ANN: BITMAIN has Tested Its 28nm Bitcoin Mining Chip BM1382

[cesarm]

They are still doing the machines ? Bad news.

I would much prefer if they were mining with the machines... receive a non-functioning machine is a much bigger issue than receive a machine that was used for 10 days... there will be no "burn-in" period ?

[aztecminer]

Anyone know how loud the S3 is and what kind of heat it puts out? I mean it's only 366 watts at the wall, right? I'm just wondering if those fans are loud or not and it is spits out a lot of heat. Do you need to have air conditioning on or this unit is able to cool itself with its fans? I'm assuming those fans are in a push/pull configuration, just pushing the air through.

And do you think one RM1000 power supply can power 2 of these? I use the Corsair RM1000 on my KNC and it has been a very good power supply with plenty of connections.

i have an entire fleet of rm1000's powering my antfarm colony without a single problem. they are great power supplies will run two ants oced... who would know the details of the S3 that hasn't shipped yet ?? i think if is comparable to the S1 that it will be able to withstand heat. we had people running entire farms of S1's in warehouses that were not well cooled. however heat kills electronics. it would be 'best practice' to run these in air conditioning. i keep my antfarm colony well cooled in a 70 - 72 degree datacenter. my S1's are ranging from 39 degrees to 45 degrees overclocked. however, many datacenters will run temps at around 76 degrees (which i think is really about saving money).

[elduderino]

Anyone know how loud the S3 is and what kind of heat it puts out? I mean it's only 366 watts at the wall, right? I'm just wondering if those fans are loud or not and it is spits out a lot of heat. Do you need to have air conditioning on or this unit is able to cool itself with its fans? I'm assuming those fans are in a push/pull configuration, just pushing the air through.

And do you think one RM1000 power supply can power 2 of these? I use the Corsair RM1000 on my KNC and it has been a very good power supply with plenty of connections.

It's a similar power envelope to the S1, so it should work.

Not sure, but the RM1000 (assuming it's 1000 Watts) should be able to handle 2 at standard clocks.

Overclocking, who knows?

[Mudbankkeith]

I have 14 Kw in my home that is ear marked for S3... now I just need some BTC to purchase with :-)

I just need Bitmain to to stop mining with my miners and ship it! . If I get a used miner I'm temped to send it back to them.  
It's like buying a new car that has 10,000 miles on it.

FAQ: Are you mining with S3 now?

Yes, we have been mining with one S3 testing sample on our engineer's desk. But we have not been able to do a large mining farm with S3 yet. We have been using S2 machine to deploy a mining farm, and that is how we own the hash rate.

The bulk PCB will be arriving tonight, and we will start first batch production tomorrow. First batch usually takes a longer time than later. We may ship the S3 -B1 on or before 14th.

I hope UPS can handle the extra load!

[dropt]

They are still doing the machines ? Bad news.

I would much prefer if they were mining with the machines... receive a non-functioning machine is a much bigger issue than receive a machine that was used for 10 days... there will be no "burn-in" period ?

So you would much rather them mine for themselves during the most profitable period of time the S3 is likely to ever see instead of giving it to you so you can take home those earnings?

You don't need to mine with HW for 10 days to know it's good to go.  When you buy a car, does it have 10,000KM (or miles) on it in lieu of "factory testing"?  Of course not.

[dropt]

Some of you guys are ridiculous. 

So Bitmain simply shipping out x PH/s of the current network is somehow worse for the expected roi of the s3 rather than adding x PH/s to the network? One option will increase the hash power directly and the other will not, I know which one I would have preferred.

What are you trying to say?  Your post that I quoted implied that you wish(ed) Bitmain jacked the hashrate mining for themselves before shipping the units.  IMO it'd be better to have them NOT jack the hashrate and the early orders get the benefit of a lower difficulty.  In this scenario the last guy to get his unit has the same difficulty he would have had if BITMAIN mined with all of the units simultaneously before shipping, however miners before that last guy gets the benefit of a lower difficulty.

