Anyone that thinks GPU mining will die is trying to get you to quit or has too much tied up in their ASIC bricks. I panicked four years ago because of naysayers and widespread selloff/panic. If only I had stuck with it, my $20k+ investment in mining would have been worth $250k - $500k today. Alt Crypto currency cannot survive without mining, and since crypto is here to stay, mining is here to stay. GPU mining is waaaayy safer than ASIC mining, especially now. I bought an S9 back in December when it was making $30 a day after power costs. I sold it the following week; if I had kept that S9, I would have lost $2k in two months as you can now get them for $2k and they only make $6 a day. $5k spent in graphics cards will pull in way more than $5k spent on ASICS. You can't win with ASICs because ASIC algo difficulties ramp up way too fast to keep up. They only ones making money on ASIC mining are the chinese companies that are in bed with Bitmain and Bitmain themselves. The house always wins and in this case the house is Bitmain and their cohorts. They are mining with tomorrow's SHA-256 ASICs today.  But the chinese do not have control of the GPU market. GPUs will hold more than 75% of their value for at least 6 months after the new GTX cards are launched because no one will be able to get the new cards, they will be in just in time for the holiday rush of gamers and miners this fall. Additionally, new GPU pricing will reflect the market demand; if they give 30% better mining performance then they'll be 30% more expensive, so don't expect previous gen GPUs to simply tank. Gamers will still need GPUs and previous gens will be the only affordable option for them. ASICs cannot replace GPUs because Bitmain simply does not have the ability to produce or obtain the GDDR5X/GDDR6 ram that Samsung has a monopoly on. Cellphones and GPU's are top priority for Samsung. Bitmain would have to sink billions into R&D just to bring out a competitive ASIC to current GPU levels. It doesn't work the same way SHA-256, X11, etc ASICs operate.