Is that like the old 486 sx and dx chips, the same chips but the sx had cache disabled, yet there and the dx had it enables. Both chips way back then cost the same to produce as they were identicle except enabling internal cache, yet the dx sold for way more just to enable a feature already manufactured in the cheaper chip but blocked.
Kinda, yeah. Although microchips everywhere work this way.
For example, all three Ryzen 7 chips come off the same machine. They just test them out to see which ones manage to successfully validate at the higher clockspeeds. Many of the 6-core and 4-core versions released later will just be 8-core chips where one or more cores are just defective, or failed validation at the desired speed. So they just disable them and sell it as a lower-end chip.
(if past generations are anything to go by, some idiots will attempt to unlock these defective cores and then whine incessantly when it doesn't work)
In GPU production, often the lower range models are just cut down versions of the big one. Chop off a couple hundred processor units so you get more devices out of the same wafer. In the case of the 1080ti, Nvidia actually did something new and didn't cut it down. It's a full-on Titan. They just put on slightly less memory (11gb vs 12gb, but actually with memory that is clocked faster) and a better cooler. In comparison the 980ti was actually cut down a bit, and had half the memory of its Titan variant.
Oddly enough, it's also a larger price gap.
Titan X (Pascal): $1200, 1080ti: $700
Titan X (Maxwell): $1000, 980ti: $650
My guess this is a preemptive strike against AMD's "Vega" GPUs. "Hey, only $700 and you get full Titan power! Why buy this piece of junk from AMD?"
And for ****'s sake, Nvidia, why did you name both generations the "Titan X?"