[Threader]

the more things change the more they stay the same back in octocber I needed to spend 21K USD ~1.2 TH to get 1.1 btc per day.
Even with the Antminers awesome low price I would now need 70 Antminers ~35TH to get 1BTC/day. about 31K USD plus PSU's

So much for the at home miner to make 1 BTC a day    I give up.

Once my hive of S1s stop making more $$$ than power costs after reducing their wattage. I'm out.

[acidmilk]

I think alot of people here dont undersand the fact that alot of us assumed they allready had s3s on the network. Meaning when they shipped there wouldnt be additional hashpower added just reallocated. This means less roi

[spazzdla]

I think alot of people here dont undersand the fact that alot of us assumed they allready had s3s on the network. Meaning when they shipped there wouldnt be additional hashpower added just reallocated. This means less roi

I did assume this.

[el_rlee]

If that's their only S3 payment address (which I'm not sure about) https://blockchain.info/address/1QB8Ds5KbGYBLQa5RyDQ2sVUeSKWf7qgkZ this would mean ~50k S3's will hit the network in the next couple of days, that's ~24PH or roughly 19% of the current network if I'm not mistaken.

Distributed Hashpower I might add. Thanks Bitmain for selling to the small man's operations.

[dropt]

If that's their only S3 payment address (which I'm not sure about) https://blockchain.info/address/1QB8Ds5KbGYBLQa5RyDQ2sVUeSKWf7qgkZ this would mean ~50k S3's will hit the network in the next couple of days, that's ~24PH or roughly 19% of the current network if I'm not mistaken.

Before you run away, check the address you posted.  Go to the last page and look at the date of the first transaction. 

[cesarm]

They are still doing the machines ? Bad news.

I would much prefer if they were mining with the machines... receive a non-functioning machine is a much bigger issue than receive a machine that was used for 10 days... there will be no "burn-in" period ?

So you would much rather them mine for themselves during the most profitable period of time the S3 is likely to ever see instead of giving it to you so you can take home those earnings?

You don't need to mine with HW for 10 days to know it's good to go.  When you buy a car, does it have 10,000KM (or miles) on it in lieu of "factory testing"?  Of course not.

You didn't considered the presupposition that the machines will only be sent after 10th of July.

It's better to receive a working and fully tested machine on 15 of July than receive a new not fully tested machine on 15 of July.

The question is not if it's better to have a new not fully tested machine on 5 of July versus having a fully tested machine on 15 of July.

[drmadison]

The other thing people need to realize is that as the total hashrate goes up, the impact new releases have is lessened...

Let's say 10 PH is added to the network over the next month. That used to be huge, now that's a less than 10% bump over the course of a month.
The days of 30% difficulty increases is dwindling, as the amount of hashrate needed to make up that 30% goes up.

There's a reason the difficulty calculators are pointing to only a ~5% difficulty increase this next bump, and with the new hardware coming following that we'll probably be looking around 10% after that.

And for someone saying that it takes $35k to make 1BTC a day...
Complaining that you have to spend 35k to bring in 20k in revenue in the first month (yes I realize there are additional costs...cooling, power, etc.) is just foolish. Especially factoring in the break-even period of ~3 months. How many businesses can you get going for under $50k that will start turning a profit 3 months in? Not many...

[Collider]

If that's their only S3 payment address (which I'm not sure about) https://blockchain.info/address/1QB8Ds5KbGYBLQa5RyDQ2sVUeSKWf7qgkZ this would mean ~50k S3's will hit the network in the next couple of days, that's ~24PH or roughly 19% of the current network if I'm not mistaken.

Distributed Hashpower I might add. Thanks Bitmain for selling to the small man's operations.

Your napkin maths is seriously flawed. It would be around 10PH IF all the money came from S3 sales, which it didn´t.

There was a considerable amount of funds in that wallet before S3 sales opend.

