Jump to content

Asus Ageia Physics Card release date?


SteelHack

Recommended Posts

i dunno about that Specter. It seems worth it to me ...

in 2 years your video card will be outdated so what?

things always change, It just depends on what ATI, Nvidia, AMD, and intel do that make the card not worth the price. For the short term thoe it looks worth it.

Link to comment
Share on other sites

  • Replies 86
  • Created
  • Last Reply

Top Posters In This Topic

i dunno about that Specter. It seems worth it to me ...

in 2 years your video card will be outdated so what?

things always change, It just depends on what ATI, Nvidia,  AMD, and intel do that make the card not worth the price. For the short term thoe it looks worth it.

That's true, and I can see that point of view.

But I got to thinking about what Papa6 said about CPU's, and how now that we have gotten to dual core, that is just the beginning.

I think we're going to see a point now where alot of this extra stuff, especially physics, can be and will start being handles by a core on the processor. Afterall, that's what multiple cores are designed for.

I think in the next 2-4 years, video aside, you are going to see where just about everything is handles by the CPU, and functions are split up amongst the cores.

Look at dual core now. For the most part, that second core sits there idle. Sure, there are a very, very few apps that can utilize multiple cores, but that is changing.

And I think that gaming is going to be one of the first to really utilize the added cores. And why not use the idle one for physics/ It would be better that way. Even having to boost RAM and take advantage of 2-4 GB, even more. That would actually make more sense to spend money on RAM, because almost everything benefits from more RAM these days.

Sure. That add-on physics card might be good now. Maybe. But I think that by the time game developers get onboard with it to the extent that it makes this board a practical purchase worth the expenditure. they will have the games using the other cores on the CPU's for the job, and it will be a moot point.

JMO. :thumbsup:

Link to comment
Share on other sites

But why not do things this way. since we aren't but "just" tapping dual cores, why not thread the physics through the second core? core 1 will do the game and core 2 will process the physics.

No offense, but the reason why not is the exact same reason why you arent going to use the second core instead of a graphics card. Because it cant do it. A CPU isnt optomized for graphics processing, or physics processing. Besides games with multithreading (most new games will have it) are doing something even better. Rather than just using one core for stuff and one for physics, they use both for everything. More efficient use of space that way.

The simple fact is that without this card you wont get what you will have with this card. Not for a long long time anyway. Thats not even up for debate. The question is whether what you get from the card is worth what you pay for the card.

For me personally: I have parents who are paying for food, clothes, house, etc. I have a job giving me money. And I have a 50 dollar Best Buy gift certificate that is reserved for the purchase of GRAW (I figure they will scam me less on software than hardware, and I need to use it for something <_<). So the question is what do I do with that money. I want some scuba gear. Expensive stuff and I wont use it much. I want GRAW, but thats taken care of like I said. Then I want computer hardware. New graphics card maybe? I have GeForce 6800GS which will almost certainly play GRAW at 35+ FPS on my little monitor (1024x768) which is my definition of playable. I could buy a new one, but I would rather wait until there is some new cool shader effects, direct X version, etc out. And there is no game out there that would benefit significantly from a processor upgrade. With a 3800+ x2 my computer isnt processor limited. So what am I going to do? I am going to wait until I see reviews of this card to confirm that it actually works, then I will buy one.

Thats just me. I am not saying its right for everyone, and obviously my financial situation is sort of unique in that if I blew every dollar I have all it would mean is that I dont have any more to blow. But again, physics is something that really excites me in a game, even more than graphics.

Link to comment
Share on other sites

FROM Insane Snyper: The simple fact is that without this card you wont get what you will have with this card. Not for a long long time anyway. Thats not even up for debate. The question is whether what you get from the card is worth what you pay for the card.

ok, if you read back a page or two, you'll see some of us don't think the extra $100-$300 is worth the AGEIA PhysX Card. As for your comment about cpu's not being able to process physics, I'd disagree. Physics is nothing more than math. Every processor today has enhanced floating point units for math computation. If these PC game DEVS do release a multi thread game(s) I would say it's safe to say that the physics will be pushed through one of the cores, would have to. Not everyone is going to go out and buy the AEGIA PhysX card. PLus today, CPU's do have to process physics. I believe Havoc is the top of the line...how is the havoc physics computed then? The processor. Not arguing with you but coming from honest reality.

Link to comment
Share on other sites

But why not do things this way. since we aren't but "just" tapping dual cores, why not thread the physics through the second core? core 1 will do the game and core 2 will process the physics.

No offense, but the reason why not is the exact same reason why you arent going to use the second core instead of a graphics card. Because it cant do it. A CPU isnt optomized for graphics processing, or physics processing. Besides games with multithreading (most new games will have it) are doing something even better. Rather than just using one core for stuff and one for physics, they use both for everything. More efficient use of space that way.

The simple fact is that without this card you wont get what you will have with this card. Not for a long long time anyway. Thats not even up for debate. The question is whether what you get from the card is worth what you pay for the card.

