Bests & Worsts Reviews from Amazon

I'm writing this for people who are unaware of the hardware requirements of image processing or research with parallel processing (i.e. the people thinking along the MS paint and minesweeper reviews). There are cases where double precision is necessary to prevent round off error from propagating through an algorithm, so you can't really compare this card to a similar card that has crippled double precision. There are also cases, such as 3D computed tomography, where a very large amount of memory may be needed to fully load the data in. There are machines that have 500+ gb of RAM for this, so you can see why a researcher or company would pay so much just to have a little more memory on a GPU. (Obviously 500 > 12, I'm just giving an idea for how big the data can get and why someone would want more memory) This card could also mean the difference between waiting 2 hours for code to run, vs 1 minute. If a company is paying an engineer $40/hour, then this card makes up for its cost in the time it saves. Perhaps there is a better AMD alternative, but there are several things that might lock a user into sticking with Nvidia, such as Matlab or CUDA. So I guess I wasted a couple minutes writing this review because the MS paint and minesweeper joke reviews were so common, and I don't understand why someone would think they are funny. It would be like laughing at a super car because it gets bad gas mileage or has limited passenger space. No one does that. My only conclusion is that those people are unaware that there might be some guy who actually needs to get from point A to point B that fast, which is image processing in this analogy. Nvidia knows that if you absolutely need to go this fast, you can pay for it. Someone might be able to make the argument there is another card more suitable for these purposes; however, the argument that this card is excessive doesn't make sense, and it isn't funny. tl;dr No one would ever buy this for gaming. They would buy it for research/development employing double precision parallel processing.