Message boards :
Number crunching :
Confused..about what GPU support is coming 1st
Message board moderation
Previous · 1 · 2 · 3 · 4 · 5 · 6 · Next
Author | Message |
---|---|
Send message Joined: 6 Apr 08 Posts: 2018 Credit: 100,142,856 RAC: 0 |
|
Send message Joined: 12 Apr 08 Posts: 621 Credit: 161,934,067 RAC: 0 |
So to sum it up..they don't care; as long as the WU's are being crunched..Even though their server can't handle the load, lol. I did not say they did not care ... I said they should not care ... A significant difference... The point of BOINC is that you can do science on the cheap. That in turn means that some things may not work so well... in this case, underpowered server for the load means a number of problems that to us means sparse work... The project is working on that, but, because we are doing science on the cheap ... we trade cost for time ... should you have an extra couple million in the back yard I am sure if you send it to Travis and Co. they will upgrade the server, hire more help ... right after they get back from the cruise ... :) |
Send message Joined: 6 Apr 08 Posts: 2018 Credit: 100,142,856 RAC: 0 |
but, because we are doing science on the cheap ... Cheap? The PCs we use are not cheap, and neither is the electricity to run them. The GPU cards are not cheap either, and neither is the electricity to run them. Whatever you say about the servers - MilkyWay is getting our computers, GPU cards (from some of us) and electricity - which are not 'cheap'. |
Send message Joined: 12 Apr 08 Posts: 621 Credit: 161,934,067 RAC: 0 |
but, because we are doing science on the cheap ... Ahem ... your individual cost is not an issue. And, as computers go, they are cheap... To build effectively a supercomputer the MW project invested maybe $20K in hardware, probably less ... and yet, they have what amounts to supercomputer computing capacity ... a steal ... You are looking at it from YOUR perspective, not the project's ... and not from the perspective of what supercomputer time and hardware costs ... So, yes, cheap ... For my part, any computer I can build for less than $3K I consider very cheap ... my workstations tend to go over $10K, though with a 5-10 year lifespan I tend to get my money's worth out of them ... |
Send message Joined: 6 Apr 08 Posts: 2018 Credit: 100,142,856 RAC: 0 |
You are looking at it from YOUR perspective, not the project's ... Not looking anywhere really, just concerned that you are belittling my 'cheap' efforts. It's not 'cheap' for the project either. Everyone can clearly see that the total cost of the project is HUGE with the all the crunching contribution included, and that contribution is not 'cheap'. By all means describe your contribution and efforts as 'cheap', but less the the 'we' since I certainly object to be included in your 'we'. but, because we are doing science on the cheap ... |
Send message Joined: 24 Dec 07 Posts: 1947 Credit: 240,884,648 RAC: 0 |
youre welcome to come back in a few months and see if its more to your liking then. A light is at the end of the tunnel....and it's a farm of 4870 graphics cards glowing red hot with all the work they're now getting! Now to get my 4850 and 4870 working under XP...been putting off doing the dummy vga plug, but now is the time. |
Send message Joined: 21 Feb 09 Posts: 180 Credit: 27,806,824 RAC: 0 |
youre welcome to come back in a few months and see if its more to your liking then. I set my second card to extend the desktop, and make it the main monitor. Then drag all your icons over, then plug the monitor in the second card. No need for a dummy plug :) |
Send message Joined: 29 Aug 07 Posts: 327 Credit: 116,463,193 RAC: 0 |
youre welcome to come back in a few months and see if its more to your liking then. Someone over at SETI said that just plugging in an S-Video cable to the extra card would activate it, no need for a dummy plug. Calm Chaos Forum...Join Calm Chaos Now |
Send message Joined: 24 Dec 07 Posts: 1947 Credit: 240,884,648 RAC: 0 |
Sweet....I'll try both. Shame I'm at w#$^ and have to wait until 9pm tonight. |
Send message Joined: 21 Aug 08 Posts: 625 Credit: 558,425 RAC: 0 |
Since Misfit hasn't asked already, if you ever feel the need to be generous, I've got dibs... ;-) |
Send message Joined: 27 Aug 07 Posts: 915 Credit: 1,503,319 RAC: 0 |
For my part, any computer I can build for less than $3K I consider very cheap ... my workstations tend to go over $10K, though with a 5-10 year lifespan I tend to get my money's worth out of them ... Oh I had thought about it. But the only time I get a response out of Paul is when I bring up the WIKI; which for some strange reason doesn't rhyme with TIKI or even SQUEAKY. me@rescam.org |
Send message Joined: 12 Apr 08 Posts: 621 Credit: 161,934,067 RAC: 0 |
You are looking at it from YOUR perspective, not the project's ... I did not belittle anything. I just point out that you are looking at your costs which you seem to think are high. Which begs the question, if it is too high for you, why did you invest the money in the system and connect it to MW and use all that electricity? As a volunteer you have total control over your costs and how much you contribute. By all means describe your contribution and efforts as 'cheap', but less the the 'we' since I certainly object to be included in your 'we'. If you are here, and you are volunteering and bearing the cost of running your systems then you are in the "we" like it or not ... The point is that the project has explicitly passed the cost of doing the work onto the volunteers. That we eagerly accept the burden in the name of doing the science means that the project does not have to bear those costs. Which means for them the project can be done on the cheap with far less investment. So, in the BOINC community we are asking questions and doing calculations that have not been done before, or at this scale before because no one wanted to sink the money into the question ... now we can do things cheaply ... from the project perspective ... well ... now we are looking at other questions ... |
