It's not the taste of water I object to. It's the after effects - Ronald Knox
AGEIA CAME BACK to us with information regarding its second PhysX chip.
This chip is now called PhysX 100M, and as the name says - it is planned for implementation in different notebooks, starting with Dell's The Beast concept, now known as XPS M1730.
Back in April, we reported about Ageia making a die-shrink and preparing a laptop module. All of this was based on upcoming die-shrink, meaning original 130nm PPUs were going the way of do-do bird. Now, Ageia's board is available in PCIe form, at least if your name is Alienware, Dellienware or Acer. Of course, all the attention of the die-shrunk PPU should go into Dell's latest gaming machine, known under the codename "The Beast".
Reason why Ageia got a good mobile design was the fact that PPU now only eats and dissipates around 10W, which is quite mobile friendly. If you team up the card with 128MB GDDR3 memory, it's rather easy to figure why an Ageia MXM card is now a hot property with notebook vendors. Dell has the exclusive for now, but we'll see what will happen in near future.
Of course, don't expect a new Ageia chip to be the size of Texas, as some sites rumoured in the past. µ
This has been around for how long? I haven't seen one good reason to even want one. Its turning into a flux capacitor, you don't know what it really does but apparently you need one.
Who needs a PPU when most games can't even keep 2 CPUs busy let alone 4?
Dodo, shurely?
So now you can spend over €200 for some crap doing nothing. But hey, at least it only consumes 10W...
and now it's even more expensive...can't they (Ageia) realise,the fact that NO ONE buys that shit.
@Labrador
"Who needs a PPU when most games can't even keep 2 CPUs busy let alone 4?"

I would not use the word 'need'
It's not that current games can't keep 2 CPUs busy. It is that the developments would not want to keep all CPUs busy. Because that would mean most people with single core CPUs will not be able to enjoy much of the game

It takes not a lot of effort to write increasingly complex physics algorithms for a more realistic game.
It seems ALL the next gen physics solutions have something wrong with them.

Nearly everyone seems to hate the PPU, refusing to give it a chance and actually chanting for it to go away.

Havok FX has crawled under a rock, but continues to keep gamer interest despite having worse game support than the PPU and potentially more expensive hardware costs.

There is only one announced game that actually uses dual core for more than just FPS gains (Alan Wake), let alone quad, and it will take many years till CPUs do anything in games to the level of the PPU and Havok FX.

Next gen physics is increasingly sounding like a pipe dream; with the reality being very small and slow improvements as we have previously done, but what can you do about it?
There's some value to offloading physics calculations to a dedicated processor. If its been designed from the ground-up to deal with physics(only) algorithms it can be designed to be a lot (orders of magnitute) more efficient at it than a CPU which has to be designed for general purpose.
Also lets be honest, even the latest games physics usually suck. Shoot a heavy crate and it zooms across a room. WTF?
Also, with Vista being such a resource-hogging mess, your quad-core cpu is already going to be fully busy with getting permission from microsoft to do anything each time you press your mouse button.