Pricing is crazy right now. Amazon has 1080ti at close to $1500 on some models. Never knew about this mining market. This crap keeps up I will not be able to hang with the prices.

For the last 3 days, the only cards Newegg itself has had in stock when I have checked a few times each day that are worth even THINKING about mining on have been ONE Zotac 1060 3GB model (at a somewhat inflated price) and 8 or 9 2GB 1050 models.
Nothing AMD RX 470/570/480/580, nothing Vega, nothing else 1060, nothing at all 1070/1070ti/1080/1080ti
 
I've also been getting ZERO notifications on anything I still have notifications on (basically ALL of the 1070 ti air-cooled models with ball-bearing fans).
 
I don't think it's the Miners at this point, I think it's "no production or almost no production out of Nvidia" on Pascal GPU chips, and ditto on AMD out of their higher-level GPU chips.
 
 

AHowes
So memory crisis?? But if so why is there no official word on it all?

No, GPU. There was a GDDR5 shortage months ago but that's long past and I've not heard anything that would indicate it's still going on. It takes months of advanced ordering to do a run of GPU chips for Nvida to use themselves or distribute to their AIBs. The orders are based on a forecast which is based on sales estimates plus likely a bit of cushion for overstocking and defects. With this additional mining boom it's not out of the realm of possibility that forecast was too modest and they've simply run out and why he's saying it feels like it's a "No Production" issue. Because in the end if Nvidia has the parts, they are going to produce.
 
Like I mentioned above, a buddy has had great success at Microcenter every Friday morning. However that could be Microcenter's warehouses having plenty of back stock to distribute to their stores and not actually receiving new production shipments from the manufacturers. The bigger question is if this is truly a "No Production" issue, what does Nvidia do? We know that Volta/Ampere is coming and if you believe some of Alienware's cryptic tweets about their new laptop revisions, prior evidence could suggest it might be as early as Q2. Is it possible to do a Pascal production run along side a Volta? If not do you interrupt and potentially create a delay in Volta to fill the Pascal void created by unforeseen circumstances when forecasting orders?
 
It needs to be said we are all really speculating and none of us really know anything. It's a weird situation and I cant ever remember seeing anything like this in the PC hobby.

The EVGA GTX 1080 Ti SC black Edition use to cost around $729-$750 USD retail for a few months after launch. Afterwards, gradually the price increased. Now a used version is worth over $1,000+ on ebay, lol

The limited availability of cards is amazing. Power supplies too. As someone who is building a new rig here in Q1, its a bit disappointing especially this long after the ti release.

I really hope that the next generation is able to do something that can help fix the supply and demand issue. It's gotten kind of crazy and no way I would pay the crazy prices third party sellers want.

Hoggle
I really hope that the next generation is able to do something that can help fix the supply and demand issue. It's gotten kind of crazy and no way I would pay the crazy prices third party sellers want.

Could if nvidia and amd both decide to block out mining on there cards.

Seen that.. a bunchbof 1080 cards sole for like $196 bucks ea.

PR Stunt.

AHowes

Hoggle
I really hope that the next generation is able to do something that can help fix the supply and demand issue. It's gotten kind of crazy and no way I would pay the crazy prices third party sellers want.

Could if nvidia and amd both decide to block out mining on there cards.

 The only way to do that would be to block GPU Compute entirely.
 NOT happening.
 
 As far as Nvidia allegedly asking it's dealers to stop selling to miners - not going to happen except perhaps on BULK sales, as there is no way for the dealer to tell a "miner" from a "small scale system builder".
 PR stunt no more - and Nvidia is a hypocrit to MAKE this claim given how much of their PROFITS for last year came from miners (more so for AMD but BOTH have benefited a lot this time around).
 

aka_STEVE_b
How exactly do you verify selling to only gamers......??? 
 
Sounds great in theory....

I'm thinking someone coming in and ordering fifty at once is most likely not going to be regarded as a gamer. 50-way SLI, anybody?
 

Why do so many gamers ASSUME that miners all are "running them at max load 100% of the time"?
To be pointedly blunt, GAMERS are far more likely to push their cards a LOT harder than miners are - you don't see "overclock competitions" out of MINERS, there's a reason for that.
ELECTRICITY COSTS MONEY, and burning electric for very little or no return HURTS PROFITS.
 
