The aluminum-plated underside of the M65 RGB Elite is home to more than the mouse's sensor. The RGB LED that illuminates the Corsair logo above also vents light out the back of the mouse, which you can see in the image above. There are also three removable weights that are hiding under those three slotted screws. Each weight is six grams, and when all three are installed, the mouse clocks in at 115 grams; by shedding the weights, then, you can drop it to 97 grams. Some FPS players won't use a mouse above 100 grams, but other gamers, myself included, prefer meatier mice, so the removable weights help it appeal to a broader audience.
The M65 RGB Elite is equipped with the same the Pixart PMW3360-derived PMW3391 optical sensor as the Ironclaw. The PMW3391 in my Ironclaw passed all the challenges I threw at it and never let me down. However, it's still worth performing the same tests again because the sensor itself is only one component of mouse input performance. It's critically important to determine whether the mouse messes with the raw data from the sensor. Both Zak and Jeff have had bad experiences attempting to game with mice that screw with sensor data in different ways.
I didn't notice any strange mouse behavior with the M65 RGB Elite during my time playing Warframe and Ironsight. I couldn't make the sensor spin out or lose tracking by slamming the mouse down at odd angles. However, I moved on to more objective tests to verify my subjective experience. I hopped into CS:GO and ran a simple test with raw input on and mouse acceleration off. I moved the mouse horizontally back and forth between two books. The crosshair consistently moved between the two endpoints set by the books, confirming what my experience gaming seemed to show: The mouse has no innate mouse acceleration.
Once graphed, the mouse data collected by MouseTester looks similar to the results from the Ironclaw. As I said in my Ironclaw review, the curves are a little less smooth than some of the curves produced by other PMW3360 variants, but a handful of outliers at the millisecond level shouldn't produce any noticeable tracking issues.
Just like the PMW3391 in the Ironclaw, the M65 RGB Elite's PMW3391 updates roughly once a millisecond. This update period translates into a 1000Hz polling rate, which is the advertised max polling rate.
I actually switched from the ancient dinosaur that is MS Paint to Paint 3D for my angle-snapping test because I was having difficulties with MS Paint (did I mention it's a dinosaur?). Regardless, my quickly drawn lines and spirals show no marks of angle-snapping. All told, my tests confirm the dependability of the PMW3391 and reveal no processes built into the M95 that tinker with the raw sensor data.
You can read my full comments on Corsair's iCUE peripheral software in my Ironclaw review. However, to summarize, iCUE needs a little help, particularly when it comes to mice, in the ease-of-use department. It isn't always apparent how to access certain options without a bit of digging.