Vote, would u like to see folding/physx Perf articles

Would u be interested in a PhysX/Folding Performance Article


  • Total voters
    37

Lord_Exodia

Supreme [H]ardness
Joined
Sep 29, 2005
Messages
7,008
Would you like to see HardOcp do a folding and PhysX performance analysis on cards. The folding doesn't only have to be on Nvidia hardware, it could be on amd as I know alot of folders use amd as well.

I've been wondering how; gts 450, gt 430, & gtx 460 folds and does physx in a real world gaming scenario.

Is that something you would find worth while reading? Please vote and feel free to comment.
 
For the record, we do include the usage of PhysX in our gameplay evaluations and in our testing of games in regular reviews in the highest playable settings, we treat it is a game feature, just like any other graphic setting.
 
PhysX Really doesn't add much, and its a gimmicky proprietary API, so I don't care about it. Folding is all well and fine, but is applicable to such a small minority, its going to be a waste of money for [H] to do it. If you want folding performance numbers, I'm sure you can find someone who's done it. Also, you can probably make a reasonable guess on performance based on the architecture. I just don't think either of these really makes sense for [H] as a business, and on top of that, the majority of people won't care.
Edit: Brent- I think he was referring to PhysX performance as in- using one of these cards as a dedicated PhysX card with, say, a GTX470 or 480.
 
Last edited:
For the record, we do include the usage of PhysX in our gameplay evaluations and in our testing of games in regular reviews in the highest playable settings, we treat it is a game feature, just like any other graphic setting.

I know and I really appreciate that but it doesn't give me any way to measure to any other card. I guess what I would like to know is what value one card would have over another when it comes to PhysX and folding.

For example: How much better would a GTs 450 be than a GT 430 for physX and if I go all the way up to a 460 768 how much better would it be and how much can they fold for me in a weeks time.

Oh and yes as a dedicated card. Mafia 2 does alot when you dedicate a card to physX. maybe future games will take this approach. A dedicated physX card would be perfect to fold with as it'd be a wast to just buy a gpu for physX, why not fold with it when it's not crunching physx
 
Last edited:
PhysX Really doesn't add much, and its a gimmicky proprietary API, so I don't care about it. Folding is all well and fine, but is applicable to such a small minority, its going to be a waste of money for [H] to do it. If you want folding performance numbers, I'm sure you can find someone who's done it. Also, you can probably make a reasonable guess on performance based on the architecture. I just don't think either of these really makes sense for [H] as a business, and on top of that, the majority of people won't care.
Edit: Brent- I think he was referring to PhysX performance as in- using one of these cards as a dedicated PhysX cards.

I see your point. I think it'll be worthwhile and I dont see it taking up too much time. For the folding they can leave the cards folding for a week and stop them and meausre the results and post them. For the game testing, yes as a dedicated card in mafia and some other benchmarks, 2 or 3 will be fine. No need to overkill. It wont take alot to find the raw facts.

I also want to say that I dont want to pain amd in a bad light but I would just like to see what you get out of that.

I think a better vote for you would have been option 3. You have valid points.
 
I also want to say that I dont want to pain amd in a bad light but I would just like to see what you get out of that.

I think a better vote for you would have been option 3. You have valid points.

You are probably are right- I should have voted 3, but I also don't fold and think PhysX is a gimmick, so I don't CARE about them either. I guess I would vote both 2 and 3 if I could :) Whatever the case, I think you may underestimate the time and work that goes into that type of testing. If you have someone playing through a few different game sections each both with PhysX and no PhysX, with and without a dedicated PhysX card and various combinations of cards, the time adds up pretty darn fast, and thus the cost of the article increases fairly quickly, especially if you add in the writing/editing of the article and making the charts and graphs, etc. Folding may not take so much time, but its gonna load up a couple of machines for that time which then can't be used for the other reviews and testing that they would be doing- all of this for something that is only important for an incredibly small percentage of people.
 
