Bitcoin Forum
June 30, 2024, 09:11:59 PM *
News: Latest Bitcoin Core release: 27.0 [Torrent]
 
  Home Help Search Login Register More  
  Show Posts
Pages: [1]
1  Bitcoin / Mining support / Re: BITMAIN Antminer S3 support and OverClocking thread on: August 27, 2014, 06:15:58 AM
Fair enough, but it's my understanding that modern network adaptors carry out the crossover function automatically.  However, I'm not sure that all of these adaptors will autosense if a crossover cable is subsequently connected.

Yes, it's called Auto MDI-X and as long as at least one side of the link has the capability, it doesn't matter whether you have a crossover cable or a straight-through cable, the interface will sense it and act accordingly. The speed and duplex need to be Auto.
2  Bitcoin / Mining support / Re: BITMAIN Antminer S3 support and OverClocking thread on: August 24, 2014, 09:52:39 AM
Bit layout for the freq_value is as follow:



OD can have values between 0 and 3
R can have values between 0 and 31
F can have values between 0 and 127
BS can have only 0 or 1

From page 6 of BM1382 spec (https://www.bitmaintech.com/files/download/BM1382_Datasheet_v3.0.pdf) you get:

NR = R + 1
NF = 2 * (F + 1)
NO value from OD value (see Table 4 from same spec page)

Fref = Fin / NR
Fvco = Fout * NO
Fout = Fin * NF / (NR*NO)

So Fout is "chip_freq" and can be calculated based on valued from above ("OD", "R" and "F"). BS select "high-band" or "low-band" (see page 7 from BM1382 spec for valid freq values for each band).

chip_freq = 50 * ( F + 1 ) / ( ( R + 1 ) * POWER( 2 , OD ) )

For timeout I used following formula:
timeout = 4000 / chip_freq

I suggest to calculate all 3 Fref, Fvco and Fout and make sure all 3 value are in range described on page 7 for the band specified by "BS" value.

Hi folks, given the above, I have thus calculated EVERY POSSIBLE valid frequency and register value. I don't know about the time out, so for that I just used JakeTri's timeout formula to get the time out and then took the integer value:

Assuming:                                 
Fin =   25                              
                                 
Fref   Fvco      Chip Speed      Register Value            TimeOut
12.5   800      100.0000      0x   0   F   8   6      40
25.0   800      100.0000      0x   0   7   8   3      40
12.5   825      103.1250      0x   1   0   0   6      38
12.5   850      106.2500      0x   1   0   8   6      37
25.0   850      106.2500      0x   0   8   0   3      37
12.5   875      109.3750      0x   1   1   0   6      36
12.5   900      112.5000      0x   1   1   8   6      35
25.0   900      112.5000      0x   0   8   8   3      35
12.5   925      115.6250      0x   1   2   0   6      34
12.5   950      118.7500      0x   1   2   8   6      33
25.0   950      118.7500      0x   0   9   0   3      33
12.5   975      121.8750      0x   1   3   0   6      32
12.5   1000      125.0000      0x   1   3   8   6      32
25.0   1000      125.0000      0x   0   9   8   3      32
12.5   1025      128.1250      0x   1   4   0   6      31
12.5   1050      131.2500      0x   1   4   8   6      30
25.0   1050      131.2500      0x   0   A   0   3      30
12.5   1075      134.3750      0x   1   5   0   6      29
12.5   1100      137.5000      0x   1   5   8   6      29
25.0   1100      137.5000      0x   0   A   8   3      29
12.5   1125      140.6250      0x   1   6   0   6      28
12.5   1150      143.7500      0x   1   6   8   6      27
25.0   1150      143.7500      0x   0   B   0   3      27
12.5   1175      146.8750      0x   1   7   0   6      27
12.5   1200      150.0000      0x   1   7   8   6      26
25.0   1200      150.0000      0x   0   B   8   3      26
12.5   1225      153.1250      0x   1   8   0   6      26
12.5   1250      156.2500      0x   1   8   8   6      25
25.0   1250      156.2500      0x   0   C   0   3      25
12.5   1275      159.3750      0x   1   9   0   6      25
12.5   1300      162.5000      0x   1   9   8   6      24
25.0   1300      162.5000      0x   0   C   8   3      24
12.5   1325      165.6250      0x   1   A   0   6      24
12.5   1350      168.7500      0x   1   A   8   6      23
25.0   1350      168.7500      0x   0   D   0   3      23
12.5   1375      171.8750      0x   1   B   0   6      23
12.5   1400      175.0000      0x   1   B   8   6      22
25.0   1400      175.0000      0x   0   D   8   3      22
12.5   1425      178.1250      0x   1   C   0   6      22
12.5   1450      181.2500      0x   1   C   8   6      22
25.0   1450      181.2500      0x   0   E   0   3      22
12.5   1475      184.3750      0x   1   D   0   6      21
12.5   1500      187.5000      0x   1   D   8   6      21
25.0   1500      187.5000      0x   0   E   8   3      21
12.5   1525      190.6250      0x   1   E   0   6      20
12.5   1550      193.7500      0x   1   E   8   6      20
25.0   1550      193.7500      0x   0   F   0   3      20
12.5   1575      196.8750      0x   1   F   0   6      20
12.5   800      200.0000      0x   0   F   8   4      20
25.0   800      200.0000      0x   0   7   8   2      20
12.5   1600      200.0000      0x   1   F   8   6      20
25.0   1600      200.0000      0x   0   F   8   3      20
12.5   825      206.2500      0x   1   0   0   4      19
25.0   1650      206.2500      0x   5   0   0   3      19
12.5   850      212.5000      0x   1   0   8   4      18
25.0   850      212.5000      0x   0   8   0   2      18
25.0   1700      212.5000      0x   5   0   8   3      18
12.5   875      218.7500      0x   1   1   0   4      18
25.0   1750      218.7500      0x   5   1   0   3      18
12.5   900      225.0000      0x   1   1   8   4      17
25.0   900      225.0000      0x   0   8   8   2      17
25.0   1800      225.0000      0x   5   1   8   3      17
12.5   925      231.2500      0x   1   2   0   4      17
25.0   1850      231.2500      0x   5   2   0   3      17
12.5   950      237.5000      0x   1   2   8   4      16
25.0   950      237.5000      0x   0   9   0   2      16
25.0   1900      237.5000      0x   5   2   8   3      16
12.5   975      243.7500      0x   1   3   0   4      16
25.0   1950      243.7500      0x   5   3   0   3      16
12.5   1000      250.0000      0x   1   3   8   4      16
25.0   1000      250.0000      0x   0   9   8   2      16
25.0   2000      250.0000      0x   5   3   8   3      16
12.5   1025      256.2500      0x   1   4   0   4      15
25.0   2050      256.2500      0x   5   4   0   3      15
12.5   1050      262.5000      0x   1   4   8   4      15
25.0   1050      262.5000      0x   0   A   0   2      15
25.0   2100      262.5000      0x   5   4   8   3      15
12.5   1075      268.7500      0x   1   5   0   4      14
25.0   2150      268.7500      0x   5   5   0   3      14
12.5   1100      275.0000      0x   1   5   8   4      14
25.0   1100      275.0000      0x   0   A   8   2      14
25.0   2200      275.0000      0x   5   5   8   3      14
12.5   1125      281.2500      0x   1   6   0   4      14
25.0   2250      281.2500      0x   5   6   0   3      14
12.5   1150      287.5000      0x   1   6   8   4      13
25.0   1150      287.5000      0x   0   B   0   2      13
25.0   2300      287.5000      0x   5   6   8   3      13
12.5   1175      293.7500      0x   1   7   0   4      13
25.0   2350      293.7500      0x   5   7   0   3      13
12.5   1200      300.0000      0x   1   7   8   4      13
25.0   1200      300.0000      0x   0   B   8   2      13
25.0   2400      300.0000      0x   5   7   8   3      13
12.5   1225      306.2500      0x   1   8   0   4      13
25.0   2450      306.2500      0x   5   8   0   3      13
12.5   1250      312.5000      0x   1   8   8   4      12
25.0   1250      312.5000      0x   0   C   0   2      12
25.0   2500      312.5000      0x   5   8   8   3      12
12.5   1275      318.7500      0x   1   9   0   4      12
25.0   2550      318.7500      0x   5   9   0   3      12
12.5   1300      325.0000      0x   1   9   8   4      12
25.0   1300      325.0000      0x   0   C   8   2      12
25.0   2600      325.0000      0x   5   9   8   3      12
12.5   1325      331.2500      0x   1   A   0   4      12
25.0   2650      331.2500      0x   5   A   0   3      12
12.5   1350      337.5000      0x   1   A   8   4      11
25.0   1350      337.5000      0x   0   D   0   2      11
25.0   2700      337.5000      0x   5   A   8   3      11
12.5   1375      343.7500      0x   1   B   0   4      11
25.0   2750      343.7500      0x   5   B   0   3      11
12.5   1400      350.0000      0x   1   B   8   4      11
25.0   1400      350.0000      0x   0   D   8   2      11
25.0   2800      350.0000      0x   5   B   8   3      11
12.5   1425      356.2500      0x   1   C   0   4      11
25.0   2850      356.2500      0x   5   C   0   3      11
12.5   1450      362.5000      0x   1   C   8   4      11
25.0   1450      362.5000      0x   0   E   0   2      11
25.0   2900      362.5000      0x   5   C   8   3      11
12.5   1475      368.7500      0x   1   D   0   4      10
25.0   2950      368.7500      0x   5   D   0   3      10
12.5   1500      375.0000      0x   1   D   8   4      10
25.0   1500      375.0000      0x   0   E   8   2      10
25.0   3000      375.0000      0x   5   D   8   3      10
12.5   1525      381.2500      0x   1   E   0   4      10
12.5   1550      387.5000      0x   1   E   8   4      10
25.0   1550      387.5000      0x   0   F   0   2      10
12.5   1575      393.7500      0x   1   F   0   4      10
12.5   800      400.0000      0x   0   F   8   2      10
25.0   800      400.0000      0x   0   7   8   1      10
12.5   1600      400.0000      0x   1   F   8   4      10
25.0   1600      400.0000      0x   0   F   8   2      10

For the above paste, in the interest of space I only presented speeds up to 400.

Note that if the frequency does not show on this list, it does NOT exist, (except for ones above 400mhz) as these values were calculated from EVERY possible binary value of the register, then filtered for valid values based on the criteria in the datasheet.

I do not know the implications of the Fref and Fvco values, so I sorted everything from lowest to highest. For a clock speed of 200mhz is it best to use a Fvco of 800 or 1600? furthermore, within those results, is it best for Fref to be 12.5 or 25? Given BitMain's example in the datasheet, evidence would suggest they want us sticking in the low range bank, with Fref=25. If there's any good reason for that then that reduces the possibilities to:

Fvco   Chip Speed      Register Value            TimeOut
800   100.0000      0x   0   7   8   3      40
850   106.2500      0x   0   8   0   3      37
900   112.5000      0x   0   8   8   3      35
950   118.7500      0x   0   9   0   3      33
1000   125.0000      0x   0   9   8   3      32
1050   131.2500      0x   0   A   0   3      30
1100   137.5000      0x   0   A   8   3      29
1150   143.7500      0x   0   B   0   3      27
1200   150.0000      0x   0   B   8   3      26
1250   156.2500      0x   0   C   0   3      25
1300   162.5000      0x   0   C   8   3      24
1350   168.7500      0x   0   D   0   3      23
1400   175.0000      0x   0   D   8   3      22
1450   181.2500      0x   0   E   0   3      22
1500   187.5000      0x   0   E   8   3      21
1550   193.7500      0x   0   F   0   3      20
800   200.0000      0x   0   7   8   2      20
1600   200.0000      0x   0   F   8   3      20
850   212.5000      0x   0   8   0   2      18
900   225.0000      0x   0   8   8   2      17
950   237.5000      0x   0   9   0   2      16
1000   250.0000      0x   0   9   8   2      16
1050   262.5000      0x   0   A   0   2      15
1100   275.0000      0x   0   A   8   2      14
1150   287.5000      0x   0   B   0   2      13
1200   300.0000      0x   0   B   8   2      13
1250   312.5000      0x   0   C   0   2      12
1300   325.0000      0x   0   C   8   2      12
1350   337.5000      0x   0   D   0   2      11
1400   350.0000      0x   0   D   8   2      11
1450   362.5000      0x   0   E   0   2      11
1500   375.0000      0x   0   E   8   2      10
1550   387.5000      0x   0   F   0   2      10
800   400.0000      0x   0   7   8   1      10
1600   400.0000      0x   0   F   8   2      10

But i imagine people are willing to try other than just these values, since folks are asking for speeds between 237.5 and 250 (the only one is 243.75 but with two options) and i assume 243.75 has been tried/tested by now?

I can post my excel file so people can check my work, and/or if people want to see valid chip speeds up to 3000mhz (i'm pretty sure you will NOT be able to run at 3000mhz! so don't get any bright ideas) Smiley

If any of this proves useful, feel free to donate BTC at 1Lq8HMwSCR65RpApfYmWpCrP5wobsbcLDE
3  Bitcoin / Bitcoin Discussion / Re: Is it just me, or does this chart look too uniform? on: August 01, 2014, 07:28:18 AM
I believe that at some unknown point in the future, largely (if not entirely) due to reward halving, mining will cease to be a profit-based bitcoin-generating activity, and will become a cost-based transaction-processing activity. I fully expect motivations to change drastically when that time comes. By then, I won't be surprised to see a different pattern.
4  Bitcoin / Pools / Re: IS GHASH.IO Cheating Miners? Paying LESS THAN 50% of expected? on: August 01, 2014, 07:11:14 AM
I've been mining on ghash.io since i sold off my ghs from cex.io to buy into my own asic hardware. At about 1.8 TH/s i've been pulling down about .034 to .036 or so per day. You said "best conditions" and by your numbers i should be getting 0.048312 per day. So if i assume NOT "best conditions" then should i think .036 instead of 0.048312 is reasonable?

also, something that might affect it - i've been mining on cex/ghash for about 8 months, and plotting/graphing the entirety of it for analysis purposes. The thing that i have found to be very intriguing is that when the difficulty changes, it really doesn't seem to have much affect on my mining production. like literally at all. with a slow decay in production over time, it's almost as though they linearize the decay that happens due to difficulty drop, and by that i mean make it into a constant decay rather than a stepped change like difficulty by adjusting based on the expected difficulty change which is amazingly accurate, especially the closer it gets to the actual diff change. my theory is that this is how they make their profit since they have a 0% pool fee. and honestly, i kinda like it... i'm ok with them making their profit in an innovative way outside the box of what everyone would normally assume.

if this is true, then it means what you should be getting vs. what you are actually getting would be different immediately before and immediately after a diff change.

whether i could and/or should be getting that extra 0.012312 per day, well, i have to do some analysis on that. when i'm not focused on building out my mining op i'll spend some time deciding whether to go to another pool and which one. but i like the fact that ghash is tied to cex, and the reason i got into cex was not specifically for the cloud mining but rather the commodity trading of ghs. regardless of the fact that their ghs is more expensive than hardware, regardless of how much i had, i was able to nearly double and quadruple it by seizing lucrative trading opportunities. because of that i was able to make thousands of $USD worth of bitcoin in profit, and sell out to by 20 s3's this week.
5  Other / Off-topic / password security on: July 31, 2014, 02:54:30 AM
Maybe I'm just old-fashioned.



I'm guessing that at some point in technology history, someone must have thought "no, we don't need to do that..."

why is this so common? It seems I only started seeing this when I got started with bitcoin.
Pages: [1]
Powered by MySQL Powered by PHP Powered by SMF 1.1.19 | SMF © 2006-2009, Simple Machines Valid XHTML 1.0! Valid CSS!