Pages:
Author

Topic: Mining Equipment Manufacturers - page 18. (Read 55714 times)

full member
Activity: 196
Merit: 100
June 13, 2013, 12:41:01 PM
i dont think it will grow THAT fast. GPU miners will stop at some point, where mining makes no sense to them and no new GPU miners will join. next FPGA miners will stop. Also with a incredible difficulty the ROI of ASICs will shrink and people will not hurry into them.
legendary
Activity: 826
Merit: 1004
June 13, 2013, 12:25:27 PM
We have started just a nice discussion here.
I have been thinking about that model used here - constant network growth on each retarget. I don't think it's possible even for a year.
I created a chart with data from this year and +12% on each reatrget up to december 2014. That's exponential growth up to 26,000 Thash by the end of next year (dec 2014)! Just make a chart with all that data and you will see that sharp exponential growth.
I don't see how all that hash power can be added in a year. I see something like linear growth up to 3,000 or maybe 5,000 Thash in that time.

EDIT: uploaded a (stacked) chart https://www.dropbox.com/s/5o66n1h13eccdmb/hashrate12.png
EDIT2: and another speculation, 3PH next year: https://www.dropbox.com/s/rage8ununtfowom/hashrate-3P.png

I pointed this out to some person in a different thread. I think he was using a a fixed 30% increase, and one of the increases he had calculated meant adding 2 million Th/s.

Basically, there's a maximum hash rate that each company can bring online per time period and the sum of those values corresponds to the maximum change in difficulty.
sr. member
Activity: 308
Merit: 250
decentralizedhashing.com
June 13, 2013, 09:19:05 AM
I think your Terrahash prices are way off. They're about 10 times more expensive than what it says in your chart on the first page...
I guess they changed prices.  Is the ellipses to show your disdain for such an error?  lol
full member
Activity: 532
Merit: 100
June 13, 2013, 02:37:33 AM
I think your Terrahash prices are way off. They're about 10 times more expensive than what it says in your chart on the first page...
sr. member
Activity: 308
Merit: 250
decentralizedhashing.com
June 12, 2013, 07:16:52 PM
Maybe you should change BFL 5 gh/s to 'YES' on shipping. They have shipped some 1k units in the last two weeks. If Avalon is shipping so is BFL.
They might ship some product, but are they currently shipping anything that you can buy in their store?  Being that this is about lead time I have to take into account only the miners that are ordered today.
full member
Activity: 224
Merit: 100
June 12, 2013, 04:29:03 PM
Maybe you should change BFL 5 gh/s to 'YES' on shipping. They have shipped some 1k units in the last two weeks. If Avalon is shipping so is BFL.
sr. member
Activity: 308
Merit: 250
decentralizedhashing.com
June 12, 2013, 03:56:12 PM
Yifu talked about a lot in the Bitcoin 2013 video. From deliberate sabotage from shipping companies,, pricing, USB controllers that can't handle certain volts, procuring certain other parts...

I'm not sure why a shipping company would do that.  Wouldn't it hurt their reputation?  What incentive is there?
Pricing?
USB controllers and other parts aren't needed to send out chips, unless I'm misunderstanding you.

...
I'm not sure, obviously, what the future holds, but I'm fairly certain it is neither linear nor exponential.
I agree with that. Exp growth is absolutely not sustainable.

It could hold until the next block reward halving. That doesn't seem too infeasible at the moment. However I'm thinking growth will cool off to 5% as the ASIC fevel cools down. A few companies completely failing to deliver will achieve that, by cooling down the demand.

Another variable to add to the scenarios is expected delivery dates. You cannot count amortisation since day 1 when most companies will take a very long time to deliver, except ASICMiner.
I agree that the difficulty will continue to have some crazy slopes and rises.  The average difficulty of all total rises has been about 17%.  It's handy to say 17%, but it discounts all of the major changes that have happened.  Basically this is what I'm doing with the 1 year estimate.  There are situations where a 12% average may end up actually giving the ROI of a 15%, if the growth is weighted toward the beginning of the cycle.  That's why I added the higher percentage columns, it looks like we will have a fairly low growth cycle compared to current estimates, but it could be heavily weighted toward the near future.
.m.
sr. member
Activity: 280
Merit: 260
June 12, 2013, 03:18:20 PM
I would support a guess, that we will see 1 maybe 2 Ph/s in the end of September.
see here - feel free to amend :

https://bitcointalksearch.org/topic/m.2452732

full member
Activity: 224
Merit: 100
June 12, 2013, 10:08:36 AM
Quote
While I don't believe anyone has heard from Avalon regarding the matter, I don't think we can assume that the delays in batch 2 & 3 means that the chips will be delayed by that amount.