An important question to ask is whether or not adding the time for these extra tests would pay off with more hits/readers to the article.
I'm not sure there is enough of a base of dedicated physx card users or F@H users to draw from.
Possibly some1 from the [H]orde could answer the F@H performance question without spending time reviewing it?
 
Hmm, then perhaps this information could be added in the reviews from now on. I mean folding for 2 days non stop results, and say dedicated physix support on 2 or 3 games compared to 2 other cards.

So in this GT 430 review released today they would have compared it to GTs 450 and GTS 250 in physX and folding.

I figure it'll be easier to revisit the cards after the gaming performance reviews with a quick 1-2 page performance analysis rather than include it buy hmm I guess so far the consensus thinks otherwise.
 
People with ATI cards = not interested.

I'm shocked, shocked I say! :eek:

What a troll man- You must not realize that many of us really don't have an affinity for either brand and just buy whatever suits our needs the best, and has the best performance for the price. I for one only have these current 5750s because I was given one, and wanted a little more performance, thus I bought a second as nothing else could compare to the performance of a 5850 for me only paying $130. I had a 4870 before this, but that was because I got it for $130 as well. Before that I had a string of Nvidia cards, from the Geforce Ti 4xxx series through the Geforce 7300GT that was replaced by the 4870. I just want the best I can get for my money, and PhysX and Folding don't play into that equation for me since I find them both uninteresting.
 
i checked not interested. Folding is a neat concept and I will participate eventually (once I graduate and can afford a new build). As far as Physx is concerned, there are too few games that utilize any significant amount of GPU Physx processing that the interest level for such a thing is low for me.
 
Hmm, then perhaps this information could be added in the reviews from now on. I mean folding for 2 days non stop results, and say dedicated physix support on 2 or 3 games compared to 2 other cards.

So in this GT 430 review released today they would have compared it to GTs 450 and GTS 250 in physX and folding.

I figure it'll be easier to revisit the cards after the gaming performance reviews with a quick 1-2 page performance analysis rather than include it buy hmm I guess so far the consensus thinks otherwise.


like i said in the GT 430 thread.. F@H has to many variables.. every WU is different.. so to even remotely show a performance difference you would have the run the exact same WU.. but then that only shows the performance for that exact WU.. theres something like 40 different WU's that GPU's run.. so you would have to show numbers for all 40 WU's.. then after 3 months all those numbers are useless because those WU's have been replaced by new WU's..

i understand your reasoning for it.. i'd also like it, but in the real world the numbers are meaningless and just add more time to the reviews.. maybe when Pande Group extricates their head from their asses and actually releases an openCL client for ATI cards then maybe they could run a quick 2 day test to show the performance of openCL vs cuda.. now that would be far more interesting..
 
People with ATI cards = not interested.

I'm shocked, shocked I say! :eek:

You really need to stop trolling. You troll nearly every possible places just to start the flame war.

I have a 480, and I am not interest at all... Do I have ATI cards? yes on my laptop.. :rolleyes:
 
I could see it now..
"Heres a lineup of the top end offerings from Nvidia and how they stack up for physx. We spent hours looking for games that support Physx and the only one worth our time was Batman. Heres how the cards did".

8800gt: Debrits on the ground shuffled in the air.. slightly... we think.
460gtx: Same thing. Small pieces of paper and leaves occasionally would move.
480gtx SLI: While the frame rates are obviously better than the other two cards, there was no performance difference in the shuffling and skurrying of small shit on the ground.

There you have it folks, Physx at its finest.
 
I could see it now..
"Heres a lineup of the top end offerings from Nvidia and how they stack up for physx. We spent hours looking for games that support Physx and the only one worth our time was Batman. Heres how the cards did".

8800gt: Debrits on the ground shuffled in the air.. slightly... we think.
460gtx: Same thing. Small pieces of paper and leaves occasionally would move.
480gtx SLI: While the frame rates are obviously better than the other two cards, there was no performance difference in the shuffling and skurrying of small shit on the ground.

There you have it folks, Physx at its finest.

Hey funny, but mafia 2 does cool stuff with physX, if you dedicate a card to it, it does even more. Read the release notes on it
 
Back
Top