Generally weekends are lull periods in new hashrate. Though there could be a 5%+ jump in the following difficulty adjustment with the bulk pricing spondoolies and bitmain have for the weekend |
|
|
|
2 boards on my unit are dead and the pump works fine. machine is fried wat a darn bummer... praying bitmain blesses me but that other guy earlier is still waiting but i think his was a cooling issue... mine is just simply toast ://// pleade bless me with some reliable s3's bitmain... please  and help me with shipping and a hefty coupon on my next batch of s3 orders sunday. that sounds fair to me... equal exchange for s3's with reimbursed shipping in form of coupon for my sunday s3 order. @dogie i understand they do not reply right away. i just pray they follow up well and this doesnt turn into a me nagging them and making me into the bad guy (even tho i have the worlds most beautiful paperweight) open up the unit and check all the cables. two totally dead boards is strange, and seems like theres a common issue. Perhaps they are sharing a weak rail on the PSU or have loose cables. if you're tech-savvy at all, its very easy to open the case and spend 5 minutes with a multimeter to check for 12V and 0.7V presence on the bad boards, and if any status LEDs are lit |
|
|
|
6x 1.4 TH/S PRISMA = 8.4 TH/S mining farm
nice setup  |
|
|
|
Chinesesavoir and Mindtrip both had X's in their ASIC status area. And when you have those X's your (guaranteed) hashing power drops. And after changing the voltage, which I've already done to 740, the X goes away but eventually comes back. Now with 2 X's for even less hashing power. What if I wanted to sell in the future and as the specifications indicate 10008GH/s and the unit can't even get to that, then I'm pretty much screwed. I'm just wondering how every is getting 1000TH/s with no problems and I have 6 of them struggling to keep above 950GH/s. I guess I'm just expecting more support and attention like I'm use to with other companies I have purchased from.
Are you here as a mascot/cheerleader for Bitmaintech or really here for support purposes?
sometimes there are just a few bad or low-preformance chips in every batch. if raising the voltage by 0.01-0.03V doesnt help then its probably not resolvable. even 2 bad chips are only a ~5% below spec, which is almost nothing. There no reason why you cant resell it at a fair price even if its 5% under spec (you might get 5% less for it of course). not sure if you expect bitmain to spend a week RMA-ing your boards over a ~$25 issue unless its drastically outside the +/-5% spec |
|
|
|
still not reply from bitmain
emailed them my order id like they asked this morning around 2-3am pst.
My c1 paperweight is so cool. It has a theoretical hashing of 1ths............except my unit is DOA...
What's the deal bitmain?
Losing 5$/day on this unit. I know it has not been long for them to respond so I will remain patient.
curious to see how these people sort this issue out.
This is the most beautiful paperweight i have ever seen.
how is the unit setup, whats the power supply, what happens when you turn it on? Are you sure the PSU is even providing 850W+ on the correct 12V rail(s)? Its unlikely to have a unit thats totally DOA - most problems are caused by 2-3 common issues. |
|
|
|
all of sp-tech miners are user-friendly and configurable.
So, here's what you need to consider.
Price per GH
Watts per GH
Resale value.
I have been mining with SP-Tech equipment since SP10 (bought in March) and I am STILL running those SP10, alongwith my other SP-Tech miners.
I have found that the customisability of the miner (downclocking) is second to none. I am running each of my old SP10 at max 833watts and getting 1.2TH/s
That is still a profitable miner - EIGHT months from the date I received them. Of course, they have paid me back the purchase price long ago.
I'm now mining with the SP20 and I am underclocking these also;
745watts mining at > 1.35TH/s - i worked out that these will still mine profitable right up until difficulty hits 100billion.
If I were to sell on my old SP10's I'd easily get a decent price for them.
Spondoolies, somehow have created the most efficient, most stable and most trustworthy miners, ever.
in addition, noise is an issue for a lot of people. The SP20 is louder than an S3+ or an S4. regarding resale - Bitmain products generally have the best value as they are very home user friendly - even S1 units can still sell locally for $40-70 each. Spondoolies gear is sometimes a harder sell. not sure why you underclock the SP20 - do you pay >$0.20/kwh for power? Most calculations show you will make more in the next month at 1.7TH/1100W then you will by massively sacrificing hashrate to save some electricity. |
|
|
|
More S1 upgrade kits please!!
you could feasibly use the C1 as a pair of upgrade kits, but its not ideal. Not sure why bitmain isnt selling upgrade kits for ~0.3BTC/kit |
|
|
|
Not melt, but how hot do they get? I have a 750W PSU running on an Atminer S3+ and the cables can get warm to the touch. At 288 watts they must get really hot.
