ComputerGenie
|
|
January 05, 2017, 04:38:43 PM |
|
... It's just there are bad business people not treating their equipment right that make it seem that there is a high failure rate.
AKA hobbyists
|
If you have to ask "why?", you wouldn`t understand my answer. Always be on the look out, because you never know when you'll be stalked by hit-men that eat nothing but cream cheese....
|
|
|
pondule
Newbie
Offline
Activity: 24
Merit: 0
|
|
January 05, 2017, 05:39:30 PM |
|
... It's just there are bad business people not treating their equipment right that make it seem that there is a high failure rate.
AKA hobbyists I tripped over the power cable to one of the PSUs running my S9 and it spiked the mains and now that hash board doesn't power up. PSU is still good so I suspect the regulator has died but it's too cold to take it apart because two S7s aren't enough to keep my flat warm. Is that what you mean by hobbyist?
|
|
|
|
ComputerGenie
|
|
January 05, 2017, 05:42:10 PM |
|
...Is that what you mean by hobbyist? When I stop lmao, I will answer that.
|
If you have to ask "why?", you wouldn`t understand my answer. Always be on the look out, because you never know when you'll be stalked by hit-men that eat nothing but cream cheese....
|
|
|
philipma1957
Legendary
Offline
Activity: 4256
Merit: 8595
'The right to privacy matters'
|
|
January 05, 2017, 06:13:17 PM Last edit: January 05, 2017, 06:34:10 PM by philipma1957 |
|
If you are a trucking company would you buy the truck that costs more and need more fuel but drive on the road all the time, making money. And if something is defective you have very good customer service. Or the one that is cheaper, needs less fuel but you have troubles all the time, you have to argue with the customer service about very cheap spare parts even if you bought hunderts of trucks from them.... ...
As for your analogy.... That's so flawed that I don't even know where to begin.... Since 66 is the LCM... | Unit Cost | Power Usage | TH/s | Power/TH | (kWh) cost USD | Yearly Power Cost/TH | Yearly Power Cost @ 66TH | Purchase Cost for 66TH | A7 | $1,222.29 | 1,000.00 | 6 | 166.67 | $0.10 | $146.39 | $9,661.87 | $13,445.19 | S9 | $1,232.00 | 1,185.80 | 11 | 107.80 | $0.10 | $94.67 | $6,248.40 | $7,392.00 |
Edit: purchase price alone, you could buy 11 S9s (not counting the 1 that the power cost difference makes), throw 5 in the trash (having never used them) and still save $. your price is off on the a7 they are $888.00 before shipping not 1,222.29 I paid 2,032 for 2 with 1 controller so if you use 1016 not 1222.29 you total for 11 = 11176 that counts shipping and gives you 5 rasp pi's my last s9 was 60 to ship so 1232 + 60 = 1292 x 6 = 7752 so s9 = 7752 avalon 7 = 11,176 if 1 s9 breaks that is 16.667% loss if 0 A7 breaks that is 0% loss so 7752 x 1.16667 = 9044 act cost of the s9 not 7752 so compare 9044 to 11,176 next if the old s9 is really 83% block luck 9044 x 1.16667 = 10551 is what 66th of old s9 costs you vs 11,176 power you did at 10 cents do it at 5 cents a7 = 4831 not 9661 s9 = 3124 not 6248 So this would be a lot closer to realty for people that can do 66th of gear and purchased the old s9's cost of the old s9's were a lot higher then the numbers used here. but some of the time blocks were 25 not 12.5 But that just numbers. So if you want to prove a point at least use the right prices for the avalon 7's it still shows that s9's are better not a lot but still better
|
|
|
|
ComputerGenie
|
|
January 05, 2017, 06:22:27 PM |
|
your price is off on the a7 they are $888.00 before shipping not 1,222.29
I paid 2,032 for 2 with 1 controller so if you use 1016 not 1222.29 you total for 11 = 11176 that counts shipping and gives you 5 rasp pi's
Edit 2: *based on retail price in yxt's singature So, @ $1016, you only get to throw 4 unused S9s in the trash instead of 5
|
If you have to ask "why?", you wouldn`t understand my answer. Always be on the look out, because you never know when you'll be stalked by hit-men that eat nothing but cream cheese....
|
|
|
tournamentdan
Member
Offline
Activity: 118
Merit: 10
|
|
January 05, 2017, 06:23:37 PM |
|
... It's just there are bad business people not treating their equipment right that make it seem that there is a high failure rate.
