Riding the Crypto wave

[hat]

It does when I earn, for example, 1 mBTC per day, which might be worth $3 today, but $6 tomorrow. Of course those are completely out of my ass numbers, but it shows how the profitability changes. Or, if you prefer, maybe you'd rather I call it perceived profitability, since that's how I see it, but it might not be technically the correct way to see it. I see you throwing around concepts and terminology as if you studied some financial shit, and not knocking you for that, but I haven't, and what it boils down to for me is money is money.

 

[notb]

It does when I earn, for example, 1 mBTC per day, which might be worth $3 today, but $6 tomorrow. Of course those are completely out of my ass numbers, but it shows how the profitability changes. Or, if you prefer, maybe you'd rather I call it perceived profitability, since that's how I see it, but it might not be technically the correct way to see it.

You aren't mining BTC!
Your mining efficiency could be shown in BTC, but that's already recalculated using the relation between BTC and cryptocurrencies you mine (mostly ETH, I presume).

It's a simple concept. I can't believe you don't understand this after weeks of discussing in this thread... :-(
This just shows that more people should be forced to get a grasp of the "financial shit".

I see you throwing around concepts and terminology as if you studied some financial shit, and not knocking you for that,

I studied physics, but that's irrelevant here.
The interesting part is: why do you even consider "knocking me" for using financial terminology?

And to be honest, I'm trying to be pretty light on the financial stuff and math in these discussions. It's not like I'm attacking you with stochastic processes and partial differential equations...

but I haven't, and what it boils down to for me is money is money.

The big problem here is that you (like most miners) don't really understand what money is and how it works.
I assume that money for you is just cash in your pocket. A thing that you can give someone for goods.
To be honest, this is enough for 99% of people on Earth in their jobs and private lives. It's even enough for most people working in banking.
But it's not enough to analyze a complicated financial instrument.

 

[silkstone]

You aren't mining BTC!
Your mining efficiency could be shown in BTC, but that's already recalculated using the relation between BTC and cryptocurrencies you mine (mostly ETH, I presume).

It's a simple concept. I can't believe you don't understand this after weeks of discussing in this thread... :-(
This just shows that more people should be forced to get a grasp of the "financial shit".


I studied physics, but that's irrelevant here.
The interesting part is: why do you even consider "knocking me" for using financial terminology?

And to be honest, I'm trying to be pretty light on the financial stuff and math in these discussions. It's not like I'm attacking you with stochastic processes and partial differential equations...

The big problem here is that you (like most miners) don't really understand what money is and how it works.
I assume that money for you is just cash in your pocket. A thing that you can give someone for goods.
To be honest, this is enough for 99% of people on Earth in their jobs and private lives. It's even enough for most people working in banking.
But it's not enough to analyze a complicated financial instrument.

If the value of BTC changes in the time between BTC payout (from ETH) and USD payout, then the amount of money @hat makes is dependent on BTC price. Long term changes, less so. The ROI given in USD per day is soley dependent on the ETH price at the time, as you say.

@hat - You can imagine how it works ias follows: You mine ETH (or whatever) and are then paid the going rate for that ETH (be it $175 or $400/ETH) that is converted into BTC at the going rate for BTC. If you were to cash out the BTC, be it high or low, you would only get as many USD as your ETH gave you.

The middle step (ETH -> USD) doesn't actually occur, but the effect is the same.

 

[notb]

If the value of BTC changes in the time between BTC payout (from ETH) and USD payout, then the amount of money @hat makes is dependent on BTC price.

Of course. I mentioned that few posts back - if he's depositing his mining results in BTC, then the BTC price becomes important. But this is just an unnecessary speculation added to the equation.

However, you have little choice with services like Nicehash, because of the weekly payments in BTC. So even if you decide to get paid ASAP, you're already affected by few days of ETH/BTC fluctuations.
In fact Nicehash is clearly lying to its customers in this sentence:
"No exchange needed; get paid directly, your payment does not depend on exchange rate fluctuations."
[https://www.nicehash.com/?p=ethereum-mining]

 

[silkstone]

Of course. I mentioned that few posts back - if he's depositing his mining results in BTC, then the BTC price becomes important. But this is just an unnecessary speculation added to the equation.

However, you have little choice with services like Nicehash, because of the weekly payments in BTC. So even if you decide to get paid ASAP, you're already affected by few days of ETH/BTC fluctuations.
In fact Nicehash is clearly lying to its customers in this sentence:
"No exchange needed; get paid directly, your payment does not depend on exchange rate fluctuations."
[https://www.nicehash.com/?p=ethereum-mining]

As far as I Understand, the conversion from ETH to BTC is instantaneous, in that regard the payment to BTC does not depend on fluctuations.

