Bitcoin dilemma solution
I wonder if it would be a good platform for resolving the Bitcoin mining dilemma. Those GPU units can usually handle serious number crunching routines.
Amazon has chugged Nvidia's new virtualized GPU technology to spin-up a new class of rentable instances for 3D visualizations and other graphics-heavy applications. The "G2" instances, announced by Amazon on Tuesday, is another nod by a major provider to the value of Nvidia's GRID GPU adapters, which launched in 2012. These …
Not sure if it can run games, previously I am sure they couldn't. Also, even if it can, whilst it may be good for visualisations, it probably wont be good enough speed wise to play games on, the main reason being lag, but i guess it depends on what games you want to play, anything requiring a real time connection is going to suffer from input lag I would expect. MIght be interesting to try it though!
"it probably wont be good enough speed wise to play games on, the main reason being lag"
There was an interesting piece I read about that (it may even have been on el Reg). IIRC it claimed that the lag would be at least as good as modern consoles (at the time). It would not be good enough for a hard core PC gamer, but good enough for a casual gamer.
If you could use technologies like VMView PCoIP or Citrix HDX, you could probably make yourself a reasonable platform for gaming, however, by the time everything is licensed up, you'd probably be better off buying a gaming PC. Unless you're setting up an OnLive type service, where I could see it being beneficial.
a new class of rentable instances for 3D visualizations and other graphics-heavy applications
I think I'm missing something here, because I don't understand who would want to use a cloud-based GPU. Surely the network lag would negate any benefit of the super-fast graphic processing.
The only application I can think of would be something like off-line rendering of CG animated movie scenes or suchlike.
What else? Medical imaging maybe ... ?
It would be fast enough for 3D visualisation; it's not a game, you're not looking for a competitive advantage of few milliseconds. The powerful graphics are so that large assemblies of hundreds or thousands of parts can be accurately viewed, not to shunt frames out a rate of 60 per second.
The advantages are that you can rent the software by the hour, you are not limited by the RAM of your client machine so large assemblies can be viewed smoothly, and when you are ready to render (or simulate) you can throw more CPU/GPUs at it nearly instantly. Also, engineers in different locations can work collaboratively on the same model.
Archicad users would love that, trust me. Even on some today's monster, some rendering take hours. And that's without the fanciest effects (raytracing etc). But architects need to render something 2 or 3 time a year only. They would be happy to be off spending 6k on a computer that works only when it decides to, or some handy geek is around.
Adobe could use it for it's cloud too, or even as option in poser/gimp/pixlr/etc.
That's the main potential usage I see at least.
Sure it can. Grid K2 is essentially a GTX690 without the video output ports (I have a GTX690 modified into a Grid K2 in the ESXi test box under my desk). How well it will run it depends on how many clients you are running off a single GPU.
As a rough comparison, a single Quadro 6000 (essentially a clocked down and shader reduced GTX480) on ESXi vSGA managed about 6 simultaneous 800x600@25fps Borderlands sessions (rendering and encoding into video). Of course, you need a client end that can decompress the desktop video stream coming off the server in realtime.
Biting the hand that feeds IT © 1998–2019