Originally posted by: batmang
Originally posted by: Sirlaughalot87
But why are the market going the "multi" GPU way! I don´t want to use bad scaling tech as SLI or Crossfire, i want ONE card, i don´t want two of those energy leechers in my case, i don´t want two cards that can malfunction; One card is more then enough...as it should be.
Why twin turbo when you could use one large turbo.
(This makes no sense just like this thread.. but I'm posting it anyway.)
In this case, that's the exact opposite analogy.
One turbo takes longer to spin up (and spin down) than two small ones. This causes 'turbo lag', or the perception of unresponsiveness. Latency is critical in the stoplight grand prix, quarter mile and even track racing. In almost all cases you can anticipate need and try to build up a full head of steam in the turbo before you need the power, but this requires a lot more driver attention and skill than the guy with the great big chuffing V8 with cylinders the size of Folgers coffee cans, a roots or Lysholm screw type supercharger and 600 ft-lb of axle snapping torque off idle.
The other approach to multiple turbos is multi-stage turbo: the first turbine feeds the second one.
In any event, both two smaller vs one larger approaches reduce 'latency' -- the time you need to wait between sticking your foot in the turbo and feeling your face peel back. The second approach also allows for moar airflow and pressure! Since everyone knows air + fuel in = power out...
In the case of multi-GPU it's the other way around. Multiple slower GPUs will give you significantly more input lag than one fast GPU if used in AFR mode.