Assuming you have one connector per blade, each connector has only ~180-200W. Almost certainly those cables would npot support 250W+ without having a fan directed at the wiring to cool it. when possible, 16awg wiring is a must. It is capable of 250-320W per pcie connector whereas 18awg is only good for 180-240W |
|
|
|
1) https://bitcoinwisdom.com/bitcoin/difficulty difficulty is slowing, and its unlikely there will be any 10%+ increases until a dramatic price rise occurs - though next week may see a big jump due to the black friday deals by SPTech and Bitmain. 2) most people mining assume the bitcoin price will rise such that profitability is extended 3) you can underclock/undervolt to 0.5w/GH to extend the lifespanIf you look at your calculations you basically assume it wont be profitable past june because you are assuming the price of bitcoin is still only $375 at that time, when (i personally believe) a price of $800+ is likely. If it isnt, you can expect the influx of new equipment to grind even further to a halt as all 1w/gh+ equipment is turned off. explain? https://bitcointalk.org/index.php?topic=796839.0 clearly states that 500W/1000GH is the underclocking limit of the S4 |
|
|
|
4F81 is 400Mhz - did you pencil mod at all? I'm still having some issues with my modding. Ive got 4 units: unit 1: Resisitance in place = 2.05 (+/- 0.06). Voltage = 0.75 (+/- 0.03). Frequency/TO = 206.25/50. All chips hash except for 2 of them - ~102GH unit 2: Resisitance in place = 2.1 (+/- 0.06). Voltage = 0.76 (+/- 0.03). Frequency/TO = 200/65. All chips hash except for 6 of them - ~89GH unit 3: Resisitance in place = 2.13 (+/- 0.06). Voltage = 0.76 (+/- 0.02). Frequency/TO = 212.5/50. All chips hash except for first chip on first board - ~102GH unit 4: Resisitance in place = 2.13 (+/- 0.06). Voltage = 0.76 (+/- 0.02). Frequency/TO = 206.25/60. About 8-16 chips dont hash - ~75.5GH I'm having trouble finding much pattern here. some things i noticed: - timeout >70 results in no hashrate and beeping. chips are generally all visible though - timeout 50 works better than 60+ at frequencies ove 206.25Mhz, but at 200 or 193MHz 60 or 65 timeout seems to work best - at lower frequencies, the last chips in each board tend to not work. at higher frequencies the first chips seem to be a more common issue. - majority of 'x' chips are in the last segment on each board. - if the frequency is too low, it seems to cause issues. whats stable at 200-206MHz is not very stable at 193MHz and a large number of chips read 'x' at 175Mhz any advice? My goal is to move them to a small structure like Chaositec did - but first i want to get each board to hash 90GH or better. I cant tell if i should be setting my voltage even lower or if that will make things less stable so an update is due since a few things happened 1) I disconnected the fans for 40 minutes. on all 4 units the maximum temperature reached while still on the original S1 heatsinks/frame with zero airflow was about 63-69 Celsius. Surprisingly this had little (if any at all) negative effect on spees or hashrate, and in fact seemed to improve some of the units i had issues with. Reconnected the fans and after 3 minutes they were back to running silently at 32-36 C. 2) after the above, more chips appeared to respond. I imagine this is due to the effect of the pencil modded lead tending to 'cure' in place and slightly reduce voltages further. ( i have not re-checked yet) 3) I tuned units 3+4 to 212.5/65 and am getting 102Gh from one (2 dead chips) and 109Gh from the other (no bad chips). Still no wattage readings, but i imagine ~1w/gh as this puts the hardware pretty close to the S2 specs. I have no doubt now that the units can be stacked with spacers like chaositec did and have a single fan maintain temperatures at 55C or less without much noise. Hopefully the ebay standoffs arive soon to do this. The goal will be to run 16 boards (~800w) on a pair of CS650M |
|
|
|
>30% off Black Friday Deals Special Sales for ANTMINER S4, 0.409USD/GH/s, FREE International Shipping. Shipping on Dec. 2nd