[el_rlee]

If that's their only S3 payment address (which I'm not sure about) https://blockchain.info/address/1QB8Ds5KbGYBLQa5RyDQ2sVUeSKWf7qgkZ this would mean ~50k S3's will hit the network in the next couple of days, that's ~24PH or roughly 19% of the current network if I'm not mistaken.

Distributed Hashpower I might add. Thanks Bitmain for selling to the small man's operations.

Your napkin maths is seriously flawed. It would be around 10PH IF all the money came from S3 sales, which it didn´t.

There was a considerable amount of funds in that wallet before S3 sales opend.

there was ~35k BTC coming to the wallet, divided by miner price ~0.75BTC makes ~50k miners which have half TH each, makes ~~25PH, no?

But of course it might be that there has been money on that address from sales of other miners...

[Collider]

If that's their only S3 payment address (which I'm not sure about) https://blockchain.info/address/1QB8Ds5KbGYBLQa5RyDQ2sVUeSKWf7qgkZ this would mean ~50k S3's will hit the network in the next couple of days, that's ~24PH or roughly 19% of the current network if I'm not mistaken.

Distributed Hashpower I might add. Thanks Bitmain for selling to the small man's operations.

Your napkin maths is seriously flawed. It would be around 10PH IF all the money came from S3 sales, which it didn´t.

There was a considerable amount of funds in that wallet before S3 sales opend.

there was ~35k BTC coming to the wallet, divided by miner price ~0.75BTC makes ~50k miners which have half TH each, makes ~~25PH, no?

But of course it might be that there has been money on that address from sales of other miners...

Are you blind? There is less than 17K bitcoins in that wallet, and not even all of these came from S3 sales.

[el_rlee]

If that's their only S3 payment address (which I'm not sure about) https://blockchain.info/address/1QB8Ds5KbGYBLQa5RyDQ2sVUeSKWf7qgkZ this would mean ~50k S3's will hit the network in the next couple of days, that's ~24PH or roughly 19% of the current network if I'm not mistaken.

Distributed Hashpower I might add. Thanks Bitmain for selling to the small man's operations.

Your napkin maths is seriously flawed. It would be around 10PH IF all the money came from S3 sales, which it didn´t.

There was a considerable amount of funds in that wallet before S3 sales opend.

there was ~35k BTC coming to the wallet, divided by miner price ~0.75BTC makes ~50k miners which have half TH each, makes ~~25PH, no?

But of course it might be that there has been money on that address from sales of other miners...

Are you blind? There is less than 17K bitcoins in that wallet, and not even all of these came from S3 sales.

Can you read? I said it might be that there is money sent to that address for buying other miners in the past.
Why don't you do something meaningful and add up the S3 sales?

[Collider]

Yes, it might be that bitmain has sold antminers before the S3 and the funds went to their address 

[dropt]

Can you read? I said it might be that there is money sent to that address for buying other miners in the past.
Why don't you do something meaningful and add up the S3 sales?

And it's absolutely not a "might".  You missed my previous post, go to the last page and check the dates of those first transactions. 

[dropt]

You didn't considered the presupposition that the machines will only be sent after 10th of July.

It's better to receive a working and fully tested machine on 15 of July than receive a new not fully tested machine on 15 of July.

The question is not if it's better to have a new not fully tested machine on 5 of July versus having a fully tested machine on 15 of July.

The date of shipping doesn't matter in the context of the conversation, and the reason for the shipping date has already been attributed to BITMAIN not having production PCBs in house until tonight.  Further, what exactly constitutes "working and fully tested"?  Your proposition is that BITMAIN should hold your unit for 10 days to burn in test because then you're likely to receive a good working unit.  My argument is that burning in for 10 days vs 1 day does not greatly decrease the probability that you receive bum HW.  Wasting 9 days burning in the machine costs you, the end-user, hugely.  If you don't agree then how about BITMAIN sends me your barely tested HW and I'll burn-test it for 10 days before sending it off to you.