Look at cklolivas(cgminer dev) post in this thread. Goes into great detail about this.

If you want, you can change the queue in /etc/init.d/cgminer

I use 32 and it seems to be a good compromise.  Some people use 1, others have kept 4096 - doesn't seem to have a huge effect, but can save you some CPU cycles if you make it lower.

It's funny, I've never considered such a hosted service, and I just took a closer look, but the numbers don't add up.

Buying a S3 from Bitmaintech is ~ 390 USD, at GAW Miners it's 499.95 USD. Their hosting service is 0.9 USD a day after a free month versus 1.22 USD on engery costs for a day of running the S3 here at my place (calculated with 340W instead of the indicated 366W on the GAW Miner S3 product page). So the two prices including running costs reach equilibrium after ~ 256 days at a total of around 702 USD, which is way beyond the point to which you can make a profit. That calculation is done without VAT because it would not apply to my (one of the advantages having a company at hands), with our VAT of 20% for the hardware on import and on energy the picture would be different of course.

Calculated the other way around: The hosting service of GAW Miners for 0.9 USD equals a price of 0.11 USD per kWh. Combine that with the premium on the hardware purchase and you will never reach ROI anyway.

However you turn it, it will be a loss.

Combine that with the premium on the hardware purchase and you will never reach ROI anyway.

However you turn it, it will be a loss.

Do any of you here know why the queue on the config is set so high? Is there a big gain from setting it to a lower number or why did they program 4096 as the queue. The discarded amounts are so high...

Do any of you here know why the queue on the config is set so high? Is there a big gain from setting it to a lower number or why did they program 4096 as the queue. The discarded amounts are so high...

I have been running both of my Ant's OC'd to 500 GH/s and 490 GH/s and both temperatures are well within normal ranges and all chips are stable and hashing with no X's.  One of them has been running a few hours as if you look up a post or two you'll see where we came up with the newer smaller incremental frequency steps, and that is the one running at 490 GH/s up from the 237.5 clock which was in the 470-480 GH/s range, while the other has been running for 72 hours at 500 GH/s.  

Highest temperature reached by any miner is 44C, while averaging 40-43 C in use, so I would say replacing the thermal compound is unnecessary.

That being said, if you do decide to go ahead and upgrade/replace the thermal compound, do not use any compound with Silver in it as it is conductive, and make absolutely sure that whatever brand you decide to go with DOES NOT CONDUCT ELECTRICITY - YOU WILL FRY YOUR ANTMINER!!!!!

Hahaha, no no definitely didn't remount the plate upside down. However, if my heatsinks were machined/cast anywhere NEAR as well as the ones in that unit I probably wouldn't have any problems.

Have the same question... After I wrote my message to them, couple hours later the "message" feature just dissapeared from their website.

CX500 is $30 AR not $50. I've bought those and would buy again if I needed a good cheap small PSU. Just like I'd buy EVGA 1300 again if I needed a good cheap powerful PSU. I don't know what Antminer warranty has to do with anything, or living conditions.

"Doing it wrong" are your words not mine. See, I don't tell how you should run things, but you seem to be so worried that others are doing it not according to your rules that you resort to calling them "dumb" etc. I will not wish I had spent $20 more, or $40 more as the case might be. I will be happy that I can afford a spare and zero downtime rather than losing revenue while RMAing a "high quality" PSU (yes, those break too). To each their own.

Your explanation on "how power supplies work" does not make any sense at all.  I think it only leads to more confusion.  How did you come up with 81.5% load on the CX600 with an S3?  Please elaborate.  What "added power draw from the wall" are you referring to?  It is the actual power draw at the wall that computations should be extrapolated from because it is the only concrete number we can get out of the S3 or any device for that matter.  Let us indulge ourselves:

AC power draw at the wall * PSU efficiency = DC power draw on +12V rail of PSU
DC power draw on +12V rail of PSU / max load on +12V rail of PSU * = % of load exerted on +12V rail of PSU

Therefore (assuming an OC'd S3 draws 450WAC at the wall using a CX600):