However, as there is a minimum BTC withdrawal amount, any BTC to USD conversion would be subject to fluctuations.

 

[hat]

You aren't mining BTC!
Your mining efficiency could be shown in BTC, but that's already recalculated using the relation between BTC and cryptocurrencies you mine (mostly ETH, I presume).

It's a simple concept. I can't believe you don't understand this after weeks of discussing in this thread... :-(
This just shows that more people should be forced to get a grasp of the "financial shit".


I studied physics, but that's irrelevant here.
The interesting part is: why do you even consider "knocking me" for using financial terminology?

And to be honest, I'm trying to be pretty light on the financial stuff and math in these discussions. It's not like I'm attacking you with stochastic processes and partial differential equations...

The big problem here is that you (like most miners) don't really understand what money is and how it works.
I assume that money for you is just cash in your pocket. A thing that you can give someone for goods.
To be honest, this is enough for 99% of people on Earth in their jobs and private lives. It's even enough for most people working in banking.
But it's not enough to analyze a complicated financial instrument.

I may have used the term "mining bitcoin" before, but I didn't use it where you quoted me. I already understand I'm not mining bitcoin, but saying "mining bitcoin" is easier than saying "running a program which automatically mines one of many cryptocurrencies based on profitability at the time, which then pays me in bitcoin". I didn't consider knocking you, I only threw that in in case you were to get offended or something, since I mentioned your postings touching on some of the technical details behind the "financial shit", that being things like money, and risk, things you seem to think most people don't understand because most people don't hold a master's degree in said financial shit. I've admitted many times I don't understand everything about this, including the way the market actually works. Maybe I don't understand the definitions of risk or money you're talking about. What I do understand is this: I saw an opportunity to make money (there's money, that is, in my pocket) by mining bitcoin (okay, not directly mining bitcoin, mining other coins and getting paid in bitcoin). I also understand it's risky, primarily because I spent a good amount of money to get started (there's the risk, this whole thing could grind to a halt and I'll make pennies on the dollar), but also because my earnings can vary wildly based on the state of the market (there's more risk, the volatile nature of the market itself). So... I know what money is, I know what mining is, and I know what the risks are (to me, personally), so I feel I know enough to proceed.

If the value of BTC changes in the time between BTC payout (from ETH) and USD payout, then the amount of money @hat makes is dependent on BTC price. Long term changes, less so. The ROI given in USD per day is soley dependent on the ETH price at the time, as you say.

@hat - You can imagine how it works ias follows: You mine ETH (or whatever) and are then paid the going rate for that ETH (be it $175 or $400/ETH) that is converted into BTC at the going rate for BTC. If you were to cash out the BTC, be it high or low, you would only get as many USD as your ETH gave you.

The middle step (ETH -> USD) doesn't actually occur, but the effect is the same.

Yeah, I get that. That's why I say bitcoin price influences my earnings... because, well, it does. The amount of bitcoin I earn changes based on how other coins, which I mine, are doing, but because I get paid in bitcoin, when the price of bitcoin changes, even though I'm not actually mining bitcoin, that also affects my earnings. If everything stayed the same except bitcoin shot up to $10,000 I'd be making over 3x as much as I am now.

Of course. I mentioned that few posts back - if he's depositing his mining results in BTC, then the BTC price becomes important. But this is just an unnecessary speculation added to the equation.

However, you have little choice with services like Nicehash, because of the weekly payments in BTC. So even if you decide to get paid ASAP, you're already affected by few days of ETH/BTC fluctuations.
In fact Nicehash is clearly lying to its customers in this sentence:
"No exchange needed; get paid directly, your payment does not depend on exchange rate fluctuations."
[https://www.nicehash.com/?p=ethereum-mining]

See... how is that unnescessary? I get paid in BTC, so BTC price fluctuations directly affect how much money I earn from this. Again, wild example, but 1 BTC is $2734 right now, and if it becomes $10,000... that's important!

 

[silkstone]

I may have used the term "mining bitcoin" before, but I didn't use it where you quoted me. I already understand I'm not mining bitcoin, but saying "mining bitcoin" is easier than saying "running a program which automatically mines one of many cryptocurrencies based on profitability at the time, which then pays me in bitcoin". I didn't consider knocking you, I only threw that in in case you were to get offended or something, since I mentioned your postings touching on some of the technical details behind the "financial shit", that being things like money, and risk, things you seem to think most people don't understand because most people don't hold a master's degree in said financial shit. I've admitted many times I don't understand everything about this, including the way the market actually works. Maybe I don't understand the definitions of risk or money you're talking about. What I do understand is this: I saw an opportunity to make money (there's money, that is, in my pocket) by mining bitcoin (okay, not directly mining bitcoin, mining other coins and getting paid in bitcoin). I also understand it's risky, primarily because I spent a good amount of money to get started (there's the risk, this whole thing could grind to a halt and I'll make pennies on the dollar), but also because my earnings can vary wildly based on the state of the market (there's more risk, the volatile nature of the market itself). So... I know what money is, I know what mining is, and I know what the risks are (to me, personally), so I feel I know enough to proceed.



Yeah, I get that. That's why I say bitcoin price influences my earnings... because, well, it does. The amount of bitcoin I earn changes based on how other coins, which I mine, are doing, but because I get paid in bitcoin, when the price of bitcoin changes, even though I'm not actually mining bitcoin, that also affects my earnings. If everything stayed the same except bitcoin shot up to $10,000 I'd be making over 3x as much as I am now.



See... how is that unnescessary? I get paid in BTC, so BTC price fluctuations directly affect how much money I earn from this. Again, wild example, but 1 BTC is $2734 right now, and if it becomes $10,000... that's important!

Only retroactively.

If the BTC price shot up to $10,000 or 4x the price, Nice hash would pay you 4x fewer BTC, assuming the price of ETH {or whichever coin} remained constant

 

[notb]

As far as I Understand, the conversion from ETH to BTC is instantaneous, in that regard the payment to BTC does not depend on fluctuations.

However, as there is a minimum BTC withdrawal amount, any BTC to USD conversion would be subject to fluctuations.

Yes, this is exactly what I meant. The BTC inflow is instantaneous, so you're only affected by BTC/USD fluctuations afterwards. Still, you're getting a variable mining efficiency because of the ETH/BTC conversion.

 

[hat]

How does one use those USB risers? It seems that the graphics card would no longer fit where it normally goes, because the riser card would be hitting something (namely another slot)...

 

[verycharbroiled]

How does one use those USB risers? It seems that the graphics card would no longer fit where it normally goes, because the riser card would be hitting something (namely another slot)...

Some people hang the gpu in the air from something, some build frames, some set the gpu on the shelf or something behind the mobo.. basically the gpu can go anywhere within the riser and psu cables reach and they can get decent spacing between the cards.

 

[hat]

Seems tricky to install securely... I may end up getting another mobo when the time comes. AM3 processors, like the Athlon II series, will work in AM3+ boards, correct?

 

[P4-630]

Cryptocurrency miners are renting Boeing 747s to ship graphics cards

http://www.pcgamer.com/cryptocurrency-miners-are-renting-boeing-747s-to-ship-graphics-cards/

 

[notb]

Cryptocurrency miners are renting Boeing 747s to ship graphics cards

http://www.pcgamer.com/cryptocurrency-miners-are-renting-boeing-747s-to-ship-graphics-cards/

Now this is a funny story. But I won't laugh. I'll estimate.
They're mining Ethereum in Iceland. If we assume that:
- they ship cards from Hong Kong,
- shipping by air takes 15h,
- the cost of renting a 747-800F is $20k/hour,
- shipping by sea would take 30.6 days
the whole transport will cost $300k and save them 30 days.
That's $10k/day - more or less what 7k 1070s would make at current pricing.

Even assuming that they're shipping all the necessary stuff (not just GPUs), a 747-400F would still handle 500m^3 of graphic cards.
An MSI 1070 Gaming box is ~0.008m^3, which gives us 500*125 = 62500 graphic cards.

This is easily worth the fuss, even at current awful mining profitability - leaving huge margin error for the assumptions above.

EDIT: thanks @silkstone.

 

[silkstone]

Now this is a funny story. But I won't laugh. I'll estimate.
They're mining Ethereum in Iceland. If we assume that:
- they ship cards from Hong Kong,
- shipping by air takes 15h,
- the cost of renting a 747-800F is $20k/hour,
- shipping by sea would take 30.6 days
the whole transport will cost $300k and save them 30 days.
That's $30k/day - more or less what 20k 1070s would make at current pricing.

Even assuming that they're shipping all the necessary stuff (not just GPUs), a 747-400F would still handle 500m^3 of the graphic cards.
An MSI 1070 Gaming box is ~0.008m^3, which gives us 500*125 = 62500 graphic cards.

This is easily worth the fuss, even at current awful mining profitability - leaving huge margin error for the assumptions above.

$10k per day at 30 days

 

[hat]

Well, my rig burned up again. I'm not sure what happened this time, it didn't look any more damaged than before. All I did was brush the 24 pin male and female ends with a toothbrush and rubbing alcohol like I did before, but this time I tried using a shop vac, primarily on the female connection on the board, to try to suck up and blow out any loose crap that might be in there.

My two +12v leads on the male 24 pin end have looked better. They're burned up a bit and frayed a bit, it looks a little worse now than when it did when I took the picture back on page 47. I'm tempted to cut off the burned/frayed ends and try to pull a little slack, and stick some salvaged connectors from a sacrificial PCI-E connector or 24 pin connector or something on the fresh ends... but I can't say with 100% confidence that repair will be successful. I'd do it in a heartbeat if I had another power supply, which is first on my list of things to get. A nice top tier power supply. Then I can use this one (once it's repaired) in a secondary machine. Until then, I've used MSI Afterburner to set the power target to 80%. I was running 100% before, so this should shave off 60w or so... hopefully that 24 pin won't be so stressed now.

 

[infrared]

Hmmmm, get in touch with the psu mfg and see if they'll send you a new cable. There can't really be much wrong with the pin on the board other than being dirty, but the wire/crimp/connector can be bad. I'm sure if you show them a picture of the burned up 24pin connector they'll sort you out, or at the very least sell you a new one which hopefully would fix the problem. I'm guessing either the connector isn't a tight fit on the pin or the connector isn't crimped onto the wire well enough. Either way it sucks you're having issues :/

I've been mucking around with Nicehash today, my two gtx titans combined (original kepler 6GB cards) are getting £1.58/day (daggerhashimoto @35.2MH/s) and the 1080Ti @ 90% tdp is getting £2.20/day (EquiHash @ 635Sol/s). Neither really makes sense to keep running imo. That's a lot of money in electricity and stress on components.

 

[hat]

Well, the 24 pin cable isn't modular, so... a new cable won't do much. That's why I say I want to cut the ends and try to stick new connectors on.

Yeah, the Kepler cards aren't great at this. My 600 Ti did about 70 cents a day when profitability was good. The 1080Ti might be worth running... it's outperforming both those older Titans by itself, so...

For reference, my 1070s combined are doing 795 Sols/s in Equihash right now. That's with +350 vram and 80% power target.

 

[infrared]

Ah that makes sense, good idea, it's pretty easy to crimp up a new connector

Thanks for the reference, I might keep the 1080Ti going a bit longer then. It's only using 220W, the titans are using 400W for 72% of the output lol

 

[hat]

All of the gods dammit...

My rig abruptly shut down again, and wasn't coming back on. So I took the 24 pin back off and plugged it back in, nothing. I took it back off and brewed up a few ideas, most of which won't work.

My first thought was to sacrifice the end of a SATA power cable from an old modular power supply. The end that connects to the power supply is the same end on the 24 pin connector. So I cut one of the wires and the connector I was after wouldn't push or pull out, so I just murdered the plastic housing and took it out. These connectors look awfully small. I could possibly uncrimp one with a razor blade or something, but sticking a new wire in and re-crimping it... not sure. This idea is on hold because I'm unsure I could pull it off, and I don't want to risk destroying my halfway working 24 pin and not being able to put it back together.

Then I thought about cutting all the wires off another 24 pin, cutting all the wires on my current one, and just splicing them together, effectively making a new 24 pin connector... but that's a lot of damn spliced wires. And a lot of damn heat shrink tubing I don't have. I could go really ghetto with electrical tape, though. And I don't have a soldering iron, although I don't really think soldering the wires would be absolutely necessary. This seems like it could work, but it somehow seems less than ideal... wouldn't this also increase resistance due to all the spliced wires?

Which brings us to option three: frankenstein machine. Simply use both power supplies at the same time. Jump the existing 24 pin from the Corsair CX600 with a paper clip, and use the 24 pin from the second power supply. That way, the second power supply powers the board, and the CX600 powers everything else. I'm unsure of the potential consequences of running power to components that aren't running, though. The CX600 would have to be on to power the other components before I hit the switch to turn the computer on triggering the second power supply plugged into the board, and thus I assume some power would be being fed to everything else for a short time before the computer is actually running.

Of course, I could save up some money and just buy a proper power supply. But then, since the 24 pin female on the board is somewhat damaged, isn't that high potential for another problem causing high resistance connection? I can't really get in there to clean it up nice... all I can do is brush the connector on the board and hope for the best. I guess I would need a power supply AND board to really be good here... but a nice z68 or z77 board would be a great way to breathe new life into an aging system. I could clock up my CPU and RAM a bit. I've been stuck running 1600MHz RAM at 1333 for ages, because of the H67 chipset I use. Then again, option three frankenstein PC would be a good way to test this. If the 12 lines on the second PSU burn up as well, I know the board can't handle it.

