Topic: [ANN] sgminer v5 - optimized X11/X13/NeoScrypt/Lyra2RE/etc. kernel-switch miner - page 86. (Read 877846 times)

That's not linux, sorry, I've messed everything up. Looked at .a dump and talked about dl, but that's windows PE import library, by mingw or cygwin. No actual code there. Well, you could have some fun with IDA and actual .dll, a good way to kill a week or two

Looks like usual import library. IMHO those 8 bytes gets into executable at link time (or is it just empty dataless placeholder?), at load time dl fills it with resolved function address, at execution time code performs indirect jump by this address (dl patches not JMP code  in read-only text segment, but read-write data, so library image could have different base address for each process and still be shared). I thought that gnu linker works directly with .so files but nothing prevents it from this scheme. Well, in PE there are fancy things like forwarded exports but sure that's not the case.
EDIT: yes, something strange with sections, I'll try to feed this library to IDA, it's good at such things

Oh, right, I used yacminer, too.


I was wondering how you got 500kh/s on 280X, as that's about what I've got.

Hi Wolf0,

Any chance you would compile the sgminer for linux 64bit?

Super!!!! thank you so much !!!
280x earned by 260mh\c   ,  270x ... and not given the addition of speed ..... what am I doing wrong?

Isn't it in SG5? I remember boosting the speed of it for a client.

I've extracted the libs and rebuilt the (modified) kernel with those, it became 104 bytes fatter and hashrate went from 377 to 374 on a 290x@936 MHz; don't know how it works on your stock kernel, though...
Still want the files?

Any way to run on linux? Extracting libopencl from Ubuntu vivid package?

Yes I tried everything, with xI of 1024, 2048 and worksize of 128.

Will try the 14.12 cataylst version.

Don't know what I am doing wrong but I get the old 60Mhash/s rate for my 270x

Yes I guess you are correct

According to:
https://bitinfocharts.com/darkcoin/


Only 2584 DRK were mined in the last 24 hours and the block reward for miners is only 2.80 or so which is only around $7000

Right now the average hash is 100Ghash/s which if everybody is mining with an 280X or equivalent leads to 16,666 GPUs or so. So only $0.40 per day per 280X equivalent video card. Very sad considering back in January 2014 an 280X was pulling in $10/day.

It costs money to make the silicon design and you have to insure the yield on the wafer is high enough to manufacture ASICs equipment at a profit; there is 10 time less DRK each day compared to LTC, plus masternodes take another 3.75 away.  Therefore, you are looking at 14 times lower volume of coins available compared to LTC. Plus, each year there there is 7% reduction. Furthermore, DRK's developers has done 3-4 hard forks in the last year, so Wolf0, it will need a mighty high price to justify the inherit higher risks of engaging in process of producing ASIC equipment.  

Well, its very strategic on GPU's, because you can resell GPU's for decent money at anytime................plus the Litecoin transition out of GPU mining left the GPU alt coins over saturated with GPU mining hardware..............it has taken 6 months to replace GPU Litecoin with another alt coin (DRK)................profitability will rise this year and next..............Darkcoin is limited to 21 million coins ..........it will last a lot longer then GPU Litecoin, it will need to pass $28 per DRK before anyone considers making any ASIC mining hardware for it.........it has 10x lower daily output then LTC, plus an annual reduction of 7%..........so it could be many years of profitable GPU mining on DRK........allowing larger miners to make successful long term investments.

However, by 2018 and 2019 there will be more alt coins reaching maturity for GPU mining (1st Block Halvings or Reductions)

Here is top four, which varies with profit switching multi-pools (X11 averages are):
1) 124GHs (Peak is 144Ghs) Darkcoin Block Reduction 7% per annum 74%-----$66 million of GPU hardware.
2) 10.5GHs Vertcoin Block Halving 2018 6.3% -------$0.56 million of GPU hardware
3) 10GHs Startcoin Block Halving 2016 6%-----------$0.53 million of GPU hardware
4) 8.5GHs Feathecoin Block Halving 2019 5%-----------$0.45 million of GPU hardware

Other GPU crypto-currencies with weak mining support equal about 13Ghs 8%------$0.69 million.  Noteworthy coin with weaker mining support is Spreadcoin, with less then 20 million coins and a talented developer crew.

Total GPU hash power is approximately: 167GHs

You can see the headache left when GPU Litecoin ended unexpectedly; you've had $68 million of GPU mining hardware mining alt coins only worth $25 million dollars. This time around the larger miners won't make the same mistakes and nor will everyone else

Should be Lyra2RE.

So 3 years to break-even with those 750ti GPUs?

WOW. Much worse then buying an BTC Asic.

I've based that on publicly released open source software or leaked mods widely available, we all know that their are many mods to boost hashing power on Neoscrypt, Lyra2, X13 and so on.

I'll repost it from February, with a number of improvements, which Thomas Paine famously recommended doing in The Age of Reason. I'm not incline to causing such controversies myself and I just apply it to what I wrote in February.


American are saying they're paying 14 cent for electricity and Chinese are saying they're paying 10 cent for electricity. Finally, in Iceland electricity cost is around 9 cent; not the rumoured 3-5 cents. Thirdly, I have updated figures for the current Fiat Currency prices.

"As a digression, for anyone investing in DRK.

The current ROI on a 4 card 750TI PC ($880) hashing machine on Darkcoin is:

POW 62.5% mining versus Masternode deduction of 37.2% on block reward, not everything goes through masternodes, effective block reward is 3.41DRK for miners.

China 10 cent for electricity and BTC price is $260

Average difficulty for 3 year breakeven on investment $880 = (Difficulty at 3600) 0.023BTC or $5.98.

Should you factor in 20% per year profit margin for the miner, yes, it is a miserly profit margin

0.0276BTC or $7.18.

These calculations won't really bite home until 2016, when old Litecoin GPU machines (to many) fail in large numbers, miners will either need to increase their selling price or quit GPU mining altogether (DRK supports up to 70% of GPU cards mining today). The latter would monopolize DRK into miners (who know what the are doing) and save in DRK. Many will refuse to sell their DRK on the exchanges until Darkcoin owners realize their long-term ambitions. This will push the price up this year and next year.  

Equally, there is 7% block reduction per annum and these calculations should be adjusted each year. Current Difficulty spike today is 4200, but this temporary spike disappears when miners figure out this lesson: save in DRK whenever it is undervalued, just redo the numbers yourself.  

Darkcoin has about 70% of GPU hash at moment, it fluctuates between 50%-70%. It is 124GHs at the moment.

Or, it is the equivalent of 20,667 280X's or 38,750 750TI's. 90% of GPU's are AMD, so that nvidia comparison is hypothetical

Besides, Darkcoin is still good for rises in price up to $5.18 this year, I did a post looking into the degree of undervaluation on Darkcoin at the beginning of February based on GPU investments from the large mining groups and master-node investment (1000DRK per masternode): I bought extra DRK.  

Yes but now its the #3 minable coin. It should of made GPU more profitable again with the large market cap.

I guess there are millions of GPUs out there hashing away.

Mining x11, gpu0 hashes but no submitted shares and no hardware errors, gpu1 is ok.
Tried switching the two gpus but still gpu0 doesn't work (i.e. the gpu in the top pci slot).
I'm using wolf0's binaries.
Mining other algos works fine.
Also tried older sgminer versions but no change.
Any clues?