AKA hobbyists I tripped over the power cable to one of the PSUs running my S9 and it spiked the mains and now that hash board doesn't power up. PSU is still good so I suspect the regulator has died but it's too cold to take it apart because two S7s aren't enough to keep my flat warm. Is that what you mean by hobbyist? Well I didn't quite mean it like that. I think the difference between the two is how much money you have to invest. And with out a doubt there are stupid people with money to invest. And smart people with little money to invest. This thread does have a lot of hobbyists. And I have learned a good bit off of them. For example. There were many people on this very thread that wrote "do not buy the first few batches of the s9. They will have problems just like the S5 and s3 did". And I did not buy until batch 5. Which must be the reason why I am not having any problems. So it sounds like other people did not use the same wisdom.
|
|
|
|
|
ComputerGenie
|
|
January 05, 2017, 06:29:33 PM |
|
...There were many people on this very thread that wrote "do not buy the first few batches of the s9. They will have problems just like the S5 and s3 did". And I did not buy until batch 5. Which must be the reason why I am not having any problems. So it sounds like other people did not use the same wisdom.
Like I said, I'm not saying that they are without issues (I have some from batch 4 that run ~10 rather than ~13, but I'm "OK" with it because the ROI on downtime and shipping would take 27 months to equalize). I'm simply saying that you have to be completely unconcerned with ROI, or just a complete fanboy, to say that the 2 rigs (S9 and A7) are even in the same market for comparison.
|
If you have to ask "why?", you wouldn`t understand my answer. Always be on the look out, because you never know when you'll be stalked by hit-men that eat nothing but cream cheese....
|
|
|
ComputerGenie
|
|
January 05, 2017, 06:50:05 PM |
|
After edit: ...
OK, so using your revised set of numbers, for the same $ purchase, you get 8 S9's, 4 backup control boards, and 1 backup hashboard.... Which gives you 2 rigs on the shelf (to run when your imaginary "16.667% loss" takes place, twice...)... So, with the S9 package, we have spare parts for all of the controllers (2 from the shelved rigs plus the 4 for equal cost), 7 spare hash boards (6 from the shelved rigs), greater power efficiency (enough to buy a 9th rig just from savings, by your own numbers), and someone's still a fanboy?
|
If you have to ask "why?", you wouldn`t understand my answer. Always be on the look out, because you never know when you'll be stalked by hit-men that eat nothing but cream cheese....
|
|
|
firetreeactual
Legendary
Offline
Activity: 952
Merit: 1003
|
|
January 05, 2017, 07:00:46 PM |
|
Well I didn't quite mean it like that. I think the difference between the two is how much money you have to invest. And with out a doubt there are stupid people with money to invest. And smart people with little money to invest. This thread does have a lot of hobbyists. And I have learned a good bit off of them. For example. There were many people on this very thread that wrote "do not buy the first few batches of the s9. They will have problems just like the S5 and s3 did". And I did not buy until batch 5. Which must be the reason why I am not having any problems. So it sounds like other people did not use the same wisdom.
I disagree. It's not how much you have to invest; it's what proportional importance mining has to your overall situation. A hobbyist does something because they enjoy it, and can afford to "lose" money on it...perhaps. I've been mining since the beginning, and I think that to assume because I run around 24TH at this point in time that I'm a hobbyist is...frankly...ridiculous. I'm partially disabled and live primarily on VA disability and SS. Mining provides a significant percentage of my net spendable (or invest-able, depending). OTOH, there was at time back in the day when I had a net worth of ~USD15M (1980 dollars). The crash in '07, plus nearly dying, drained every pence of my resources, and changed all that. So...I think those sorts of comparisons are not very helpful, because I'm not unique in that way. My last point is that one should not have to think about waiting for the first few batches to show up the real bugs before buying...or that there should be a situation requiring...basically...running the mfr's burn-in beta tests at the expense of your customers' power and frustration. Now, THAT sucks.