Sample CHips have shipped.


Quote
So much to respond to here!  I didn't know that the manufacturing wasn't the problem.  But if this is the case what is holding them back?  Is there a link to anyone from the company talking about this, or other proof that they have the chips made.. just can't figure out how to get the home pick-up for UPS to work? 


Yifu talked about a lot in the Bitcoin 2013 video. From deliberate sabotage from shipping companies,, pricing, USB controllers that can't handle certain volts, procuring certain other parts...
donator
Activity: 980
Merit: 1000
June 12, 2013, 05:57:05 AM
...
I'm not sure, obviously, what the future holds, but I'm fairly certain it is neither linear nor exponential.
I agree with that. Exp growth is absolutely not sustainable.

It could hold until the next block reward halving. That doesn't seem too infeasible at the moment. However I'm thinking growth will cool off to 5% as the ASIC fevel cools down. A few companies completely failing to deliver will achieve that, by cooling down the demand.

Another variable to add to the scenarios is expected delivery dates. You cannot count amortisation since day 1 when most companies will take a very long time to deliver, except ASICMiner.
sr. member
Activity: 308
Merit: 250
decentralizedhashing.com
June 11, 2013, 05:55:55 PM

Thanks, I'm trying to help people figure out what's going on so we can have more decentralized hashing Smiley

Looking deeper into your doc (and if you are not using anything different) I'd suggest a quick script for Google Spreadsheet that can be useful for BTC exchange rate.

On one of my recent works I've added the script "MtGox Bitcoin Last Value" from the "tools" dropdown menu, then simply adding the function

Code:
=getBitcoinLastValueFromMtGox()

I got automatically the price in dollars of 1 BTC Smiley
That's very interesting!  I wonder if it works for other currencies.

So much to respond to here!  I didn't know that the manufacturing wasn't the problem.  But if this is the case what is holding them back?  Is there a link to anyone from the company talking about this, or other proof that they have the chips made.. just can't figure out how to get the home pick-up for UPS to work?  Smiley

Wish I had time to do a comprehensive check, I'll get back to this ASAP.
hero member
Activity: 588
Merit: 500
June 11, 2013, 01:56:41 PM

Thanks, I'm trying to help people figure out what's going on so we can have more decentralized hashing Smiley

Looking deeper into your doc (and if you are not using anything different) I'd suggest a quick script for Google Spreadsheet that can be useful for BTC exchange rate.

On one of my recent works I've added the script "MtGox Bitcoin Last Value" from the "tools" dropdown menu, then simply adding the function

Code:
=getBitcoinLastValueFromMtGox()

I got automatically the price in dollars of 1 BTC Smiley
sr. member
Activity: 302
Merit: 250
June 11, 2013, 09:52:19 AM
Quote
They were moved back because from the information I have found, it seems quite unreasonable to believe that 9-10 weeks is but a remote possibility for the people unlucky enough to not have made an order yet.  Though I really should go change the lead time for companies that are selling their own chips in miners...  That could happen even before the 9-10 week said lead time..

For the chips, they are within one week of the designated shipping time of sample chips. You will find the chips aren't the problem in their production right now...
^^^^ THIS ^^^^^

While I don't believe anyone has heard from Avalon regarding the matter, I don't think we can assume that the delays in batch 2 & 3 means that the chips will be delayed by that amount. There are production problems with the batches (as Avalon has stated) that I highly doubt relate to the chips. So then the question is, will Avalon intentionally hold back delivery of 61 batches of chips just because the miner batches 2 & 3 are delayed? Based on the chip sample deliveries (which were just a bit late and not months late), I speculate the answer is No. They will ship the chips relatively soon.

Now that might be wishful thinking on my part (I'm waiting on chips), but I think that argument is just as valid as the one (if not more so) you've used to bump Avalon chip miner dates by several months.
full member
Activity: 224
Merit: 100
June 11, 2013, 05:07:15 AM
Quote
They were moved back because from the information I have found, it seems quite unreasonable to believe that 9-10 weeks is but a remote possibility for the people unlucky enough to not have made an order yet.  Though I really should go change the lead time for companies that are selling their own chips in miners...  That could happen even before the 9-10 week said lead time..

