Ageia phys-X. Does it worth the money???

[l1zard]

Hi all. i am builting a new system:

asus P5B DELUXE/WiFi
C2D E6600
2Gb(2x1gb) G.Skill DDR2-800 4-4-4-12
x1950xtx

and i was wondering if Ageia phys-X worth those 250$. I saw this sell factor trailer and i must say i was surprised. But the titles that support this card r still limited. On the other hand this card is first of and its kind and as far as i know Nvidia and ATI r both working on their own physics cards. So i think that it would be better to w8 and see some more staff. What do u think,should i buy this card? Would u?

 

[uber_g]

no

wait for the ATI/nVidia physics

 

[slim142]

wait, when nvidia releases theirs, AGEIA will become bankrupt

 

[someonewhoisdead]

separate physics cards will never allow physics to mean jack in gameplay, because developers can never count on everyone having the card.

it will only be as gpu and cpu tech increases that better physics systems will play larger roles in games...

y'know, when you can blow up a building and bury everyone inside... physics processes that arent just eye candy

 

[Mousemonkey]

wait, when nvidia releases theirs, AGEIA will become bankrupt

Thats a bit harsh, most likely correct, granted, but still a bit harsh :lol:

 

[prozac26]

It depends.

If you want bragging right and tell your friend "y0, lookz, i gotz a pointless $250 card in my c0mputerZ taking upz airzflow, and doing nozing". Then yes, it's worth it. :lol:

Seriously, it's not worth it. Nothing is using the card to make it worthy.

 

[niz]

It depends.

If you want bragging right and tell your friend "y0, lookz, i gotz a pointless $250 card in my c0mputerZ taking upz airzflow, and doing nozing". Then yes, it's worth it. :lol:

Seriously, it's not worth it. Nothing is using the card to make it worthy.

meh. Maybe no games really use it yet, but no doubt H/W physics will be the next 'big thing'. What you have is like the 3DFX voodoo 1. Its the first in a new market, so better than the competition (currently none). However, as soon as other H/W physics products come out, it won't be competetive.

There's been talk that nVidia are planning on adding support into their drivers to use a 3rd video card solely as a physics processor.

Thats really a good idea because it now gives us something to do with our old GPUs after we upgrade. Also even a bottom-end nVidia GPU has more computing power than the current PhysX card.

 

[enforcerfx]

Hi all. i am builting a new system:

asus P5B DELUXE/WiFi
C2D E6600
2Gb(2x1gb) G.Skill DDR2-800 4-4-4-12
x1950xtx

and i was wondering if Ageia phys-X worth those 250$. I saw this sell factor trailer and i must say i was surprised. But the titles that support this card r still limited. On the other hand this card is first of and its kind and as far as i know Nvidia and ATI r both working on their own physics cards. So i think that it would be better to w8 and see some more staff. What do u think,should i buy this card? Would u?

Really not worth it. Invest that 250 into a DX10 graphics card. Believe me, you'll thank me later

 

[NightlySputnik]

PPU (Physic processing unit) will become mainstream the day somebody (Microsoft!!!!) release an open standard for it, and not a day before. It looks like it might come with first update to DirectX10 from Microsoft maybe in late 2007 I'd say. Something like DirectX 10.1.

Love it or hate it, Microsoft is pretty much the only software company that has the muscle to do this.

Let's only hope they get it right the first time... this time. ATi and NVidia would probably gladly join in if it wouldn't represent to much change to their silicon.

What do you guys thing of this???

 

[3lfk1ng]

Around game are scheduled to used ageia next year. Only 5 of them are going to be good, UT2007 is one of them....but the card is still not worth it.

 

[NightlySputnik]

we or at least i do not bloody well care as this as been discussed not only in the many dedicated threads relating to this topic but alos in all the offshoots from other threads.

great flame wars between trolls and others have been fought and we don't need another. this questions should not have been asked so let this thread die for the sake of the everyone.

That's strange... stranger

No offence, but if you don't care, just stay out of the topic ... I gave an opinion and ask what other guys think about. Nobody's forced to answer, and that should have applied to you.

:idea: Most of us are here EXACTLY BECAUSE we like to speculate and tap ourself in the back when we're right. Those of us who want hard fact are better to: a) stay on the official news site, b) ignore whatever is speculations.