|
To infinity and beyond...on two 741s and one of only 3...nope, make that 4...full nodes in Hawaii...on <30A. (I have other gear on the Hoth ice planet)
|
|
|
firetreeactual
Legendary
Offline
Activity: 952
Merit: 1003
|
|
January 05, 2017, 07:21:11 PM |
|
Well...from the look of it (haven't had time to read into it)...Canaan is working on the next iteration, which I would expect for them to be doing at this point. That's the great thing about open source. I should take a look at the code and see what I can see.
|
To infinity and beyond...on two 741s and one of only 3...nope, make that 4...full nodes in Hawaii...on <30A. (I have other gear on the Hoth ice planet)
|
|
|
-ck
Legendary
Offline
Activity: 4242
Merit: 1644
Ruu \o/
|
|
January 05, 2017, 08:04:25 PM |
|
Don't get me wrong, I'm not saying that S9s aren't a steaming pile of bovine dung; I'm saying that, dung or not, they pay the bills better than the current alternatives.
Not if they're not finding blocks... I am not sure how to interpret this from a practical point of view. do other pools have the same phenomenon? What about slush's or other non-bitmain associated pools? If they do-I will subscribe to this argument of "poison" S9 v1 If they don't-? Are you saying the bad luck this pool is having is not really bad luck, just S9 not finding blocks on this pool?
Meanwhile the latest S9v1 statistics on the pool since the S9 existed:
They've found 82 Blocks, but expected to have found (BDR) 110.666 ...
Kano does a luck report regularly to find block withholders - they show up as making no blocks at all over an extended period when they should have made some (it's random so you can't detect it until they've been hashing many blocks' worth). He reports it here regularly though I'm not sure everyone understands the implications of the report. The S9 batch 1 across all miners as you see above by now should have found 110 blocks based on the total hashes they've submitted to this pool but have so far only found 82. This is by far worse than any other device that we can positively identify. We can only speculate why this is, but the batch 1 firmware also had some other stratum extension that was never working on any pools so it was removed in later firmware. It is possible there is some kind of bug in the original firmware. This is why he was urging everyone with S9s that was still on v1 firmware to upgrade. I'm aware that you lose some kind of frequency control or fan control or something by upgrading but that's nowhere near as bad as (potentially) losing blocks. As for whether other pools are affected, we can only report accurately what happens here, and only speculate what is happening on other pools based on people grumbling about bad luck on almost every other pool thread. It certainly looks like global luck has been down across all pools since the S9 came out, but I can't say anything for certain.
|
Developer/maintainer for cgminer, ckpool/ckproxy, and the -ck kernel 2% Fee Solo mining at solo.ckpool.org -ck
|
|
|
philipma1957
Legendary
Offline
Activity: 4256
Merit: 8595
'The right to privacy matters'
|
|
January 05, 2017, 08:12:44 PM |
|
Don't get me wrong, I'm not saying that S9s aren't a steaming pile of bovine dung; I'm saying that, dung or not, they pay the bills better than the current alternatives.
Not if they're not finding blocks... I am not sure how to interpret this from a practical point of view. do other pools have the same phenomenon? What about slush's or other non-bitmain associated pools? If they do-I will subscribe to this argument of "poison" S9 v1 If they don't-? Are you saying the bad luck this pool is having is not really bad luck, just S9 not finding blocks on this pool?
Meanwhile the latest S9v1 statistics on the pool since the S9 existed:
They've found 82 Blocks, but expected to have found (BDR) 110.666 ...
