the use of cuda 42 or Cuda 50 or 55 for NVIDIA graphics cards or nvidia GPU
This will severaly decrease the computing time for many tasks
For comparison i think that 7 hours computing could be brang to less as one hour
This increase the productivity of each of our computers and bring to the cern much more results in less time as usual with the same number of computers in the world
The cruncher are receiving as prime for their truly crunching by cern a higher number of cobblestones in the same time.
Everybody is winning in the party

Guy PFLIEGER
France

I just want to excuse me against you to have forgotten Something:
I forgotted to say to you in the last message:

Happy Crunching and best Regards

The Reason was that i combined at Sunday several activities and i wrote the message after 22 hours of combined work so i was really tired
I get to bed early in the Morning and after sleeping i had to do my cooking
I was also yesterday searching muschroom and i found enough to do today my cooking and some were prepared to be freezed for the winter
In regards of muschroom in Relationship with possible radiation comming from Tchernobyl problem, I can say that the dried muschroom where controled with a dosimeter and you can find a very little noise comming from the muschroom. The difference is so little that we can say now that the Relationship with the data published by crirad direct after the destruction of Tchernobyl show a very important decreasing of the radioactivity in the muschroom in alsace (except in ground recognized containing natural uran in Relationship with possible mining)
North of France is usually a raining part of France and if you look the property of uran is to be slovly dissolved in water and with the years the ground was cleaned with the rain and the melting snow in ground that did not contain uran in the rock Under the ground

Now i finish my meeltime and i eat my muschroom

Happy Crunching
Best Regards

Guy PFLIEGER
France

Bon appétit Guy!

---

I eated Lepista nuda with ognon and Brussel sprout
Yesterday and today i found also Laccaria amethystina.
One part i prepared for an omelette and the rest was light cooked in melted butter and as soon as it was cold i putted them to freezing for the winter
So I have hability to have special mushrom for the feasts

They where searched outside of mining zone in a forest with ground of clay

This was excellent
but i have a problem with my hearth and the specialist said to be care with the consum of mushrom

Thank you for your messages

Now i think listening "Hold the line" from TOTO

Nice evening for everybody!

Guy PFLIEGER
France

At first i want to speek from the computer market of GPU
The most selled GPU are NVIDIA GTX...

No, the most selled gpu are Intel integrated gpu.
Second: you can use Cuda only on Nvidia gpu, OpenCl can crunch on Nvidia, Amd, Intel, FPGA, ARM, Mali, etc

then the answer is easy:
To satisfy the most of users the choice should be Open CL
and then if CERN develop under opencl and CPU tasks then it is possible that the most of user can use his GPU

Guy PFLIEGER
France

document to read with link for the devollopers (Sorry; in French)

https://fr.wikipedia.org/wiki/OpenCL

https://fr.wikipedia.org/wiki/Compute_Unified_Device_Architecture


Guy PFLIEGER

France

download of the languages:
https://www.khronos.org/opencl/
https://developer.nvidia.com/cuda-zone


happy crunching!

Guy PFLIEGER
France

At first i want to speek from the computer market of GPU
The most selled GPU are NVIDIA GTX...

No, the most selled gpu are Intel integrated gpu.
Second: you can use Cuda only on Nvidia gpu, OpenCl can crunch on Nvidia, Amd, Intel, FPGA, ARM, Mali, etc

Then how come there are projects which only use ATI but not Nvidea?

---

Then how come there are projects which only use ATI but not Nvidea?

Which project? All opencl app support both Amd and Nvidia gpu.
The only "problem" may be the different optimization of the code.

Then how come there are projects which only use ATI but not Nvidea?

Which project? All opencl app support both Amd and Nvidia gpu.
The only "problem" may be the different optimization of the code.

Oh, I stand corrected, there aren't any. I'm sure there used to be. Did OpenCL not used to run on Nvideas?

So is there any point in games or BOINC having Cuda code if all cards can just use OpenCL?

---

Hello,

I confirm we plan to enable GPU computing in SixTrack with Boinc support. We are developing a new library SixTrackLib which would enable GPU offloading for a large subset of SixTrack simulations. As some of you argued, we are focusing on using OpenCL 1.2 because it allows to run on AMD, Intel and Nvidia GPUs, although is less advanced than other options (OpenCL 1.2 dates back to 2011).

Generally GPUs have very strong 32bit floating point capabilities, but we will not be able to profit from them. The calculations needed by SixTrack requires variables with large dynamic range, which is very different from tasks like machine learning. SixTrack computes the trajectories of the particles as they move around the accelerator. There are 6 main quantities (3 positions and 3 velocities for each plane: horizontal, vertical, longitudinal) that oscillates around an equilibrium trajectory. Each position and velocity is mainly affected by the velocities and positions of the corresponding planes and very weakly from the other planes. This weak interactions are small in the short term, but become very important in the long term (that are the typical scale of our simulations). A weak interaction correspond to the operation of adding a small quantity on a large number, and if we the number does not have enough bits, we loose the information and the simulation would be wrong.

Despite this, GPUs also support 64bit floating point calculations but at a lower speed, typically 16/32 times more slowly, however still the computing capabilities they provide are comparable and often exceed the ones of CPUs. Even the latest embedded GPUs in notebooks are fairly capable. There are also GPUs that are strong in 64bit arithmetic such as Nvidia Quadro GV100, Tesla V100/P100, Titan V, AMD Radeon VII, Firepro W9100, W8100 although they are very expensive and less common in workstations (instead more common in supercomputers and the cloud providers).

For this reason, we would like to enable GPU offloading in SixTrack because we believe there are substantial additional computing resources that we would be able to access for our simulations.

We are still in the process of developing SixTrackLib, validating the physics and tackle the integration into the main SixTrack. Our time scale is still probably a year or two to expose the functionality into LHC@Home, but we already performing simulations using SixTrackLib in GPU for specific problems.

Please do not hesitate to ask if you have further questions and my apologies for the slow answer....
Riccardo from the SixTrack team

Great.
Thanks for explaining.

That is very useful. A longer time frame will help guide our purchases. Some of the new 7nm GPUs may have not bad 64-bit capability, maybe at reasonable cost.
At least we can look for it. Thanks.

I have two nVidia cards on 2 different PCs and I load the latest drivers from nVidia. I use them on SETI@home, Einstein@home and GPUGRID, The funny thing is that sometimes the latest drivers fail, especially in SETI, and you have to load older drivers.The drivers are meant for gamers, not science tasks, I am not a gamer.
Tullio

---

Thanks, for the comments.

I have my FP64 ready, both Titan V & Radeon VII.

You can get older Quadro/Radeon Pro cards for an OK price, these drivers are more tested and slower release cycles for these reasons.