The only point I would dispute is that ATI support in the Client is promised, but now it is backed up one version in the schedule. I am not sure of all of the reasons but one is likely that DA asked for an ATI card, they promised it, and UCB still has not gotten it.

Now, were it my project I would have gone out and bought one, albeit a cheap one, to integrate the new GPU into the code base.

Don't forget, there is also the need to add support of OpenCL. The only bright spot there is that Charlie Fenton (the OS-X developer) may add that for us when Apple releases Snow Leopard with OpenCL support built in. We are also waiting video drivers with CUDA support for OS-X. Unknown to me if Apple is doing the work there, or Nvida is ... but either way, we are still waiting.

What burden that places on projects to add support for OS-X is not at all clear to me. In other words, it would not be a "slam-dunk" to expect that a project such as GPU Grid or SaH to have an OS-X version of the applications right after a driver release that made it all possible.

I am not sure of all of the reasons but one is likely that DA asked for an ATI card, they promised it, and UCB still has not gotten it.

Um, no.. they got it but it didn't work. The whole ATI support is being tested in-house;

Don't forget, there is also the need to add support of OpenCL.

If projects want to have CAL or OpenCL support is totally up to them. It's the science application that uses that, BOINC only uses C for the detection.

---

Jord.

The BOINC FAQ Service.

I find it interesting that ATI support is not 1st on the list as to the fact of price ATI vs Nvidia.

But hey maybe Nvidia is giving better support than ATI. I for one have had better luck in the contact of Nvidia corp than I am getting from ATI.

Maybe ATI is snooting the whole Boinc project thing.

Just a thought.

---

Basically i think the whole boinc and marketing seems to be all about nvidia.
At the moment especially on the low budget cards ATI absolute rulez on the sales.
Almost everyone i talk which bought a new one seems to have bought an ATI card based on the 4XXX series, since they give much more bang for the buck.
Simply because nvidia give much less performance for the same price, in fact only a few i know bought an 2XX card, when i asked them why they stated to run cuda alos which made them choose the more expenssive nvidia

There is another way to reduce the hammering of the server -- eliminate support for ATI GPU <smile>. That is where the server stress started. I'm not advocating that (even though I have no ATI 38xx or 48xx cards deployed -- not a gamer or photo processor or CAD/CAM guy), but as long as folks are seeking to solve a problem, add that as grist for a potentially flaming discussion.



Of course, the easy way to stop the hammering of the server would be to reduce the credits to that of *cough*seti*cough*, or even below that. Then watch how quickly people will stop crunching and hammering here - after some very unpleasant postings in the forum, no doubt.

Excuse me while I put on my fire-resistant suit!

There was some server stress when gpu was opened up, but more so when they increased the min number of wu's on the server that would trigger new wu generation. then the server started sending out "Low" responses all the time and everyone's boinc clients started scheduling way off in the future.


what may have been a better solution would be to reset the trigger amount back where it used to be but generate more work at that level. Before travis monkeyed around with those trigger levels, there was something on the order of 80k wu's running at any time. now there are only 30-ish k.

bob>where can i get one of these scripts?
Mine [XP] says:

@echo off
REM Boinc Automatic Project Update
:TOP
"c:\Program Files\BOINC\boinccmd.exe" --project http://milkyway.cs.rpi.edu/milkyway/ update
ECHO %TIME%
SLEEP 5m
GOTO TOP

You'll have to get SLEEP from the net, or fake it using PING as mentioned elsewhere on this board

Oh, you mean the thread that was locked for administrative purposes??? Yeah, I finally figured it out using windows task manager. and a few people have PM'd me since. Now my CPU machines have a small but constant flow of work, and my GPU gets units in small batches but allows me to continue running 4 ABC and 4 CPDN wus concurrently with 3 Milkyways, when I get them.

my average has jumped up from about 1000 to 25000 a day. still a far cry from the 80k/day I was getting just 1 1/2 months ago.

thanks, v

Sounds like a rather lame excuse -- we are talking 'sub $100' here for a 3x00 or low end 48xx card. Perhaps folks over at SETI need to redirect their contributions to specifically encourage DA to do this....

