Topic: rpietila Altcoin Observer - page 208. (Read 387552 times)

Show me the GPU miner your using at 1ks+ then please.

Anyone can use a gpu to mine monero with the claymore miner...its not closed source.

show me cpu running 1khs+ speeds?

Are you kidding....-facepalm-, please Read the claymore miner thread. High end CPU's are more effective than a GPU for monero at this point.

BTW: anyone can use the claymore miner.....not 1 person...its open to everyone lmao.

you are so late, no single cpu will over take gpu speeds anymore

ok, some high end servers upto 800hs, but gpus do that too

Hashrate went up because there are lot more bitcoiners who have GPU's from mining previous coins. But, the GPU miners for monero are not as good as a high end CPU.

Making assumptions about things you dont know about wont get you anywhere, it looks pretty foolish.

I'd say, buy a high end CPU with a lot of RAM, and mine Monero, you'll get way more coins than using the Claymore miner with a GPU. I've tried..

ec2 is expensive..
i can run now 2.6khs on my gpus, its 30x cheaper than ec2
this costs me 1000w, count that

and see that when gpu miner came out, hash went from 3mhs to 12 mhs

Are you trolling....

GPU doesnt earn more coins than a high end CPU with cryptonote coins. I've tried mining with a GPU using Claymore's miner and got LESS than I did with my CPU and EC2 instances.

Claymore miner is horrible...and doesnt earn anywhere near 5% lmao.



From the Claymore miner thread:

i hope risto p. did not hit his hands on full load of shit this time.
still time to sell.
seems like we are ending to monopoly xmr, one person collecting soon 5% of daily mined coins.
yep, and only because of closed source miner, claymore, he kills the coin.

You'd be surprised.  Miners show up in a *very* big way on EC2 spot when the profit margins are high.  I've confirmed this with Amazon.

I know personally of one miner who spent $200,000 USD in the last 30 days on EC2, and another person told me they were "in a similar range."  I believe there are about 4-5 total who do this.  I will have spent $22,000, and my bids went as high as USD$0.31/hour during the peak margin times.

I suspect that there was at least $1m USD spent on AWS for mining by the top 6 cloud miners during June.  I'm probably under-estimating the total.

Also, looking back at it - the graph I posted is very hard to read, because what really matters is the fluctuation at the bottom (from 0.27 to 0.31 as the floor), not the huge spikes.  But it's there.

Thats because contrary to what dga is suggesting I doubt xmr mining does  have a significant impact on ec2 prices, this is in addition to the fact that this graph is showing 1 out of 5 regions and also 1 of the several different instances ec2 miners are using. Furthermore a spike in ec2 price 2 days before the spike in XMR net doesn't make sense for this scenario.

I know nothing about EC2 but just an observation, the spike on this graph is almost 2 days prior to the biggest spike in XMR's price.  The network hashrate also went up at the time of the price spike not the the spike on your graph.

I presume dga is referring to me. 1200 figures are a best case scenario, Ive had some higher while playing with things such as huge pages etc but the variance you can get even between instances in the same region can be easily 10% + . Your right my c3.8 costs are a little high, couple of reasons for that;

1)Im in the EU and due to region instance limits I have/had a real mish mash of instances in various locations etc. The price of a c3.8 in Ireland atm is actually higher than .32 so to guarantee your instances wont be terminated by price you sometimes pay up to .4 an hour.

2) Its actually more profitable to run multiple smaller instances that cost a little more than one 3.8 but shush


Also to dga;

Have you overlaid that graph with the other 5 regions to see if your analysis stacks up across the board ? I suspect it wouldn't.

The EC2 prices are back down - I think the EC2 miners are getting edged out by the botnets and GPUs at this point.  (Reference:  There was a self-claimed botherder in #monero with about 500kh/s.  Not verified, but his numbers were sound.)

This is the Amazon spot price history for the last month for c3.8xlarge.  It's pretty clear when XMR mining was very profitable:



(The three colors are three different zones.  If only one spikes, it probably doesn't mean the overall price is high, but when all of them go high, it means that there's huge demand on EC2)

And then you can see that about a week ago, the spot prices returned back to near the "floor" (Amazon sets a minimum price for each instance type, even if there's no demand).  That correlated roughly with a large increase in diff and drop in price of XMR.