For the chips, they are within one week of the designated shipping time of sample chips. You will find the chips aren't the problem in their production right now...
hero member
Activity: 630
Merit: 500
June 11, 2013, 04:53:26 AM
...
I'm not sure, obviously, what the future holds, but I'm fairly certain it is neither linear nor exponential.
I agree with that. Exp growth is absolutely not sustainable.
sr. member
Activity: 294
Merit: 250
http://coin.furuknap.net/
June 11, 2013, 04:29:22 AM
We have started just a nice discussion here.
I have been thinking about that model used here - constant network growth on each retarget. I don't think it's possible even for a year.
I created a chart with data from this year and +12% on each reatrget. That's exponential growth up to 26,000 Thash by the end of next year (dec 2014)! Just make a chart with all that data and you will see that sharp exponential growth.
I don't see how all that hash power can be added in a year. I see something like linear growth up to 3,000 or maybe 5,000 Thash in that time.

This is a common mistake that opponets of mining investments do; to extrapolate the future based on a limited sample size. Since February, where we had roughly 20TH, we have now grown to 7x and almost touched 140TH just a couple of days ago. In the course of four months, we've increased by 700%, which would mean that a similar growth would put the network at 1PH/s in October, 7PH/s in February 2014, 49PH/s in June 2014, and 280PH/s in December.

Fixing the growth in a number of TH/s per period will also lead to a declining growth over time, so that is equally false, but seems to be a favorite especially when arguing against PMBs. "We've grown 20TH a month the past month, just imagine how huge that number is in a year!" That is actually far better than increasing exponentially...

I wrote a bit about this in my latest article on PMBs as well, which really applies to hardware mining as well: http://coin.furuknap.net/are-perpetual-mining-bonds-scams-not-really/

I'm not sure, obviously, what the future holds, but I'm fairly certain it is neither linear nor exponential.

.b
sr. member
Activity: 308
Merit: 250
decentralizedhashing.com
June 10, 2013, 11:30:43 PM
Right, but the main person that I am watching out for isn't the company, it's the consumer.  In this case I'm having a hard time figuring out how to make it fair.  The company says they will ship in 9-10 weeks.  The company has said a lot of things about times, and none of them have bared out.  In the last few weeks there was a personal announcement about sample chips coming out in a week from Yifu.  It took, what two, three weeks just to ship?  How can I claim that a person will make an ROI of such and such when the proof and general consensus is otherwise?  There have been so many companies that have come on claiming this and that out the first day, people asking why aren't I putting them on the list. 
Here is my method for gathering data:
First look at what the claims made by the company are, sifting for inconsistencies and warning signs.  Then I'll generally check through Coin Canary and talk to Operatr about them.
If the company has been thoroughly tried and the general consensus is that they are at least up to par with basic expectations I'll put them on and watch.  There are companies right now that I feel good about, but I'm checking first.  Just announcing doesn't get you on.  In the same way just saying something doesn't necessarily compel me to post it.  I'm thinking of this as a consumer watch group rather than an aggregation of company data.  Perhaps I'm not being clear on that, there is probably a way to change the title.
Though I could see how splitting columns to something like "Company stats"  "Adjusted stats" or something that wasn't confusing might work... 
So I guess by objective, I meant that I'm trying to take all available data and post it without being attached.  I would like to know what others think is a reasonable estimate on lead time for chips bought today would be.  Maybe November is a bit much, I would be willing to compromise if you have a compelling case.  I thought 3 months was going easy on them since they are WAY past that on their other orders.  How are they going to catch up?

I want the opinion base to be as broad as possible.  This discussion is very valuable to me, and if you or anyone else in the community would like to continue it I'm always open for ideas.
hero member
Activity: 648
Merit: 500
June 10, 2013, 07:18:55 PM
I pushed all the companies requiring Avalon chips back a couple months.  There is no reason to believe that 9-10wks is a reasonable delivery time, when there are people waiting much longer than that.  If people start confirming that their chips came in, and it looks like the lead time is fixed, I'll change this.

The spreadsheet should be objectively based on the manufacturers advertised lead time. Changing information is personal opinion, highly subjective, misleading, and does not give an accurate portrayal of the companies. A column could be added to describe the likelihood a company will deliver on time, but arbitrarily changing lead times is dishonest, regardless of the intentions.

Additionaly, the prices of companies which are based on opensource designs are not being represented equally, as each is offering a different level of finish. Assembly is not necessarily the same thing from each company. Some are offering parts only, some parts and accessories, some fully assembled units with cases and power supplies.