Send message Joined: 6 Apr 08 Posts: 2018 Credit: 100,142,856 RAC: 0 |
If you are here, and you are volunteering and bearing the cost of running your systems then you are in the "we" like it or not ... I think your are missing my point, as you seem to miss everything in your ramblings. Please don't speak for me. Please don't spout your opinions with your 'we' as if those opinions include me. You also seem to want to speak for the project admin here, but I don't see anything that indicates you represent them, and you certainly don't represent me. "I" am not doing science on the cheap even if you think eveyone else is - get it? "I" do not think that "we can do things cheaply", whether as "we" you are talking about all crunchers or as a MilkyWay project staff (get them to give you a tag if you are). I am certainly not part of any of your "we" since I repeat my view that the project cost with all the contributions of volunteer time, computers and electricity bills is HUGE. |
Send message Joined: 12 Apr 08 Posts: 621 Credit: 161,934,067 RAC: 0 |
If you are here, and you are volunteering and bearing the cost of running your systems then you are in the "we" like it or not ... I could say the same thing, that you too are missing my points ... I represent no one but myself and have never claimed to speak for anyone but myself. The aggregated costs are significant... though I would not say huge ... but I have been speaking of the direct costs to the project and the individual costs to the participants. In both cases the costs are low and controllable as compared to the cost if the project were to want to perform the calculations on their own captive hardware. But it is obvious we are never going to agree on this point. |
Send message Joined: 24 Dec 07 Posts: 1947 Credit: 240,884,648 RAC: 0 |
If you are here, and you are volunteering and bearing the cost of running your systems then you are in the "we" like it or not ... People have gone out and purchased suitable GPUs specifically to crunch MW. I know I have and I know of others who have (and that includes Ice / Antonphotos). I also know for a fact that the cost of my electricity for 1 year of crunching MW is less that the cost of the GPUs I have purchased per GPU (that may change with the 3850 AGP I just purchased for my old P4). Over the lifetime, of what I imagine the project lifetime to be, the cost of electricity will far outway the cost of the GPUs I have purchased - but that is just due to the cost difference between a good GPU and a brand new computer (that does an order of magnitude less work that a GPU). With the advent of decent GPUs and GPU crunching, the cost of the base computer now for outways the cost of the GPU and we (yes everyone who has purchased a GPU for crunching) are happy that we get more bang for our buck with a GPU. |
Send message Joined: 6 Apr 08 Posts: 2018 Credit: 100,142,856 RAC: 0 |
With the advent of decent GPUs and GPU crunching, the cost of the base computer now for outways the cost of the GPU and we (yes everyone who has purchased a GPU for crunching) are happy that we get more bang for our buck with a GPU. I would go a little farther than that with my bangs for a buck and say in my case I have incurred the cost of ATI GPU cards - but that's it, just the electricity costs. I have managed to use older PCs with PCI-E slots that were destined for landfill. They aren't much good for anything else and could expire on me at any time - but then I'll just go and visit the landfill again. That is one of the key successes of the GPU cards - any old crummy PC will do so long as you can install the GPU and run BOINC. |
Send message Joined: 22 Nov 08 Posts: 136 Credit: 319,414,799 RAC: 0 |
If you are here, and you are volunteering and bearing the cost of running your systems then you are in the "we" like it or not ... You're nit-picking. It was meant in the broad scale of things. "You" were not singled out. 4870 GPU 4870 GPU |
Send message Joined: 6 Apr 08 Posts: 2018 Credit: 100,142,856 RAC: 0 |
If you are here, and you are volunteering and bearing the cost of running your systems then you are in the "we" like it or not ... No I am not nit-picking. The statement was "we are doing science on the cheap" but I object to someone saying that I am doing science on the cheap. By all means agree that you are doing it on the cheap, but please don't speak for me and say that I am doing it on the cheap. What I've contributed to MilkyWay is NOT cheap. |
Send message Joined: 21 Aug 08 Posts: 625 Credit: 558,425 RAC: 0 |
If you are here, and you are volunteering and bearing the cost of running your systems then you are in the "we" like it or not ... Nobody said that what you have contributed is...however when putting things in perspective, all of our systems allow them to get the work done for less money than what it would cost them normally for equivalent processing time on a cluster / supercomputer. Ergo, it is "on the cheap"...from a certain perspective. I'm sure Paul wasn't trying to say that you're cheap or I'm cheap or whatever, just that if there wasn't an economical savings to doing things this way, it would make a whole lot more sense to just do it on a cluster, which is what Predictor ended up doing... |
Send message Joined: 6 Apr 08 Posts: 2018 Credit: 100,142,856 RAC: 0 |
If you are here, and you are volunteering and bearing the cost of running your systems then you are in the "we" like it or not ... Say what you want, but not that "we are doing science cheap". I just completed 18,000,000 MilkyWay WUs. They weren't "cheap" as far as I am concerned, for me or for the project. |
©2024 Astroinformatics Group