Wolf's miner is no longer the one to use as a comparison, btw -- yam is faster.  I havent benchmarked it on EC2, though, because the margin is too small (or negative) for my tastes.  1050 is probably a reasonable ballpark.

dga: Perhaps Wolf's figure of 1201 hps/c3.8xl was an outlier, then, if 1000 is typical.  But your spot instance rate seems a bit high, since the amazon site persistently reports in the .256 to .288 range.  Any light on that?

nakaone: I completely agree. 

I like to track the distribution of mining costs in order to model how each typical demographic will behave.  If you had the relative numbers of miners in each major bucket, and could estimate their sales timing, it would give you a substantial market timing edge.  We're getting pretty close to that point here, but not quite.  I could try to parameterize a model from the price data, but the signal is low enough and the dimension of the parameter space high enough so that current history length would be inadequate to provide statistically robust parameters.  The more dimension can be reduced, the sooner a valid parameter set can be derived.

 

maybe I am dead wrong here but I think the current price drop is not given by the inflation and the refinancing of miners but by the market getting rid of short term speculators (classical pump and dumpers). I simply follow the assumption that in general people mining this particular altcoin are better informed (higher and deeper technical knowledge) than the usual daytrader which are trying to ride the waves and looking for short term profits.


I followed monero from the very beginning (bought some preexchange over the counter) - we had this situation probably two times before - the first short term speculation free equilibrium can probably be found on cryptonote.exchange in the range of around 0.001, the second one maybe in the range of 0.002 after the initial hype ended when it hit poloniex, I think we have a third one here at around 0.004 after the end of the hype around mintpal.

for me it is no surprise that miners are mining this coin at a loss - I remember one coin this year where this was also the case.  

Various, but if you work it out based on a rough average of ec2 costs and HR.

So the large c3.8 instances are pushing average around 1000h/s with a cost of $.32/hour , so $7.68 per day OR 0.01151 BTC per day.

At current difficulty 1000h/s is around 1.45 XMR per day. This means 1.45 XMR needs to sell for 0.01151 to break even, giving a production price of 0.008 BTC.

I thought you're crazy.
You're craziest than I thought.

Im in.

what kind of instances are you using?

I definitely think free hardware or already ROI'd hardware account for a largish percentage of the network. But Im fairly confident, from what I know, that the large majority of the network is made up of ec2 miners. Most having already mined and sold enough on the spikes to mine at a loss at least temporarily.

From my calculations the current production cost on ec2 at a difficulty of 759066027 is 0.008 BTC per XMR which puts us well below production cost. I dont think i've ever seen clearer buy signals particularly as the net hash is rising again after a small dip.

I think free hardware will account for a significant amount of emission.  Essentially, scrypt mining went ASIC so GPU miners who have fully amortized have free, otherwise idle, hardware. If they are domiciled in Kuwait or Norway they have free power too.  However, they won' t add hash nearly as quickly as cloud miners, so that when network rates increase rapidly, these cheap coins will be a decreasing fraction, and the marginal production cost will constitute a floor.  If demand is sufficient to sink all the cheap coins, the price tends to rise easily to the marginal production cost.  Rises past that depend on positive pressure from excess demand, and will tend to drop to the marginal supply cost easily.

Increasing demand is good for everyone.  Miners, holders, and buyers all benefit.  When the GUI wallet arrives, the level of accessibility will be comparable to BTC, and the available sources of demand will be much more adequate to sink the supply.  Mining will become profitable for marginal miners, and difficulty will rise, creating positive feedback on price.  Marginal miners will scale up easily, but low-cost miners will not, so the price floor created by marginal supply will get harder.

Meanwhile, I am happy to accumulate cheap coins at the best prices I am likely to see, ever.

Observation: At current rates, the 290x pays for itself in a year of operation, give or take, with 10 cent kwh.  Startup GPU mining is probably sustainable at 0045, barely.  whatmine.com lists XMR as the highest return cryptonight coin, and indicates a better return than any scrypt coin.  Expect, therefore, hashrate to climb about 2x on GPU mining, which will take the marginal cloud miner cost to .01, and the startup GPU mining level (1 year payback) to .009.