 |
|
Questions and Answers :
Wish list :
AMD64
Message board moderation
| Author | Message |
|---|---|
 Send message Joined: 28 Aug 04 Posts: 29 Credit: 1,061,281 RAC: 8,788 |
|
|
Send message Joined: 5 Sep 04 Posts: 7629 Credit: 24,240,330 RAC: 0 |
A 64 bit version was tried last year, but proved to be too unstable. The programmers have said that it may happen at some time. But there will be NO anonymous Linux because of copyright issues. |
|
Send message Joined: 1 Nov 05 Posts: 2 Credit: 8,346 RAC: 0 |
This means no anonymous platform (app_info.xml) support, right? |
 MikeMarsUK Send message Joined: 13 Jan 06 Posts: 1498 Credit: 15,613,038 RAC: 0 |
Anonymous platform should be OK (I guess, never used it), but what Les means is that the code isn\'t available, so the software can\'t be built by third parties. I beleive that you can run the 32-bit project code on a 64-bit platform if you wish (possibly using the technique that you mention). I'm a volunteer and my views are my own. News and Announcements and FAQ |
|
Send message Joined: 5 Sep 04 Posts: 7629 Credit: 24,240,330 RAC: 0 |
Some one somewhere, (perhaps on the php boards), was asking time and again about an anonymous platform, and I asked Carl, one of the programmers, if he could supply an official answer. That answer was NO ANONYMOUS PLATFORM! The code, both source and compiled, is owned by the UK Met Office, and Carl said that to provide an anonymous platform would require making it public, as the xlm part would require the source code. If you don\'t like running 32 bit code, I\'m afraid that you\'ll have to stick to those projects that DO allow for a 64 bit build. |
|
Send message Joined: 28 Aug 04 Posts: 29 Credit: 1,061,281 RAC: 8,788 |
Then how about sending the x86 application to hosts reporting the AMD64 platform, both Linux and Windows? This way, hosts running the AMD64 client would be able to run the CPDN application as well. Thanks, 
|
|
Send message Joined: 1 Nov 05 Posts: 2 Credit: 8,346 RAC: 0 |
Some one somewhere, (perhaps on the php boards), was asking time and again about an anonymous platform, and I asked Carl, one of the programmers, if he could supply an official answer. Ok... [...] to provide an anonymous platform would require making it public, as the xml part would require the source code. This is not true, I run compiled-only versions of Einstein@Home, SIMAP, and chess960@Home. |
|
Send message Joined: 13 Sep 04 Posts: 2 Credit: 932,715 RAC: 0 |
Well, it\'s sad because the AMD64 is a very efficient processor. I\'ve stop running climate as there is no AMD64 client. If unstable a year ago, thing should have changed. The new Intel processor are 64bits too, so it should be time to change. L\'Alliance Francophone is the first team in climate and more and more people are leaving climate for Simap or Seti... Hope it\'ll change soon :( |
|
MikeMarsUK Send message Joined: 13 Jan 06 Posts: 1498 Credit: 15,613,038 RAC: 0 |
There are lots of users with AMD processors running the climate project, I run an AMD X2 4600 myself. The thread is about whether the climate app is 32bit or 64bit, which is a separate issue. I'm a volunteer and my views are my own. News and Announcements and FAQ |
|
Send message Joined: 13 Sep 04 Posts: 2 Credit: 932,715 RAC: 0 |
Sure, but my plateform is x86_64-pc-linux-gnu. The 64 bits gain is quite important when used correctly, why working on a non optimised project ? |
|
Send message Joined: 28 Aug 04 Posts: 29 Credit: 1,061,281 RAC: 8,788 |
So far, two projects have added support for x86-64: SETI sends the x86 application and SIMAP, a true x86-64 application. Would you consider supporting the platform x86_64-pc-linux-gnu or x86_64-unknown-linux-gnu similarly? HashClash now supports AMD64 with a 32-bit Linux application and Chess960 with a 64-bit Linux application...