Don't worry, I'll gladly discuss anything with you that is in both our sphere of interest. It's only that I don't need to be offended because you don't "care" about what I ask... Plain and simple... :twisted:

 

[NightlySputnik]

LISTEN UP. WHAT YOU WISH TO DISCUSS HAS ALREADY BEEN DISCUSSED. YOU CAN ADD NOTHING NEW UNLESS THERE HAS BEEN SIGNIFICANT NEWS.

I HAVE NO PROBLEMS WITH WHAT PEOPLE DISCUSS UNLESS IT HAS BEEN DISCUSSED BEFORE IN WHICH CASE READ THE PREVIOUS THREADS AND COMMENT IN THOSE.

WHAT YOU WILL FIND IS THAT YOU HAVE NOTHING MORE TO OFFER SO SHOULD JUST KEEP QUEIT.

AND YES YOU DO DESERVE SHOUTED AT

LIKE I SAID BEFORE, IF YOU'RE NOT HAPPY WITH WHAT I HAVE TO SAY, GO LOOK ELSEWHERE.

SOME OF US HAVE A LIFE AND DON'T TAKE 3 HOURS A DAY TO READ ALL TREADS. SO IF I WANT TO SAY IT HERE, SHUT UP AND IGNORE ME IF YOU CAN'T STAND IT. :x

I'm a nice guy when poeple are nice with me, but I'm 3 time as bad as they are when they're not. AND THIS APPLY ESPECIALLY TO YOU A1234LE!!! HOPE YOU GET IT.

Now I'll do what you should have done before and ignore any stupidity you have to add.

Sorry for everybody elses here.

 

[l1zard]

Sorry for everybody elses here.

i am with u

 

[NightlySputnik]

Thanks,

It's nice to have somebody put a smile on my face. :lol:

 

[Talon]

Wait and see is my advice and what I'm doing. I do believe it has a very real future in concept tho. Can't wait for the day interacting with objects and dare I say even whole landmasses in the environment can be pivotal to play.

I also agree with you that not all have time to look at all posts. I was recently jumped on like you because I posted something LESS THAN 24 HRS OLD. I was told it was a subject beaten to death even tho it was brand new info and I should look to other threads. I love tech, yet I have a life and honestly can't be bothered to search thoroughly every time I want to help others with a post or get help with something for that matter.

Reminds me of DasickNinjas recent post about etiquette. Imo if I see a subject beaten to death and don't wish to comment I just move along. Doesn't hurt a thing to turn a blind eye to such trivial things. Woops got long-winded. Peace-out !

 

[kaotao]

no it aint. of course a simple bit of research would have told you that.

I agree. They're about as useless as more DX10 threads.

 

[DeathwingX]

this topic has, and i understand you posting, im like you cant exactly read each and every thread and sometimes it is so deep you cant find. my thought would be to wait until nvdia or ati comes out with there software/hardware support and then use what they give you. by the time that rools around my 7800gtx will be sitting there doing nothing, and i have a 7000 ati radeon (woot woot!! go old school!!!!!!!!) that could easily handle the physics of my games. im hoping that microsoft uses there insane muscle to push for the physics, it wold make games more realistic. right now games are limited to simpler equastions due to the processing power that is used. the cpu has to handle the os, the game, and then physics. once nvidia/ati comes out with there physics it will become very easy to implament and it will also end up selling more of their low end chips

 

[NightlySputnik]

this topic has, and i understand you posting, im like you cant exactly read each and every thread and sometimes it is so deep you cant find. my thought would be to wait until nvdia or ati comes out with there software/hardware support and then use what they give you. by the time that rools around my 7800gtx will be sitting there doing nothing, and i have a 7000 ati radeon (woot woot!! go old school!!!!!!!!) that could easily handle the physics of my games. im hoping that microsoft uses there insane muscle to push for the physics, it wold make games more realistic. right now games are limited to simpler equastions due to the processing power that is used. the cpu has to handle the os, the game, and then physics. once nvidia/ati comes out with there physics it will become very easy to implament and it will also end up selling more of their low end chips

Agree! Like I said before before being choke to death ( ), I think the best way to get physic to become mainstream will be through Microsoft.

This is because no game designer will go through the assle (wrong spelling? Sorry if that's the case... Oh! me poor french canadian ) of putting 3-4 physic engine in their game. That would be Ati, NVidia, PhysX and Havok. Add the ones not known to me... you should get it.

That would be perfect if Microsoft does it with Ati and NVidia beside them and not pushing it through their mouth.

My opinion, for those of you interested in it. :mrgreen:

Cheer!