Kano does a luck report regularly to find block withholders - they show up as making no blocks at all over an extended period when they should have made some (it's random so you can't detect it until they've been hashing many blocks' worth). He reports it here regularly though I'm not sure everyone understands the implications of the report. The S9 batch 1 across all miners as you see above by now should have found 110 blocks based on the total hashes they've submitted to this pool but have so far only found 82. This is by far worse than any other device that we can positively identify. We can only speculate why this is, but the batch 1 firmware also had some other stratum extension that was never working on any pools so it was removed in later firmware. It is possible there is some kind of bug in the original firmware. This is why he was urging everyone with S9s that was still on v1 firmware to upgrade. I'm aware that you lose some kind of frequency control or fan control or something by upgrading but that's nowhere near as bad as (potentially) losing blocks. As for whether other pools are affected, we can only report accurately what happens here, and only speculate what is happening on other pools based on people grumbling about bad luck on almost every other pool thread. It certainly looks like global luck has been down across all pools since the S9 came out, but I can't say anything for certain. If you absolutely can not upgrade to the new autotune firmware (my case until April) You can pull from this pool and point to Antpool or f2pool Just point auto tune s9's here or in my case avalon 7's. At the moment I have all my avalon 7's here and all my s9 old firmware at f2pool. I am hoping to get some bitfury and point here. The goal is to get all old firmware s9's off this pool or updated to auto firmware if you can.
|
|
|
|
ComputerGenie
|
|
January 05, 2017, 08:28:56 PM |
|
...The S9 batch 1 across all miners as you see above by now should have found 110 blocks based on the total hashes they've submitted to this pool but have so far only found 82. This is by far worse than any other device that we can positively identify....
Yeah, no. As I posted before...... ... Name CDF[Erl] S9v1 0.994766 103.6 BDR 79 Blocks ... A7v1 0.772224 6.5 BDR 5 Blocks ...
... When extrapolated, 6.5:5 is virtually* the same as 103.6:79! If you absolutely can not upgrade to the new autotune firmware (my case until April) You can pull from this pool and point to Antpool or f2pool
Just point auto tune s9's here or in my case avalon 7's. At the moment I have all my avalon 7's here and all my s9 old firmware at f2pool.
I am hoping to get some bitfury and point here.
The goal is to get all old firmware s9's off this pool
or updated to auto firmware if you can.
See above and see why you should be making the exact same appeal to A7 owners! *< 1 block difference in 103 total blocks
|
If you have to ask "why?", you wouldn`t understand my answer. Always be on the look out, because you never know when you'll be stalked by hit-men that eat nothing but cream cheese....
|
|
|
Biodom
Legendary
Offline
Activity: 3892
Merit: 4333
|
|
January 05, 2017, 08:32:19 PM |
|
Kano does a luck report regularly to find block withholders - they show up as making no blocks at all over an extended period when they should have made some (it's random so you can't detect it until they've been hashing many blocks' worth). He reports it here regularly though I'm not sure everyone understands the implications of the report. The S9 batch 1 across all miners as you see above by now should have found 110 blocks based on the total hashes they've submitted to this pool but have so far only found 82. This is by far worse than any other device that we can positively identify. We can only speculate why this is, but the batch 1 firmware also had some other stratum extension that was never working on any pools so it was removed in later firmware. It is possible there is some kind of bug in the original firmware.
This is why he was urging everyone with S9s that was still on v1 firmware to upgrade. I'm aware that you lose some kind of frequency control or fan control or something by upgrading but that's nowhere near as bad as (potentially) losing blocks.
As for whether other pools are affected, we can only report accurately what happens here, and only speculate what is happening on other pools based on people grumbling about bad luck on almost every other pool thread. It certainly looks like global luck has been down across all pools since the S9 came out, but I can't say anything for certain.
it is not that simple (re upgrade). New firmware does not work quite well on old (non-autotuned) miners, at least for me. I tried and miners overheated. realistic choice is to point S9 v1 miners to other pools and that's what people probably did and Phil mentioned this. Personally, i think that pushing this proposition might result in a loss of much hashing power of your pool, but it is your playground, you make the rules. re statistics, wasn't that 68-70 vs 100 expected before? If so, then more recent blocks are 12 vs 10 expected. I would like some specialist analyze all of it. Can organofcorti take a look at all of this? Many people would be interested. In clinical trials data based on a subset that was not initially established usually turns out to be a false signal, but who really knows here.
|
|
|
|
-ck
Legendary
Offline
Activity: 4242
Merit: 1644
Ruu \o/
|
|
January 05, 2017, 08:35:44 PM |
|
...The S9 batch 1 across all miners as you see above by now should have found 110 blocks based on the total hashes they've submitted to this pool but have so far only found 82. This is by far worse than any other device that we can positively identify....