I think it can be done. They are doing it with software now. Anfd I don't think dual core is going to be around long now. I think we are going to see more, and soon. And even so, games still aren't using all of the dual cores, if most games are using both cores at all, which they aren't yet.

THe processor not being able to do it? Yes it can. Super computers do it everyday, and if you think about it, they are nothing more than multi-core processors. THey just need to be told what to do.

And akk that physics card is, is a dedicated processor wqith onboard RAM, nothing more.

Link to comment
Share on other sites

ok, if you read back a page or two, you'll see some of us don't think the extra $100-$300 is worth the AGEIA PhysX Card. As for your comment about cpu's not being able to process physics, I'd disagree. Physics is nothing more than math. Every processor today has enhanced floating point units for math computation. If these PC game DEVS do release a multi thread game(s) I would say it's safe to say that the physics will be pushed through one of the cores, would have to. Not everyone is going to go out and buy the AEGIA PhysX card. PLus today, CPU's do have to process physics. I believe Havoc is the top of the line...how is the havoc physics computed then? The processor. Not arguing with you but coming from honest reality.

I'd have to disagree with you here. The PPU is essentially a special purpose vector processor, as opposed to a general purpose processor such as a CPU. A vector processor is one designed purely to carry out vector operations as quickly as possible and do nothing else. A general purpose CPU is designed to do pretty much anything imaginable, and while it can do vector operations aswell, all the extra circuitry it has for other things slow it down immensly.

Physics calculations are vector operations, which are essentially carrying out simple mathematical equations across large sets of data (for e.g. increasing the downward velocity of all objects by 9.8M/s/s, gravity). Most graphics operations are too (for e.g. rotating a mesh of a few hundred/thousand vertices/points by a few degrees). A processor specifically designed to carry out these kind of calculations will easily out do a general purpose CPU, so a general purpose CPU will never replace dedicated vector processors. Super computers designed for scientific modelling/simulation use vector processers, there is also software out there that lets you use a graphics card to run scientific modelling/simulation software because they're essentially very similar. Vector ops also lend themselves well to being easily multithreaded which is why a lot of the talk in graphics cards these days is about the amount of pixel pipelines a card has, and also I believe the Ageia PPU is multi-threaded.

I've kind of lost track of where I was going so I'll try and come to some sort of conclusion :) Physics can be done in software on a general purpose CPU, such as with Havok in HL2, Lockdown etc, but in that we only ever see a small number of objects moving at once (and that's for a reason). A dedicated PPU will enable orders of magnitude more objects moving, and much more realistic enviroments, while leaving the CPU free for other things. It was mentioned above that next gen consoles don't have PPU's, but the Cell (PS3) processor has 8/9 SPU's which are essentially the same thing. The 360 has no such equivalent and I predict that will become noticeable as both platforms mature. I think this also indicates where the future of PPUs will go, I think that, as with Cell, we'll see dedicated vector processors being added to CPUs, so there may be 4 general purpose cores (as Intel and AMD are prepping now), plus 4/6/8/however many SPUs to handle physics and other vector ops (video/music encodeing/decodeing, compiling, translating etc.) (edit: assuming we're not all using cell based PCs in a few years)

In fact the wikipedia link below explains it better than I can and is worth a read.

Graphics card used modelling collapsing star

http://www.newscientist.com/article.ns?id=dn7508

Vector Processors

http://en.wikipedia.org/wiki/Vector_processor

Edited by stevenmu
Link to comment
Share on other sites

I'd have to disagree with you here. The PPU is essentially a special purpose vector processor, as opposed to a general purpose processor such as a CPU. A vector processor is one designed purely to carry out vector operations as quickly as possible and do nothing else. A general purpose CPU is designed to do pretty much anything imaginable, and while it can do vector operations aswell, all the extra circuitry it has for other things slow it down immensly.

...

It was mentioned above that next gen consoles don't have PPU's, but the Cell (PS3) processor has 8/9 SPU's which are essentially the same thing.

...

You already said it in your post. There is the dedicated PPU (Ageia) and there is the future in computing for entertainment and multimedia, the Cell architecture. With a Processing Unit similar to current CPU-Cores and the Synergistic Processor Elements, but both of them with a general design not a dedicated to physics processing. They a high speed connected (not pci-e) with access to the same memory and separate caches. All this for a $300 price.

Why should one invest $200-250 in a dedicated PPU card with only a few supported games. The lifetime of that card is relative short IMHO.

How about a Cell powered board for the PC? There are so much possibilities not only gaming.

Link to comment
Share on other sites

I did read an article before about the threat cell posed to the x86 architecture and a possible cell based add on board was mentioned (altough it was only speculation). This would be an excellent solution and well worth $300 dollars. As I mentioned it would also be possible to integrate cell like SPUs with an x86 PC, or as others have mentioned video cards could integrate extra cores for physics or other non graphics tasks. The problem with all these is that they're a long, long way off. This ultimately is what makes the Ageia card worth purchasing, for me at least. It's the only thing on offer that offers this functionality, and is likely to be for some time yet.