 

[TheGreatGrapeApe]

Microsoft is involved already.

DirectPhysics will likely be in the first refresh of of DirectX10.

 

[CannedTurkey]

There are currently no titles that push the tech far enough, to actually require the card. In most cases, just installing the drivers will allow you to experience enough of the 'whiz-bang' effects in the titles that do currently support the Ageia card (especially if you've got a dual-core system).

Personally, I'd wait for UT2007, and some good review at that time, to make the decision. The Unreal 3 engine has support for the Ageia tech built in, and theoretically, if it makes a reasonable difference there, it will help a lot of additional titles.

Even so, adoption is going to be slow. It's very much a chicken and egg dilemna right now. Developers won't push it too far, because there isn't a wide adoption of the hardware yet, and the hardware won't get widely adopted until they absolutely need it for performance reasons.

 

[casewhite]

this topic has, and i understand you posting, im like you cant exactly read each and every thread and sometimes it is so deep you cant find. my thought would be to wait until nvdia or ati comes out with there software/hardware support and then use what they give you. by the time that rools around my 7800gtx will be sitting there doing nothing, and i have a 7000 ati radeon (woot woot!! go old school!!!!!!!!) that could easily handle the physics of my games. im hoping that microsoft uses there insane muscle to push for the physics, it wold make games more realistic. right now games are limited to simpler equastions due to the processing power that is used. the cpu has to handle the os, the game, and then physics. once nvidia/ati comes out with there physics it will become very easy to implament and it will also end up selling more of their low end chips

Well thought out with a few additional thoughts . I have been an IEEE certified engineer since 1975. Economics drives what hardware we get. The market that matters is the business market and for the high end the workstation market since workstation sales in 1 year equal about 3-4 years of sales in the enthusiast market. People like Boeing, Lockheed, Newport News, and Electric Boat use workstations in the thousands. Anything that improves work quality for them is immediately bought and the smart developers design for that market. Stanford Medical School already has the ATI Physics design up and running for reasearch in disease control The performance is about 400 gigaflops in double precision or 10 times the very best Core 2 's score in single precision. So it is effectively 20 times faster than core 2 and 10 times than any any quadcore that is not here yet. This is done using a Opteron 150. The CPU becomes a processing manager sending work out to where it can be done most efficiently and linking system memory. . The only disadvantage is that the graphcis cards are power hogs. For very large servers and supercomputers the IBM Cell chip for the Playstation 3 is the ultimate . DOE is building Roadrunner at Los Alamos Labs with a design that you could compare to 16,000 desktop hooked via an Infiniband network. It will be 100gflops per pair. That is 1.25 times the quadcore and again it is double precision (64 bit)where the quadcore is single precision(32 bit). Power usage is 1/3 of the quadcore. When the Cell+ comes along next summer expect the performance to double. Here is a good explaination as to what the IBM QS20 design does. It is the basis for Roadrunner.
http://www.supercomputingonline.com/article.php?sid=11894
QS20 link http://www-03.ibm.com/technology/splash/qs20/
This is the Stanford folding project. http://www.pcper.com/article.php?aid=305&type=expert Game developers will be able to build on the Cell chip combinded with either nVidia's or ATI's GPU's to produce 1 teraflop in the desktop which will be 25 times faster than core 2 and 12.5 times faster than either AMDs' or Intel's quadcores.

The business economics of the Cell-GPU combo for workstations , servers and supercomputing makes the Ageia design a dead end and it makes quad core useless since it requires special software encoding that is not likely to be funded. The US Dept of Energy has already paid for and written the code for the Cell-GPU acceleration. I am running it at home on a Opteron 154 an ATI X1800XT Crossfire setup. Intel doesn't have the IMC that is required to make this work. The CPU needs full access to memory and not through a northbridge for efficiency. NVidia should have its act together within the next few months and enter the fray.

 

[TheGreatGrapeApe]

the next few years are goinng to be very interesting but then in technology what years aren't.

900 - 1400 A.D. ? :twisted:

 

[TheGreatGrapeApe]

Yeah I think it was the PA-1

Pigmy + Abacus. :mrgreen:

 

[Coops07]

i'd learn to speak english better.

 

[NightlySputnik]

Microsoft is involved already.

DirectPhysics will likely be in the first refresh of of DirectX10.

Just what I tought and wrote before.

That's good, this put this tread at around 40+ post. It should be a good start for any search with AGEIA in it. Maybe all the 1337 poster in here will now gladly join in after all.