Digging Deeper

So there we were, left with G-Sync somehow working on a regular laptop. It just sat there on the table in front of us, showing off its smoothly sweeping pendulum. We bounced around some working theories, one of which was that perhaps this laptop series had somehow shipped with development hardware that ASUS was working on – perhaps for a future G-Sync branded laptop. There was only one way to find that out, and in a blur of iFixit tools, this happened:

img-4480.jpg

We followed the various traces and wires from the GPU output all the way to the LCD panel TCON (the circuitry that translates the incoming eDP signal and fans it out to drive the LCD pixels). Examining each and every component, we found no additional FPGA or ASIC. All hardware was what we would expect for a standard laptop:

img-1425.jpg

We even peeled back the back of the LCD itself and examined every component. No G-Sync:

img-1456.jpg

With the prototype G-Sync laptop theory dead and buried, we then moved to our other working theory. Perhaps the leaked ASUS driver *was* intended for an upcoming G-Sync laptop, but not this one, making this a case of accidental support – meaning that for whatever reason the driver *thought* there was a G-Sync panel in this laptop. It is entirely possible that NVIDIA uses a protocol very similar to the Adaptive Sync standard, and that the LG panel Tcon in this specific line of ASUS notebooks is able to handle that input in some scenarios as it is ‘close enough’ to what it can handle. If this were true, our knowledge of variable refresh rate technology tells is that the experience would not be ideal across the entire refresh rate range.

We know that you can’t simply flip a switch and let any LCD refresh variably at every possible rate. There are side effects to be dealt with, like the fact that pixels drift between refreshes, and waiting too long results in excessive flicker. Adding additional refresh cycles is possible, but that is not a simple solution as the additional refresh must be carefully timed as to not delay a subsequent frame being delivered by the GPU. Complex design problems like this have delayed AMD’s FreeSync launch, and it took a revision to NVIDIA’s own G-Sync module to get that timing just right.

Alright, back to the laptop to test this theory. We needed to push into the lower frame rates and prompt some of the other hiccups that we know to be challenges for even a real G-Sync panel to handle. I quickly reassembled the unit, fired it up, and got back to testing.

« PreviousNext »