As Prices Soar, Graphics Card Manufacturers Appeal To Cryptocurrency Miners

[USAFRet]

For all you wanna be home miners, this is what you're competing against:

 

[TJ Hooker]

@TJ why is eth designed to be resitant to asics?

OK, I'm no expert, but here's a sort of hand-wavy explanation based on what I've read:

Bitcoin hashing was primarily compute-bound. This made it open to HW acceleration, due to the fact that circuits could be designed specifically to do those specific kind of computations, faster/more efficiently than a general purpose CPU/GPU.
Eth is memory-bound, meaning performance is limited by memory bandwidth (and size to a lesser extent I believe). Apparently the memory subsystems found in GPUs are already pretty efficient (in terms of perf/doller and/or perf/watt), meaning there aren't opportunities for huge performance improvements with an ASIC.

Again, this is all just based on a rudimentary understanding from a bit of googling I did.

Edit: Derp, you asked why, not how. Oh well, I'll just leave this here anyway in case someone finds it interesting.

 

[turkey3_scratch]

Mining is never something I'll do. I would never spend real money on GPUs to make money in a different currency that can only be spent at 1% of places real money can be spent at.

 

[Dugimodo]

That's not what you do, you trade whatever coin you make for bitcoin which you can then sell on the markets for actual cash. I cashed in all my lightcoins weekly when I was mining. Some people keep them as an investment, always seemed way too risky to me.

 

[Matthew Renna]

I wanted to buy a new gpu to replace my 750 ti, I remember I saw a msi rx 480 8gb for 200 dollars early in may. I wish I had gotten that because it looks like I'm not getting a new gpu for a long time. It's not necessary that I get a new gpu because I don't game that much, but it would of been nice. I sure hope the prices go down soon.

 

[TJ Hooker]

For all you wanna be home miners, this is what you're competing against:

Spoiler

Eh, not really. Any amateur miner is going to be part of a mining pool, so it's not that you're pitting your home rig on its own against mining farms.

 

[The Paladin]

When will the prices bottom out for the 1060? The 760 gtx card I purchased last year from ebay recently died and am using a 460 gtx 768 MB card that is showing its age.

grab a 970 or 980TI for less and same power for gaming.

 

[The Paladin]

For all you wanna be home miners, this is what you're competing against:

man someone needs to get his farm out of the barn seriously nasty in there.,.. eww.

 

[USAFRet]

For all you wanna be home miners, this is what you're competing against:

Spoiler

Eh, not really. Any amateur miner is going to be part of a mining pool, so it's not that you're pitting your home rig on its own against mining farms.

Tell that to all the jokers in here wanting to start mining at home as a single operator.

 

[Dugimodo]

In terms of profit you are pitting your home rig against these farms, the more hashing power in the network the quicker the difficulty increases and the less profit you make. It's the large operations that quickly push mining out of the reach of a home rig with 1-6 cards and make it pointless.

 

[melvis72]

The other fun thing is prices are jumping day to day even though the stock is not there.

 

[crystaldragon141]

@PROBLEMATIQ Part of why this became an issue again is that the algorithm for Etherium is intentionally designed to be difficult to process with an ASIC due to the high memory bandwidth requirements.

 

[USAFRet]

When bitcoin was new, GPU was the thing. Then along came the ASIC rigs. That wasn't even on anyones radar. Poof...GPU mining was out the window.

Eth mining attempts to do an end run on the ASIC thing, by using different capabilities of a standard PC.
Some other technology or application will quickly do the same to Eth mining as ASIC did to bitcoin mining.

What might that be?
If I knew...I'd be rich, selling hardware.

 

[USAFRet]

The real question is: how cheap is their power? ...because that potentially sets a price floor.

And we can pretty much guarantee that their kwh cost is not retail, like you and I pay.
Anyone with even a medium size farm like that is not paying retail.

 

[bit_user]

Of course you could design an ASIC that's faster and/or more power-efficient. The key questions are how big would it have to be and on what process node would it have to be fabbed? And once you made it, would the price differential be enough to turn a decent profit?

GPUs fundamentally solve two problems:
1. Power-efficient programmable computation.
2. Hiding memory latency.

To accomplish #2, they employ SMT on a massive scale, with the biggest GPUs simultaneously juggling tens of thousands of threads and all their associated state. That's a big job, and it underlies much of GPUs' complexity. It takes hundreds of specialized engineers multiple years to design, verify, and fabricate these chips.

Again, what would it take to make an Etherium mining ASIC? I really can't say, but it might not have been obvious they could do it profitably, especially if there's a looming cutoff point where compute speed is no longer prized.

 

[USAFRet]

Right.
Either some new tech will come along (ASIC or whatever) that will blow away the current GPU thing...or it will simply become non-profitable due to market conditions.

 

[TJ Hooker]

In terms of profit you are pitting your home rig against these farms, the more hashing power in the network the quicker the difficulty increases and the less profit you make. It's the large operations that quickly push mining out of the reach of a home rig with 1-6 cards and make it pointless.