Might want to mention that you are forced to buy SIXTEEN units of S4. Also that coupons are not applicable. And finally that without coupons being accepted, that the price per unit works out to being $818. Which is $18 MORE than buying a single unit with a coupon. Shipping appears to be included. For a change. Customs fees I highly doubt are included. As usual. Unless you are in the market to buy 16 units, this is not an actual sale. And even then you save very very very little. Basically cheaper shipping. How about actually giving 30% off the price of a single S4, and still allow the coupons..... so it is not completely and utterly useless to everyone, as this so-called "sale" currently is. The 'sale' is actually a very fair price to medium-level and larger buyers. You can use 1 coupon ($400) per 16-pack. That makes the final price (shipping included) $12,688 - or $397/TH. Thats cheaper than 16 units purchased individually with coupons, particularly considering that the price includes about $2000 discount on shipping costs you would otherwise pay. Thats about 1BTC/TH WITH PSUs DELIVERED. in contrast, the SP20 deal works out to around $415/TH and does not include power supplies (another $40-70/TH) and do not rack-mount. noone includes the custom fees in their price - every customer pays different fees depending on country and personal/business use. IMO, would love to see a 4-pack/$3400 (shipping included) option for the smaller miners. |
|
|
|
Of course you have a point, However i guess up to date ive spent around 50-75 BTC with leegroup and have 20KWh of machines hosted in one of his miner farms. So trust hasnt been given it has been earnt this 20 BTC i can assume is as safe as all my other investments.  why would he pay you back? He can easily run away with the btc, make another account and do the same  but let's see he could - but right now pcfli has a good reputation. I'm cautious, but beleive that the arbitrage market provides great opportunity to those with the available liquidity, and that pcfli can stand to make good profits on investing. |
|
|
|
I've been running some numbers, one thing I see is that the SP-20 vs. the S4 is an under volt advantage that gains you money at the wall in favor of SP.. I may be wrong, but if they come out with new tech ANY batchs of the S4 won't even come close to ROI.. I bought some of the FIRST S4s to ship, and they are no where close to ROI . Even with the coupon, I can't pull the trigger when my first several S4's are so far under water. It takes me almost 15 days of mining to cover electricity cost alone. Anything after 15 days goes to the under water s4's .. This late in the game on the 28nm tech I can't find a reason to grab more. Anyone else have thoughts on how to make the s4 more profitable ?
With a coupon I can get 4 S4s for 3358 to my door. My calculations put me at Mid-February before i am losing money every day. All told I will spend $3358 and make about $1500 in revenue. So I lose $1800. How is anyone buying these and making this work. Even with free power and difficulty increasing at 10% I don't break even until June of next year. I would love to get into mining, but I cannot see how mining can make any sense to anyone anywhere unless you get the hardware for half what is the current going rate per mhs and free power. 1) https://bitcoinwisdom.com/bitcoin/difficulty difficulty is slowing, and its unlikely there will be any 10%+ increases until a dramatic price rise occurs - though next week may see a big jump due to the black friday deals by SPTech and Bitmain. 2) most people mining assume the bitcoin price will rise such that profitability is extended 3) you can underclock/undervolt to 0.5w/GH to extend the lifespan If you look at your calculations you basically assume it wont be profitable past june because you are assuming the price of bitcoin is still only $375 at that time, when (i personally believe) a price of $800+ is likely. If it isnt, you can expect the influx of new equipment to grind even further to a halt as all 1w/gh+ equipment is turned off. |
|
|
|
4F81 is 400Mhz - did you pencil mod at all? I'm still having some issues with my modding. Ive got 4 units: unit 1: Resisitance in place = 2.05 (+/- 0.06). Voltage = 0.75 (+/- 0.03). Frequency/TO = 206.25/50. All chips hash except for 2 of them - ~102GH unit 2: Resisitance in place = 2.1 (+/- 0.06). Voltage = 0.76 (+/- 0.03). Frequency/TO = 200/65. All chips hash except for 6 of them - ~89GH unit 3: Resisitance in place = 2.13 (+/- 0.06). Voltage = 0.76 (+/- 0.02). Frequency/TO = 212.5/50. All chips hash except for first chip on first board - ~102GH unit 4: Resisitance in place = 2.13 (+/- 0.06). Voltage = 0.76 (+/- 0.02). Frequency/TO = 206.25/60. About 8-16 chips dont hash - ~75.5GH I'm having trouble finding much pattern here. some things i noticed: - timeout >70 results in no hashrate and beeping. chips are generally all visible though - timeout 50 works better than 60+ at frequencies ove 206.25Mhz, but at 200 or 193MHz 60 or 65 timeout seems to work best - at lower frequencies, the last chips in each board tend to not work. at higher frequencies the first chips seem to be a more common issue. - majority of 'x' chips are in the last segment on each board. - if the frequency is too low, it seems to cause issues. whats stable at 200-206MHz is not very stable at 193MHz and a large number of chips read 'x' at 175Mhz any advice? My goal is to move them to a small structure like Chaositec did - but first i want to get each board to hash 90GH or better. I cant tell if i should be setting my voltage even lower or if that will make things less stable |