In any event, it does sort of fit an impression I'd had regarding DA. Then again, if there is no support for ATI cards, DA can't possibly complain about excessive credits going to the existing ATI optimized application as he has nothing to benchmark it to.

The only point I would dispute is that ATI support in the Client is promised, but now it is backed up one version in the schedule. I am not sure of all of the reasons but one is likely that DA asked for an ATI card, they promised it, and UCB still has not gotten it.

Now, were it my project I would have gone out and bought one, albeit a cheap one, to integrate the new GPU into the code base.

---

Then again, if there is no support for ATI cards, DA can't possibly complain about excessive credits going to the existing ATI optimized application as he has nothing to benchmark it to.

"Can't" and "shouldn't" are two different words...

Then again, if there is no support for ATI cards, DA can't possibly complain about excessive credits going to the existing ATI optimized application as he has nothing to benchmark it to.

If the ATI is not supported, then it must remain on the anonymous platform, continuing to hit hard the CPUers, unless the anonymous platform is banned!

---

unless the anonymous platform is banned!

Which would also disable the third-party CPU applications, which I don't think is practical.

---

unless the anonymous platform is banned!

Which would also disable the third-party CPU applications, which I don't think is practical.

I agree, it's unlikely that ATI will not be supported, so what can DA do about it, that's the question?

---

Why would he bother with it at all?
It only means that much more science is getting done that much faster.

---

Why would he bother with it at all?
It only means that much more science is getting done that much faster.

Science and politics mix poorly.

-jim

Why would he bother with it at all?
It only means that much more science is getting done that much faster.

But it is getting done on other projects, not his beloved SETI (washes mouth out with soap)

IMO the reason DA is so hell bent on cross project parity,(if the project is granting more than SETI, if less than SETI. ohh well) to push more participants back to SETI.

---

.

I dont see what he's worried about, most of us run Seti too, along with a number of other projects he seems to be involved with. Oh well I was curious thats all. Thanks for the insight. ;-) Maybe his motivation is jealousy???

---

Indeed, let us say that what he might have to say should be viewed thru the lens of his demonstrated experience with an ATI GPU client development and performance.

"Can't" and "shouldn't" are two different words...

---

Might be the case, but given the load issues that SETI has, more users there and more load is not what they are ready for these days. I've been attached to SETI shared work since the pre-BOINC days (joined over 8 years ago), but these days they get essentally a stay-awake tidbit of the cycles have available. For that matter, SETI is pretty much the LEAST science research oriented project of those to which I'm attached.

When I found GPUGrid, my Cuda cycles jumped from SETI to it. And for the reset, aside from MilkyWay, I'm doing Spinhenge, POEM, Climate and Einstein fairly heavily. My 'also-ran' project list includes SETI, WorldGrid, Rosetta and Malaria. My dead projects are ClimateBBC (which shut down gracefully) and Predictor (which shut down irresponsibly).

IMO the reason DA is so hell bent on cross project parity,(if the project is granting more than SETI, if less than SETI. ohh well) to push more participants back to SETI.

---

Then again, if there is no support for ATI cards, DA can't possibly complain about excessive credits going to the existing ATI optimized application as he has nothing to benchmark it to.

If the ATI is not supported, then it must remain on the anonymous platform, continuing to hit hard the CPUers, unless the anonymous platform is banned!

Actually, as I mentioned before, if CUDA comes online, people currently with a CUDA-capable card that are participating with their CPU would most likely switch to the other project. It will alleviate things somewhat...

Do you think there would be a problem running MW Cuda & CPU @ the same time??

---

Do you think there would be a problem running MW Cuda & CPU @ the same time??

I have no idea. I don't have the money to spend on any computer upgrades. I'd imagine there wouldn't be though, but still it means that some tasks will get processed slower here, ultimately lightening the load to some extent.