Anyways, currently I'm right back where I started. The rig did boot up again for some reason, but it seems pretty obvious to me that it's hanging on by a thread. That weakened 24 pin connection is really a thorn in my side. I think the best option would be to just get new stuff and retire the PSU and board both as both connectors aren't the greatest, but I've got a way to go before I can afford that. The only way would be to use my mining profits, and I need to take care of other things before I can use mining profits on hardware.

 

[notb]

All of the gods dammit...
[cut]

Oh man... you've been having these problems for a while now and you're still tinkering and trying workarounds instead of just buying new parts.
Sooner or later this will end in a fireball, so either fix the PC or get a decent home insurance...

and I need to take care of other things before I can use mining profits on hardware.

Now that's where things become pretty worrying. The idea of some additional cash (e.g. to buy more gear) is fine - even with shrinking profitability - as long as ROI is more or less sensible.
The sentence above looks like you need mining profits to cover taxes, credits or something like that.
No offense, but you have no profits yet. Since you joined during this discussion and bought new GPUs, I understand that your ROI is still far away. And it will become even farther when you start replacing broken gear...

 

[infrared]

@hat You can buy those connectors new online, they cost virtually nothing, just google 24pin atx connector. It'll come with the fresh ends, you can't re-crimp them. You should really do it with proper crimpers to guarantee a good connection. (they're a useful tool to have, make any cable you want up with them!)

edit: Something like this would work nicely, you should be able to do the 3 and 4 pin fan connectors too which is pretty handy.
https://www.amazon.com/dp/B001GMYU9S/?tag=tec06d-20

 

[hat]

Oh man... you've been having these problems for a while now and you're still tinkering and trying workarounds instead of just buying new parts.
Sooner or later this will end in a fireball, so either fix the PC or get a decent home insurance...


Now that's where things become pretty worrying. The idea of some additional cash (e.g. to buy more gear) is fine - even with shrinking profitability - as long as ROI is more or less sensible.
The sentence above looks like you need mining profits to cover taxes, credits or something like that.
No offense, but you have no profits yet. Since you joined during this discussion and bought new GPUs, I understand that your ROI is still far away. And it will become even farther when you start replacing broken gear...

You're not wrong. I've mentioned before that financial struggle is nothing new to me... but lost time running the rig is lost money, and I don't have money to fix things properly, hence workarouds

@hat You can buy those connectors new online, they cost virtually nothing, just google 24pin atx connector. It'll come with the fresh ends, you can't re-crimp them. You should really do it with proper crimpers to guarantee a good connection. (they're a useful tool to have, make any cable you want up with them!)

Hm but I would still have to remove the wires from my existing 24 pin connector and re-crimp those two ends from the +12v leads right? I could potentially take off the female connector from the board and put a new one on, if I had a soldering iron, but that's a project for another day.

 

[infrared]

Nope, slightly different approach. I'm guessing only a short bit of wire up to the connector is melted? cut it just above there (actually do it one at a time to keep track) so each end is fresh wire. You strip the wire and put it straight in the new fresh metal fitting, crimp it and click it into the plug. If you take your time getting this right you'd end up with a factory quality job, the cable would just be 1-2" shorter.

Here's a useful video, it's using automotive connectors but it's the identical process

 

[hat]

Well, there's no way I can un-crimp the connector and re-crimp it. Which leads me to option 1a:

Get two wires out of one of those modular connectors. Leave the crimped end alone, but cut the wire. Cut the burned up +12v leads from my PSU's 24 pin connector. Splice those two wires. Fresh connectors... spliced wire. Spliced wire has got to be better than burned up wire w/ shitty connector huh?

 

[infrared]

Ah, so you're trying to hack it with a salvaged connector from the other one.. Okay yeah I guess that'd work, and would get you up and running again quickly. It's a bit 'animal-ish' though lol.
I was suggesting you could get a brand new connector to replace what's on there with. To be fair its a big job, you'd splicing would be faster.

There might be better options that come up for you, faster postage maybe..
http://www.ebay.com/itm/Shakmods-24...242803&hash=item27e1bc6929:g:45kAAOxyP4dTegKg

edit:

I may have read your post wrong, i see what you're saying now I think. Extend the (now shortented) wire.. Yep that should work. It's a lot less work that what I actually meant of cutting and re-crimping the whole lot.

time for bed lol. Peace!