450WAC * 85% = 382.5WDC
382.5WDC / 522WDC * 100 = 73.28%

I am aware that the CX600 label states 46A on the +12V rail, hence you came up with 552W (46A * 12V).  Confusingly, the same label also states that the maximum +12V rail wattage is 522W.  Which one is a typo?  Should it have been 43.5A (522W / 12V) instead of 46A?  Anyway, I tend to be conservative and go with the lower rating and be on safe side which is why I used 522W on the above computation that resulted in a 73.28% load.  However, even with a maximum of 552W on the +12V rail as you asserted, the load would then be much lower at 69.29% (382.5WDC / 552WDC * 100).  So, where was your "81.5%load without the added power draw from the wall minus efficiency of the power supply" derived from?

I reiterate that the Corsair CX500 (also with only two PCIe power connectors) is able to accommodate an S3 OC'd to 250M without any issues whatsoever let alone a CX600.  I am not just saying this for the heck of it; I and at least another member in here actually did it.  Please read here.

Spending $400 on an Ant and $200 on a PSU does not make sense for a small scale miner. The CX or whatever it is you're so unhappy about works fine with a single Ant and that's all there is to it. Not everyone wants or needs 100A on a single rail. Smaller PSU at peak efficiency is better than underutilized large one. I started with cheap-after-rebate PSUs and it was the best decision I could make back then. Dual PSUs for GPU rigs when everyone was buying LEPAs at ungodly prices, and I could afford spares too. Somebody doing something different than you is not "doing it wrong".

I am running an overclocked s3 on a cx500 right now. 14 days and its fine...so....

I ONLY use this to power my single S3:

http://www.newegg.com/Product/Product.aspx?gclid=Cj0KEQjwx4yfBRCt2rrAs-P5vtkBEiQAOdFXbfyYxK8BNRjcG4HINOdFw3jsGBwqz5piwisbeBuPMD8aArEq8P8HAQ&Item=N82E16817139079&nm_mc=KNC-GoogleAdwords&cm_mmc=KNC-GoogleAdwords-_-pla-_-Power+Supplies-_-N82E16817139079&ef_id=U54GbwAABJyCojXI:20140808042436:s

If i were to overclock i would need something more powerful like plugging it directly into the sun or.... To the sub basement!

OK...let's make this simple.  A CX500 will absolutely, positively power an S3 at stock clock ...I've been running three batch 5s for two days now - avg 438Gh/s on the CX500's (I used 5 CX500s to run five S1s previousl - they draw nearly the same AMPs as an S3)  The CX 500's are selling at Newegg.com for $29.99 after $20.00 rebate right now.

One PSU per miner means if one PSU dies, only one miner dies (not that any of my Corsair, EVGA or any other name brand PSUs have died since I started seriously mining about 5 months ago).

I don't get why y'all are buying these huge, expensive PSUs to support multiple miners...what's the point? And the next person that tells me they are buying gold or platinum PSUs because they are "more efficient", better be able to justify the additional expense vs. actual electric cost savings over 80 or so days...show me the math that proves the "more efficient" PSUs actually ROI better for our very specific scenario.

the temps are decent so  other then the fan speed meter
 your miners are great  discarded shares are a meaningless stat.


the math for miner 1   hw/diffa + diffr + hw =  83/4965376+11776+83 or 83/4977235  = .000016675 x 100 = .00166 percent error rate.

accepted shares are 19396 on miner one

accepted shares are  19617 on miner two  close to the same for the same time   and the av gh = 440 and 438  which bears out to your accepted share number


the 5 second speeds are bs they move all over don't worry about them.

  for shits and giggles do the error rate for miner 2 

it will be about  .0333 percent  that is 99.67 good.       and 98.00 is pretty much passing.

Why do you say that?  Have you actually run an S3 with a CX600?

Even if an OC'd S3 draws 450w at the wall, it only represents a 73% load on the CX600.

Must be a change on the site.  The message options are gone for me too.  I had attempted to communicate with them via that method before, but it wasn't until I sent an email instead that I got anything back.