Topic: A4 Dominator - Pre-Order Group Buy - 280mh, roughly 1000w, $1800 + shipping - page 29. (Read 122571 times)

Hi! Guys, does anyone connect to A4 via ssh? Ssh server works, but i cant find any info about auth data for connect ...

I had a pi go bad, and moved the cubes to other pis.
When I run 3 cubes on a pi they are less stable, but im not having as much bad luck as you. Mine will run for a while, but not as long as a pi with 2 cubes.

Yeah ...looks like DATA HALL all the way from the above to me too.

Also of note....where all the LTC difficulty is coming from.....

A buddy said the unknown pools area for LTC went up way way up recently (see chart) the UNKNOWN area that is of pool shares

So assuming these are the big time Innsilicon ISP investors for the A4 data halls going up..they are running their own pools etc this makes some kinda sense on where the
difficulty on LTC rise came from (imho)

here is the link to current unknown part of LTC pools

https://www.litecoinpool.org/pools

perhaps someone can find a link with what the unknown rate was a couple months ago as a comparison.....I'm not having any luck

others with more skills add on to this ...see if the above has any merit....

anyway what I was told ...a big jump in unknown area of share of pools...on LTC ...thus probably the big A4 investors data halls hitting the LTC world

that's a lot of mining for LTC going down or sideways in price.... A4 ISP big money investors....the regular A4 consumers...this new bw.com miner data hall annoucement...sheesh
not to mention the good old knc titans still going strong on revised 3rd party gen tarkin firmware and cooling/heatsink mods etc..heh...likely good for another 2 years now in
that I myself have cut my dc/dc die temps by 10c to 15c ..sheesh...er anyway moving on.....we be full up on miners....no real ltc development (that I've heard) and pressure
from other alt coins like zcash and ethereum..fun times indeed

Also (further up in this thread I think it was stated) ..anyway ..read that innsilicon on their main page supposedly says they are LTC miners...do not mention scrypt...
thus with that and everyone mining to cash or btc asap to pay off equip ..including the big data hall boys above..this could get real interesting

It doesn't sound like they will be selling them, just building and running them themselves, unless I missed something?

Guys.... for those who read my other post on A4 Batch 3

I am missing a controller and 1 cube - still yet to be delivered by DHL...

So had to mix & match to cubers and available controllers to test them on pools.

The results were not impressive and  board still drops but probably not as bad as Batch 1 and 2, because somehow it will climb back up.


TODAY at my ISP:

We tested another format, ie. 1 x controller and only 2 cubes (140Mhs x 2 = 280Mhs) x 4 sets and the last controller (No.5) will handle 3 cubes (140Mhs x 3 = 420Mhs)

So far so good on controller 1 to 4.

Controller 5 is unstable and dropping boards, comes up but cant keep it up for minutes...

On the other hand, controllers 1 to 4 are hashing as per normal - we will see if these hold overnight.

So.... am I right to say that the boards are dropping because A4s cannot handle more than 2 cubes (6x2 =12 blades) per controller?

Is this scenario applicable to Batch 1 and 2?

Looks like Inno is about to have some competition: http://v.bitbank.com/statics/html/video.html?youkuId=14651978969474500047&youkuTit=Bitbank%20Invests%20in%20Research%20and%20Manufacture%20Litecoin%20Chips%20and%20Miners

Yea, my thoughts exactly.  I would probably buy 2 or 3, mine with them and stick a few on amazon, but if they aren't stable, then there is no reason to at this point.

If these A4s can't do automatic coin switching, then they are way overpriced as this infers these are effectively LTC mining rigs as currently architected, even with Batch 3 machines.

In the end, Innosilicon is now only claiming to mine LTC with the A4, no other coin, not even the mention of scrypt in the final publication.
http://www.innosilicon.com/html/mining-asic/14.html

The jury is still out on that, I'm afraid.  I'm ready to pull the trigger if/when we see some stability, but it hasn't happened yet.

I received a reply back for pricing from Innosilicon for pricing from single units up to group buy pricing, but I am wondering if it is worth it at this point.  I believe, if the machine works as intended at the advertised speed and whatnot that they will make money, the question is, will it be stable?