Yeah, no. As I posted before...... ... Name CDF[Erl] S9v1 0.994766 103.6 BDR 79 Blocks ... A7v1 0.772224 6.5 BDR 5 Blocks ...
... When extrapolated, 6.5:5 is virtually* the same as 103.6:79! If you absolutely can not upgrade to the new autotune firmware (my case until April) You can pull from this pool and point to Antpool or f2pool
Just point auto tune s9's here or in my case avalon 7's. At the moment I have all my avalon 7's here and all my s9 old firmware at f2pool.
I am hoping to get some bitfury and point here.
The goal is to get all old firmware s9's off this pool
or updated to auto firmware if you can.
See above and see why you should be making the exact same appeal to A7 owners! *< 1 block difference in 103 total blocks No. Sample size is too small to draw the same conclusion with only 6.5 total BDR. Plus we're not saying CONCLUSIVELY that it is a problem, but upgrading firmware, where you can, is desirable since it MAY be a problem.
|
Developer/maintainer for cgminer, ckpool/ckproxy, and the -ck kernel 2% Fee Solo mining at solo.ckpool.org -ck
|
|
|
suchmoon
Legendary
Offline
Activity: 3808
Merit: 9024
https://bpip.org
|
|
January 05, 2017, 08:48:10 PM |
|
So, @ $1016, you only get to throw 4 unused S9s in the trash instead of 5 Can you even buy a new S9? Every time I look it's out of stock. The only real advantage of the S9 is its resale value, which for some inexplicable reason is still very high. Although if you consider BTC exchange rate those ~4 BTC early batches don't sound so good. Other than that both the S9 and the A7 sound like shitty investments considering recent diff jumps and Bitmain firing up its Hashnest and whatnot.
|
|
|
|
ComputerGenie
|
|
January 05, 2017, 08:52:41 PM |
|
...The only real advantage of the S9 is its resale value, which for some inexplicable reason is still very high...
Unless you're dying and need quick liquidation or a more efficient model comes out, who cares about resale value?
|
If you have to ask "why?", you wouldn`t understand my answer. Always be on the look out, because you never know when you'll be stalked by hit-men that eat nothing but cream cheese....
|
|
|
-ck
Legendary
Offline
Activity: 4242
Merit: 1644
Ruu \o/
|
|
January 05, 2017, 08:54:14 PM |
|
While you and Kano might not be, in order to apply that statement broadly, it seems you've only been skimming every 3rd post or so. Me? I hate how much time I spend on this forum: https://bitcointalk.org/index.php?action=profile;u=19971;sa=statPanelI'd love to avoid coming here as much as I do since I have a sense of obligation now that I am a moderator, but all the other moderators have gone idle I do read far too many threads so don't assume I've been skimming (it's not exactly all enjoyable reading, nor the friendliest forum on earth.) Giving the recommendation I did did not come lightly. Even kano has avoided explicitly spelling it out the way I did and he's very conscious of the potential loss of significant hashrate if people just point their hashes elsewhere. I did not tell people to point their hashes elsewhere, but I'm recommending people upgrade IF they can AND it suits. You're right that I did not explicitly say so in the earlier post though.
|
Developer/maintainer for cgminer, ckpool/ckproxy, and the -ck kernel 2% Fee Solo mining at solo.ckpool.org -ck
|
|
|
suchmoon
Legendary
Offline
Activity: 3808
Merit: 9024
https://bpip.org
|
|
January 05, 2017, 09:05:10 PM |
|
...The only real advantage of the S9 is its resale value, which for some inexplicable reason is still very high...
Unless you're dying and need quick liquidation or a more efficient model comes out, who cares about resale value? I'm pretty sure there are more than just those two reasons for reselling miners When the resale value exceeds the reasonable expectation of what you can possibly mine with the device - you should get rid of it (or not buy it to begin with) - wouldn't you agree? Unless it's a hobby. In which case none of the ROI talk matters anyway.
|
|
|
|
|