A lot of us mine for EFFICIENCY, which on NVidia 10xx cards generally mean running them in the 60-70% range - which is a LOT easier on them than running them at OVERCLOCKED TO THE MAX like many gamers do even for 2-3 hours a day.
AMD side, most folks undervolt and tune for good hashrate at lowest possible TDP - again, NOT EVEN IN THE BALLPARK of "running them at max load".
Even for those that DO push their cards some, it's generally to no more than the 90% TDP range because that last 10% almost always costs more than the return, and in SOME cases there IS no return.
There are also a lot of algorithms in use in mining where pushing the core clock HURTS, because the memory can't keep up (ETH is a PRIME example), and the general rule of thumb there is to UNDERCLOCK the card not "push it" - and the card tends to run WAY below rated TDP no matter what TDP setting you give it as a result.
 
Miners are almost always sensitive to card TEMPERATURE as well - and anyone that knows electronics knows that hot devices die faster than cool ones.
Gamers have a bad habit of sticking with the "stock fan profile" and let their cards get WAY too hot - and *poof* something overheats and lets out the magic smoke!
 
Gamers are also more likely to turn on and turn off their machines on at least a daily basis.
Turning off an electrical device is the second-worst "not technically abuse" thing you can do to an electrical device - following a ways behind turning the device on.
Think about how often you have a light bulb fail at any time OTHER THAN right when you turn it on - same type of issue affects anything electrical/electronic, the circuit gets more stressed when it's NOT in a stable condition, and a lot more likely to fail.
This is why a lot of companies STILL insist on leaving their computers on 24/7 even when not in use DESPITE the "conserve" and "green" movements - the cost of the electric to do so is often LESS than the cost of additional machine failures over course of a year.
 
For reference, the ONLY cards I have ever done an RMA on were INFANT MORTALITY cases that would have died in the first couple weeks no matter HOW they were used.
Any card I have ever mined on that made it through the first month I still own AND IS STILL MINING (or doing D.Net work, which is crypto work that is very similar to mining) - many of these specific cards are 5+ years old now and STILL GOING STRONG.
 
I would BET that far more cards have been returned for failing in the past year from GAMERS than have been returned from miners.
 
 
It would be interesting to compare EVGA products to Sapphire products, but since Sapphire is all Team Red while EVGA is all Team Green, it is rather difficult to make any close comparison.
IMO Sapphire does a better job of engineering overall than ANYONE else, and tends to do a better job on cooling on most of their cards vs anyone else (a few SPECIFIC models from some manufacturers can argue the point, like the Gigabyte Aorus line, bit not entire LINES from a manufacturer - even EVGA had those "burn up due to poor design choices" early 1070 models).
Like EVGA, Sapphire also ignores anything but BALL BEARING fans, so both can duke it out across the board on longevity and generally beat anyone else, again with specific LINE examples but no other manufacturer examples I am aware of.
 

One of my 708ti king pin cards finnsly called it quits after 6 years of abuse thru 5500 hours. Waited for stock many weeks and could not get an evga card. Ordered a titanxp jedi to hold me over till next gen cards, sli capable titans not the titan V. If the 1070ti is on avg the cost of a titan, i was like look wife this card is cheaper then a 1080ti at 1400$ on amazons. She was like ok get it. Many thanks to the miners and resellers jacking up prices.

ty_ger07

Viper453
1999 here seen nothing like this until now, my first real graphics card was a voodoo 3 3000 pair with an AMD k6-2 450 mhz.

Me too! I had the exact same combination of hardware in the first PC I built. They were connected to a FIC motherboard and I overclocked the CPU to 500 mhz.

I didn't have my k6-2 oc. I had a 5" quantum bigfoot drive 6.5 gb, it wasn't ata 33 or 66 capable drive either :(

I remember working with 2.5 GB Bigfoot drives, didn't realise they got all the way up to 6.5
 
Weirdest form factor drives I ever saw, didn't fit well into most cases as a result (lots of wasted drive bay space) though they worked well enough.