Link to comment
Share on other sites

As for your comment about cpu's not being able to process physics, I'd disagree. Physics is nothing more than math. Every processor today has enhanced floating point units for math computation.

He's right, Papa.

Our system processors are designed to do everything, but nothing efficiently. Graphics are all math, too. Everything in a computer is all math, so if what you were saying were true, we wouldn't need GPU's either. However, we still need GPU's, processors that are designed to do that math needed for graphics in a way that is most efficient.

I read an article a few years ago by a processor engineer where he talked about exactly this. He said that it would be simple to design a processor to do video encoding at the same rate as our system processors but at 10% of the processing speed, if the proc were designed to do ONLY video encoding.

It's the same thing here -- a processor designed for physics calculations. Our system procs can do it, but NOWHERE near as efficiently as AGEIA's processor.

--Logos

Link to comment
Share on other sites

As for your comment about cpu's not being able to process physics, I'd disagree. Physics is nothing more than math. Every processor today has enhanced floating point units for math computation.

He's right, Papa.

Our system processors are designed to do everything, but nothing efficiently. Graphics are all math, too. Everything in a computer is all math, so if what you were saying were true, we wouldn't need GPU's either. However, we still need GPU's, processors that are designed to do that math needed for graphics in a way that is most efficient.

I read an article a few years ago by a processor engineer where he talked about exactly this. He said that it would be simple to design a processor to do video encoding at the same rate as our system processors but at 10% of the processing speed, if the proc were designed to do ONLY video encoding.

It's the same thing here -- a processor designed for physics calculations. Our system procs can do it, but NOWHERE near as efficiently as AGEIA's processor.

--Logos

And with all of these multi-core's out now, I can see where it would be possible and a helluva lot more efficient to make it so the additional cores handle this stuff efficiently. There is no reason not to, and why on earth can't the physics software be made to make it more efficiently handled by an additional processor core? After all, that just makes it like having two procs, like adding that physics board, and even though that add-on board is going to handle physics, software still has to tell it to, right? So it all still comes tdown to the software being inefficient.

Link to comment
Share on other sites

Its really not more efficecient to make multi core processors have different cores be specialized for different tasks. The whole point of a cpu is that it is "central", it does everything ok, but not much really well. Obviously it would be cool to buy a motherboard, and a processor that has an AI core and a physics core and a graphics core and a sound core, and an mp3 encoding core etc, but its just not practical. Either everyone would have to buy a super expensive processor that can do all of that stuff, or they would have to have a bajillion different models of processor with different combinations of abilities in different ratios. That would be hard or retailers and customers alike (not to mention manufacturers). Better to have them be separate chunks that you can plug together to build a computer optomized for whatever you are trying to do.

Link to comment
Share on other sites

And with all of these multi-core's out now, I can see where it would be possible and a helluva lot more efficient to make it so the additional cores handle this stuff efficiently. 

I agree. In a previous post, I suggested that AMD or Intel should by AGEIA and use the tech to make a core specifically for physics processing. It's an interesting proposition.

--Logos

Link to comment
Share on other sites

I guess the only way to settle this will be to see how GRAW (and other games too I suppose) utilise the Ageia card, what difference it'll make to game play, and how it'll affect the whole experience. There may in the future be prefferable ways of implementing Ageia in hardware, but for the foreseeable future we'll simply have the choice of either buying the card for the extra stuff it brings to the game(s), or not.

Personally, I plan on waiting to see what difference it's going to make (I'm sure someone will be kind enough to upload some videos with and without Ageia), and untill the initial bugs are worked out of the system and hopefully the price drops a little.

Link to comment
Share on other sites

I guess the only way to settle this will be to see how GRAW (and other games too I suppose) utilise the Ageia card, what difference it'll make to game play, and how it'll affect the whole experience.

Precisely. While this technology has extraordinary potential, whether or not that potential will ever be fully explored will depend on how many games employ it and thus push gamers to adopt it.

The Unreal 3 engine employs it, and I'm sure many games will be based on that engine. My guess is that Unreal3 is going to be this technology's best bet for success. However, we'll have to wait and see.

--Logos

Link to comment
Share on other sites

  • 1 month later...

Taken from HERE.

About GRIN

GRIN was founded to create state of the art games using breakthrough technology. GRIN has long experience in developing both realistic and sci-fi games, all while pushing the envelope in the graphical realm. Past titles include the award-winning “Ballistics”, “Bandits – Phoenix Rising”, and numerous arcade machines as well as military and civilian simulators. Founded in 1997 by the Andersson brothers, GRIN today stands fast with 45 dedicated developers with their mind set on elevating PC and console gaming. http://www.grin.se

I don't think this link has been posted.

Link to comment
Share on other sites

Hello guys. Long time reader, first time poster.

If you want to see a more real-world "practical" use of the PhysX card, check out the nearly 400MB Bink format movie of Cellfactor in action, here: http://www.cellfactorgame.com/

The video link is at the bottom left of the intro page.

Hopefully, that will change your minds about how little you think this card would be able to affect new gameplay scenarios.

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.


×
×
  • Create New...