|
|
Send message Joined: 28 Aug 04 Posts: 29 Credit: 1,061,281 RAC: 8,788 |
So far, two projects have added support for x86-64: SETI sends the x86 application and SIMAP, a true x86-64 application. Would you consider supporting the platform x86_64-pc-linux-gnu or x86_64-unknown-linux-gnu similarly? HashClash now supports AMD64 with a 32-bit Windows application too...
|
|
Send message Joined: 28 Aug 04 Posts: 29 Credit: 1,061,281 RAC: 8,788 |
FWIW, running two instances of the client, one the 32-bit client, the other, the 64-bit client, on the same 4-core system, but limiting each client to 2 cores, I can compare the relative performance of 32-bit and 64-bit SIMAP\'s HMMER: the 64-bit version is about 7% faster. By enabling vectorization (supported by default on AMD64), the SIMAP developers observed other 8% improvement. That\'s a potential performance improvement of 15% when porting the project application to AMD64.
|
|
Send message Joined: 28 Aug 04 Posts: 29 Credit: 1,061,281 RAC: 8,788 |
So far, two projects have added support for x86-64: SETI sends the x86 application and SIMAP, a true x86-64 application. Would you consider supporting the platform x86_64-pc-linux-gnu or x86_64-unknown-linux-gnu similarly? Guess what? Leiden Classical now supports AMD64 on Linux with a 32-bit application. ;-)
|
|
Send message Joined: 5 Sep 04 Posts: 7629 Credit: 24,240,330 RAC: 0 |
Well, it\'s sad because the AMD64 is a very efficient processor. I\'ve stop running climate as there is no AMD64 client. The current program is still the same as was tested for a 64 bit version a year ago, and I think that the compiler is also still the same. The program is now considered to be stable and, along with the new Intel Core Duo processor being the fastest thing there is for these models, even in 32 bit mode, one of the two programmers leaving the project in about two months, programming effort being concentrated on the \"back end programs\", (which will allow researchers safe access to the data being accumlated), and probably also \"etc\", I have a feeling that the priority level for a 64 bit model, (as well as a Mac version), is at about the same level as that at which BOINC runs. You may need to ask your self which is more important: Only run 64 bit science apps, or run whatever climate model is available. |
|
Send message Joined: 28 Aug 04 Posts: 29 Credit: 1,061,281 RAC: 8,788 |
You may need to ask your self which is more important: One doesn\'t have to exclude the other. Many projects send the 32-bit application to 64-bit clients, when they can continue to contribute to those projects that don\'t support 64 bits natively. CPDN could follow the same route creating the 64-bit application as an alias to the 32-bit application. Of course, BOINC intends to make this happen automatically in future versions of the back-end (see this), but until then, the accommodation of 64-bit clients must be done manually. Anyways, those who\'d like to try out running 64-bit BOINC, can download the x86-64 Linux client, version 5.8.15, from boinc_5.8.15_x86_64-pc-linux-gnu.tgz (make sure to read the file \"README.x86_64-pc-linux-gnu\" in it), and Crunch3r\'s x64 client, version 5.8.11, from boinc_5.8.11_windows_amd64.zip. For more information, see BoincStats Forum. HTH
|
|
Send message Joined: 28 Aug 04 Posts: 29 Credit: 1,061,281 RAC: 8,788 |
Here\'s the new recommended version for the x86-64 Linux client:
|
|
Send message Joined: 28 Aug 04 Posts: 29 Credit: 1,061,281 RAC: 8,788 |
Here\'s a development version of the x86-64 Linux client:
|
|
Send message Joined: 5 Sep 04 Posts: 7629 Credit: 24,240,330 RAC: 0 |
This is known. However, the server software here still hasn\'t been updated to the version which sends 32 bit apps to 64 bit clients. Sometime soon, it\'s hoped. |
|
Send message Joined: 27 Jan 07 Posts: 300 Credit: 3,288,263 RAC: 26,370 |
Rosetta@home did something sneaky to get around this. They just did a file copy and renamed the 32-bit app to the 64-bit version. BOINC clients think they are getting a 64-bit version for x86_64 but it\'s just the 32-bit binary renamed. Seems to work OK. |
©2024 climateprediction.net