Difficulty goes up for everyone, which means that everyone (even farms) would earn currency more slowly. The thing is the value of the currency tends to go up as well. As long as the increase in difficulty doesn't significantly outpace the increase in value, even individual miners can still make a profit (assuming they're part of a pool).

 

[USAFRet]

Just like the regular stock market.
You have to get in at the right time, and you have to get out at the right time.

For Eth....getting in was probably 6 months ago.

 

[Evan_83]

The problem with Bitcoin, and the other upstarts is that eventually the market will flood with this type of currency.

If the digital age has taught us anything, its that any data and ways of producing it can be copied, again, and again....

The concept of crypt o-currency is not unique to bit coin or any digital currency enterprise. Bitcoin and its likes will continue to rise and crash hard probably a half dozen times over the next decade due to hype and people desperate to invest in something they can retire on.

eventually the world will be flooded with these types of company's to the point each one will be barely sustainable, it will be a part of future economy but will likely be the lowest denomination of currency due to its saturation and the realization that it is also vulnerable to many of the same problems fiat currency has despite currently popular belief, which is based on hope and not reality.

These types of currencies will be relegated to non essential services like rewards programs many corporations offer now, arcade ticket vendors and carnivals games.

 

[Evan_83]

The problem with Bitcoin, and the other upstarts is that eventually the market will flood with this type of currency.

If the digital age has taught us anything, its that any data and ways of producing it can be copied, again, and again....

The concept of crypt o-currency is not unique to bit coin or any digital currency enterprise. Bitcoin and its likes will continue to rise and crash hard probably a half dozen times over the next decade due to hype and people desperate to invest in something they can retire on.

eventually the world will be flooded with these types of company's to the point each one will be barely sustainable, it will be a part of future economy but will likely be the lowest denomination of currency due to its saturation and the realization that it is also vulnerable to many of the same problems fiat currency has despite currently popular belief, which is based on hope and not reality.

These types of currencies will be relegated to non essential services like rewards programs many corporations offer now, arcade ticket vendors and carnivals games.

 

[bit_user]

The problem with Bitcoin, and the other upstarts is that eventually the market will flood with this type of currency.

Why is the US Dollar (still) the dominant currency in the world? Why is ebay the dominant online auction site? And why is Facebook the dominant social network? Because they all act as platforms which derive their value based on the number of others that you can reach.

There have been plenty of other social networks and plenty of other online auction sites. But a major reason they can't displace the big guys is that they paradoxically can't achieve comparable value until they reach comparable scale. So, there's a huge practical barrier to entry, even if the mechanics of copying an ebay or a facebook aren't necessarily that big.

If a cryptocurrency ever emerges as a currency (i.e. rather than primarily as an investing asset), then it could reach the dominant position of something like the US Dollar, where it's incredibly hard to displace just because so many people accept it and so many things are priced in terms of it.

I thought Bitcoin would reach this phase, but a couple of the core maintainers basically staged a coup and steadfastly refuse to increase the protocol packet size, which is needed for it to scale to the level of an actual currency. They've successfully managed to shut down any attempts to fork the currency or otherwise defy their will. So, I think it might eventually die off. We'll see.

I still believe in the idea of crypto-currencies. It'll happen, one day. There's too much demand and too many people trying to do it. Somebody will get it right, and reach that critical mass first. There might also be a second and maybe even a third player. But the bigger those top couple cryptocurrencies get, the harder they'll be to topple.

Probably the major risk they face that doesn't apply so much to conventional currencies is that of an exploit. Possibly involving quantum computing, I wonder...

 

[falchard]

The R9 Fury cards are also marked up. Just did a quick test on the amount of mhs I can do on a Fury X and it is running at a nice 30 mh/s. Even the little Nano is running around $700.

 

[calken]

To all those who who site ASIC as the natural progression need to understand the sales aspect. As soon as an ASIC solution is released, the miners will jump ship, choose a different currency and Eth will crash. The ASIC will then be unable to fulfill the orders to cover the expense of creating the ASIC in the volumes required to make it competitive against GPUs from a ROI perspective.

 

[ap3x]

I see allot of people talking about profitability and return on investment in the hardware. As a former GPU coin miner, I can tell you it is nowhere near as simple as returning the cost of the hardware. Electricity usage is a major concern for GPU minors and in many cases, the cost of power outpaces the coins mined. That is the real reason why ASICS are so popular. The price of Asic Kit + Power vs Mining productivity is significantly lower even though the kit is higher priced. The electricity eats into profitability throughout the entire lifespan of hardware used. It is continual operational costs that add up vs static cost that does not move on the hardware side.

 

[problematiq]

@PROBLEMATIQ Part of why this became an issue again is that the algorithm for Etherium is intentionally designed to be difficult to process with an ASIC due to the high memory bandwidth requirements.

That's not an issue. People keep comparing mining eth with bitcoin ASICs. The purpose of an ASIC is that you design it to best do the function you want it to do without needing to dedicate hardware to the unknown.