Are people finding that these are remaining stable with the most recent version of miner and firmware?  I don't want something I will have to check on every couple of hours.

Agreed and x2 to this, I had issues immediately and spoke about them here well within the warranty period.

Even if I ignore the restarts which is hard to do when the pool spends a lot of time getting back up to the reported hash rate of the miner.  My miners are not only still restarted frequently on Prohashing, but dropping cards a lot.  Which is just odd that on a straight pool like Litecoinpool that happens a lot less than the multicoin pools.  If I hadn't spent a lot of time watching the miners closely this weekend the miners would have been down 1 or in some cases 2 boards for a lot of the time.  And strangely the miner does not seem to act like it knows the board is missing.  It will keep running until something triggers the restart again, and in many cases it won't pick that board back up even with the restarts.  In most cases I have to stop cgminer, restart it a couple of times before it will pickup the board if I am lucky.  The worst case is having to go power cycle the miner.  I am still pointed at Prohashing at the moment, but I have to keep a close eye on things or they go sideways pretty frequently.  This is just hard to do when you have a day job and other responsibilities to be constantly babysitting these things.

I am starting to conclude that Inno fully intends to let us twist in the wind on this.  Email to Chloe has gone unanswered after I questioned this policy of having customers replace chips and reprogramming them and asked them to arrange for contracted support to get these B1 and B2 miners fixed or to have them send out the replacement boards.  I hope they are just busy fixing this and coming  up with a solution for their customers.  This is not keeping their promises to the community if you ship product and then expect the customers to fix their bad product which hasn't worked properly from day one.  It isn't like they weren't told right away.  This should be covered under warranty and fixed.  They acknowledge they have an issue now they need to step up and take care of the customer.

If they come through and take care of us then I will buy more miners.  Right now I won't spend any more money with Inno until they take care of this. I need to buy some more miners, but it looks like at this point I cannot with a good conscience place another order with Inno when I don't know what I am going to get or if they are going to fix an of it when there is something wrong.  I would rather keep my business here with Inno, but they are making that very difficult to do at this point.  I will have to make that call in the next few weeks as I need to buy and stand up some more rigs so I hope we get some clarity and resolution from Inno before then.  I am not holding my breath at the moment.

Inno needs to speak up and let us know what is going on.  And they need a customer focused policy that fixes the stuff they already shipped that is broken. 

Try not to think about the restarts. As some others have noted, the miners can reach 350+MHs for a few minutes after they restart, I've even had some report a 2hr average of 330Mhs from time to time, so I've ignored the restarts in search of 2hr Averages which is what I think matters and one of the reasons I've stuck with prohashing as the per miner stats are way beyond what you can get on Clever-mining or LitecoinPool. I pulled the plug on one just as it restarted the other day and you can see the last 2 lines of the cgminer.log read:

[date] Tan Restart by line goto Restart
[date] Tan Return 0

.. so there's definitely some kind of error control code in cgminer that hot restarts the cards (written by Tan who compiled the cgminer that's used). I have to say it got warm here today and my perception is that they were a little less stable than they have been lately so temp may play a role even though Inno have been reported as saying up to 65 degree's C is ok.

I'd still like to know how this works "Server Grid Reliability: Built-in high-procession temperature sensor for automatic chip protection and variable frequency operation to maximize chip life time" from http://www.innosilicon.com/html/news/13.html

I think that once this restart bug is resolved through the firmware update, we're actually looking at miners that may be capable of 300Mhs++. I'm speculating, but if the restarts are what's keeping the hashrate down and Inno can fix that, then they may get a lot closer to their original design goals.

It's a total PITA that we're lumbered with this, but early adopters are often unwilling test partners, that's just how it is.

Well they've been silent from the beginning and i dont expect miracles even now. It looks like B1 ans B2 users are left on their own. Its either wait for custom jig which is weeks away from now and will be sent to only a few customers or start taking soldering/desoldering courses.

I hope the silence is an indication that the rumors are true and that B3 has fixed the issues.

However, I have concern about how us B2 and B1 owners are going to get to that point.

Innosilicons silence on this issue is bringing about some hard questions as to why they are silent.

-Warren Buffett

Any news to report on the B3 miners?  Any improvements?