That being said, thank you for taking the time to work on this.
I actually agree with this.  I tried to keep the table as objective as possible in the manner you describe.  I'm adding BFL back in today or tomorrow for this very reason.  I took them off for more signaling reason's than any belief that they will never deliver.  Unfortunately the only signaling in pushing back Avalon's time is that it makes the people dependent on them look less competitive.  However, I did not move them back for the same reason.  They were moved back because from the information I have found, it seems quite unreasonable to believe that 9-10 weeks is but a remote possibility for the people unlucky enough to not have made an order yet.  Though I really should go change the lead time for companies that are selling their own chips in miners...  That could happen even before the 9-10 week said lead time..  wow I just realized I need to go ask everyone what they think their own personal lead time is based on what their chip order is.  Well again thanks, it's not fair to put everyone together in this way, maybe I can get away with just putting everyone who is selling full miners with a preorder of chips a month out. 
I never intended to say that people who already have preorders wouldn't get their chips until November, better go fix it..

Awesome work!  Shocked

Thanks, I'm trying to help people figure out what's going on so we can have more decentralized hashing Smiley

That is not objective, or based on the figures each company is providing.

Whether or not you believe a company can meet their timelines is irrelevant, and should at the very least be placed in a separate column. Perhaps something like a stated personal belief in the likelihood the company will ship successfully.

Speaking for myself, the lead time for any batch purchased is 9-10 weeks for delivery from avalon, and 3 days for shipping assembled miners. Adding two and a half months to my timeline is an inaccurate representation of what I am offering.

There are plenty of threads on this forum claiming x company will never ship. I believe you are trying to build a spreadsheet that accurately represents the companies in this space and what they are offering, and I thank you for the time you have invested in doing so.
sr. member
Activity: 308
Merit: 250
decentralizedhashing.com
June 10, 2013, 06:22:00 PM
I pushed all the companies requiring Avalon chips back a couple months.  There is no reason to believe that 9-10wks is a reasonable delivery time, when there are people waiting much longer than that.  If people start confirming that their chips came in, and it looks like the lead time is fixed, I'll change this.

The spreadsheet should be objectively based on the manufacturers advertised lead time. Changing information is personal opinion, highly subjective, misleading, and does not give an accurate portrayal of the companies. A column could be added to describe the likelihood a company will deliver on time, but arbitrarily changing lead times is dishonest, regardless of the intentions.

Additionaly, the prices of companies which are based on opensource designs are not being represented equally, as each is offering a different level of finish. Assembly is not necessarily the same thing from each company. Some are offering parts only, some parts and accessories, some fully assembled units with cases and power supplies.

That being said, thank you for taking the time to work on this.
I actually agree with this.  I tried to keep the table as objective as possible in the manner you describe.  I'm adding BFL back in today or tomorrow for this very reason.  I took them off for more signaling reason's than any belief that they will never deliver.  Unfortunately the only signaling in pushing back Avalon's time is that it makes the people dependent on them look less competitive.  However, I did not move them back for the same reason.  They were moved back because from the information I have found, it seems quite unreasonable to believe that 9-10 weeks is but a remote possibility for the people unlucky enough to not have made an order yet.  Though I really should go change the lead time for companies that are selling their own chips in miners...  That could happen even before the 9-10 week said lead time..  wow I just realized I need to go ask everyone what they think their own personal lead time is based on what their chip order is.  Well again thanks, it's not fair to put everyone together in this way, maybe I can get away with just putting everyone who is selling full miners with a preorder of chips a month out. 
I never intended to say that people who already have preorders wouldn't get their chips until November, better go fix it..

Awesome work!  Shocked

Thanks, I'm trying to help people figure out what's going on so we can have more decentralized hashing Smiley
hero member
Activity: 648
Merit: 500
June 10, 2013, 05:28:12 PM
I pushed all the companies requiring Avalon chips back a couple months.  There is no reason to believe that 9-10wks is a reasonable delivery time, when there are people waiting much longer than that.  If people start confirming that their chips came in, and it looks like the lead time is fixed, I'll change this.

The spreadsheet should be objectively based on the manufacturers advertised lead time. Changing information is personal opinion, highly subjective, misleading, and does not give an accurate portrayal of the companies. A column could be added to describe the likelihood a company will deliver on time, but arbitrarily changing lead times is dishonest, regardless of the intentions.

Additionaly, the prices of companies which are based on opensource designs are not being represented equally, as each is offering a different level of finish. Assembly is not necessarily the same thing from each company. Some are offering parts only, some parts and accessories, some fully assembled units with cases and power supplies.

That being said, thank you for taking the time to work on this.
Pages:
Jump to: