Tho it seems interesting, if I use VLC to render video in D3D or OpenGL, if it's a 420p video, I'm pretty sure I don't need the extra oomph to render it, so I guess the best solution would be a mix of "autobotonomously" and a switcheroo.
If it ends up working as advertised then it will definitly be a great thing, but after how much trouble the previous solutions where I'm just a bit sceptical.
I'm pretty sure the Nvidia graphics in my unibody MacBook Pro have done this for some time? Switching between the integrated 9400M and the discrete 9600M GPU's as and when necessary?
If they combine that with the external GPU concept, they will really have a huge winner. The inherent limitations of high-level internal GPUs are too many at the moment: space, heat, power consumption, speed that quickly becomes very dated, and insane pricetags that are usually *completely* out of proportion with the desktop equivalents.
Imagine being able to take your notebook, with decent CPU and Ram, but typical integrated graphics, and being able to not only hook up an external gaming GPU, which could be upgraded without needing to buy a whole new machine, and also benefit from the Optimus technology. That would be sweet.
Seems like a source of unreliability to me - I'd just want to run with nvidia all the time - the nvidia card should just have a low power idle state removing the need any onboard graphics. However I can see why they are doing it - now they can't provide chipsets for iWhatever due to intel's bullying they need another way in. This keeps intel happy, while allowing nvidia a way in to sell stuff - very smart way to make money despite the chipset lockout.
On it's own, it's hardly new or all that exciting, but if the use of Optimus Technology becomes part and parcel of external laptop GPU's then this will truly shine. Within existing constraints for laptop graphics, those with non-gaming laptops will have a feature that they will not truly appreciate.
looks fine and dandy, but the last time Nvidia came out with "multiple graphic solution" (SLI) was years ago and up to today they still haven't ironed all the bugs...
i kno theyre different things, but i just cant help myself to compare the two...
Originally Posted by TESCO-Zsömle When I first saw this, just think: BEWARE! Here comes the mighty Marketing Bullshit Monsta... And yeah! (Sh)it happens...
Very old(, long forgotten) tech, and now it's reinvented again...
Actually, it had nothing to do with AMD/ATi or Ruddy Fuddy so it was bulls**t free.
I might be totally off course, but wouldn't the constant copying of data to 2 sets of memory shorten battery life all around?
IGP's running and it's wasting power copying stuff to the descrete's memory. and same the other way round.
Wouldn't the manual solution be better as long as either could be turned completely off without the whole rebooting mess?
Originally Posted by deadsea I might be totally off course, but wouldn't the constant copying of data to 2 sets of memory shorten battery life all around?
IGP's running and it's wasting power copying stuff to the descrete's memory. and same the other way round.
Wouldn't the manual solution be better as long as either could be turned completely off without the whole rebooting mess?
you make it sound like it takes huge gobs of energy to copy information to and from ram. Also as i understand it the second chip would be shut down, even shutting it down for say 30 seconds or a minute would save more energy than it would take to keep the ram going for 10minutes. Ram is only a few watts a video card can top 100.
the only thing going manual would do is let you control when it switches rather than have an algorithm do it for you. You wouldn't be able to get that instant switch feature unless the ram is updated.
Comments 1 to 25 of 33
Reply*head desk*
x2
Rich, you should be ashamed. That is DREADFUL.
Imagine being able to take your notebook, with decent CPU and Ram, but typical integrated graphics, and being able to not only hook up an external gaming GPU, which could be upgraded without needing to buy a whole new machine, and also benefit from the Optimus technology. That would be sweet.
TWO references in one sentence? Really?
If I find them all do I get a gold or a platinum trophy?
It would be made of unobtanium.
hope to see this in H55/H57 platforms for some GPU down-time.
Very old(, long forgotten) tech, and now it's reinvented again...
i kno theyre different things, but i just cant help myself to compare the two...
Actually, it had nothing to do with AMD/ATi or Ruddy Fuddy so it was bulls**t free.
IGP's running and it's wasting power copying stuff to the descrete's memory. and same the other way round.
Wouldn't the manual solution be better as long as either could be turned completely off without the whole rebooting mess?
you make it sound like it takes huge gobs of energy to copy information to and from ram. Also as i understand it the second chip would be shut down, even shutting it down for say 30 seconds or a minute would save more energy than it would take to keep the ram going for 10minutes. Ram is only a few watts a video card can top 100.
the only thing going manual would do is let you control when it switches rather than have an algorithm do it for you. You wouldn't be able to get that instant switch feature unless the ram is updated.
-
« Previous
-
1
-
2
-
Next »
Discuss in the forums