|
|
|
Grrr.. I really want to pull the trigger on an SP31, but it doesn't make sense to me. For $1600 I could have 2x S4, running at 4.4TH overclocked. Or even better 3x for $2400 @ 6.6TH/s. Am I missing something besides the obvious increase in hardware quality, density, and customer support of SP over bitmain?  FUCK bitmain! if you want shitty, thrown together hardware with the cheapest, overpriced components on the planet - Bitmains got plenty - but don't be fooled by empty promises and costly RMA procedures which i can almost guarantee you will need. just my humble opinion based on my own experiences - others opinions may vary Trust me, I hear you loud and clear. I'm stuck dealing with bitmain right now, for compensation and an rma for my latest S4 that came with a dead hash board. Dealing with them is horrible, like pulling teeth. I've got 4 other just about perfectly working S4s, though, and I can't really complain about them since their issues have mostly been resolved. Still a few bugs to squash and they'll be solid. At the end of the day, ROI is still king though. I can get 8.8 TH/s for 200$ more if I go with S4s, vs 4.9 TH/s. I really don't mean to thread shit SP, and I hope that's not how this is coming across. I'm legitimately trying to justify an SP31 and I'm hoping for someone to weigh in. It won't take much to be honest, if only to treat myself to the crème de la crème of Bitcoin hardware. IMO the SP31 is overpriced - the SP20 however is in a sweet spot. The SP20 is now extremely competitive against bitmain S3+/C1 products, but the SP31 isnt very competitive against the S4. (granted, I expect a higher quality and denser product from the SP31 - but thats not worth the current premiums in price). considering that the SP31 is roughly 3x SP20 in speed, but costs an additional ~$800 (which effectively includes 2 inbuilt power supplies) it seems that a price adjustment is due, perhaps down to ~$2400 (the included PSUs and rackmounting ability are worth a $300 premium i think). I know with this new price Im pulling the trigger for a single SP20 over the weekend, and looking VERY closely at the farm pack - which is almost a 10% discount by not paying shipping costs). The only thing holding me back is that hey won't easily rackmount and they are going to be louder than the antminers i have right now. curious - can anyone compare the SP20 noise to the volume of a rockminer device fan (1.30A 120/38mm) as I found those fans to be too loud and eneded up swapping them all with fans from S1 units (1.0A fans - much lower pitch and volume). Having a "quiet" SP20 would be the deciding factor. |
|
|
|
|
just thought id throw some info in on my own experience.
modding resistors to measure 2.0-2.1 kohm (in circuit) results in 0.74-0.78V. This runs about 50GH/board when using 200Mhz and 65ms timeout. no wattage readings yet
|
|
|
|
Working with crew on a perspex case demo machine for hk bitcoin expo ...  sexy device. questions: 1) it requires an external controller? 2) it looks like multiple units can be daisy-chained. whats the limitations on this? (it would be nice to monitor and control 5TH+ under a single IP) |
|
|
|
The profit of this product on 27/11/2014, start on 15/11/2014  so ~0% so far this month. eh, markets do what the markets do. i'm sure the next bout of volatility will change that |
|
|
|
My favorite thing is that a bitfury system from last September continues to churn out bitcoins at ~1.2w/gh profitably. When I bought the device almost Noone expected it to be profitable past 6-9 months of use.
You must not pay that much for electricity if its still profitable. Are you out in Washington? $0.15/kwh IIRC the breakeven point for me right now is around 1.5w/GH - Ive been trying to sell off some of the older hardware to make sure i have only 1w/GH or better gear |
|
|
|
seems a bit high. 2.5% movement downwards is my best guess |
|
|
|
|
Show Posts
