Topic: [ANN][BLC] Blakecoin Blake-256 for GPU/FPGA With Merged Mined Pools Stable Net - page 166. (Read 409641 times)

I think that I got the pool working. I will do more tests today and if everything goes well I will publish it here

I'm interested, please let me know the details

Great

good luck. There have been several attempts to get one working. 

I started working on a pool for this coin, I will keep you updated.

1. the modified blake-256 algorithm is 2x-3x faster than sha-256d so you get more hashes which gives you more chances of finding the block hash when mining
2. the reward for mining only increases over time unlike most other coins Bitcoin/Litecoin
3. good over the long term and uses a more efficient hashing method than most other crypto coins  

  

Why should I mine BLC instead of other altcoins?
Or otherwise: What makes BLC better?

thanks! didn't know that
too bad my IGP is faster (58Mh/s) than my GTX  

any luck with a BLC pool?

looks like it is working, if it fails to solve the block that counts a stale in reaper when solo mining

see the image below as an example of reaper running normally

Hi I'm trying to get reaper working for solo mining
but somehow my stale is always 100%...


my configs look like:
in appdata blakecoin.conf:


blakecoin.conf in reaper.exe directory

reaper conf:

Reaper can only use my Intel IGP (somehow it fails at my GTX765M...)
any ideas what I'm doing wrong? I tried reaper on Litecoin and it works fine...

hey Bluedragon,
where can I find the white paper of Blakecoin?  Thanks

i sell 250blc -4- 0,1btc

Is there any exchange/places where I could buy some?

Good news everyone! (imagine Prof. Farnsworth voice )

I have managed to compile a version of cgminer for gpu mining blakecoin on Linux. So far I have only mined two blocks with it, blocks #29622 and #29634, they are still awaiting maturity. However they've already been confirmed several times so I am going to assume its valid. You can look at them here:
http://blakexplorer.dyndns.org:2750/b/CDFYi4Y7J
http://blakexplorer.dyndns.org:2750/b/C4EV2G1q4
Thanks for the block explorer

I have uploaded a binary and sources to mediafire. So far I have only compiled a 64bit version, but when I have time I will compile 32bit as well.

http://www.mediafire.com/download/ahaf69z3sdzeqou/blakecoin-cgminer-3.3.1-linux.tar.bz2

Many thanks, that's very generous of you. I'm taking a timeout to port my Litecoin miner to the CM1 (I only expect around 60kHash/sec so don't get too excited), but I'll keep plugging away at the blake builds as its just a matter of tweaking the build parameters and firing it off, checking back a few hours later to see if it has any chance of completing, then typically aborting it any trying again as the full build can take 12 hours or more.

I'll be looking at this.

I did send you some BLC for all the TLC you've given to FPGAs.

Thanks for the compliment. Tips gratefully accepted at BYEE2FvGv4L3yt5aEr8BQ73oWzXMu59Fwz (no need for an upfront tip, just mine blakecoin and decide what you'd like to contribute from proceeds).

The HW errors reported per-device for the Ztex are perhaps a little misleading as these are calculated as the devices are polled four times per second. So an error rate of just 1% (one in a 100) will give 4 * 3600 / 100 = 144 HW errors per hour! So I wouldn't worry too much about these, the automatic frequency management is likely to dial back the frequency to reduce the rate anyway (this may be what you're seeing with the grouping as the frequency steps up and down). I may just add a feature to completely disable this if its annoying, but it does seem to be working reasonably well and gives some protection against overheating (as evidenced by comments in the original code). Not that overheating should be an issue as the FPGA is only about 50% utilized (compare 80% for the bitcoin code), so it does run cooler. Now if I could only get that third core to route we'd really be cooking


I've been running builds all weekend, but haven't come up with anything faster yet. I have added a --cainsmore-clock command line parameter to cgminer to allow you to set the clock speed. Mine seems OK up to around 200MHz but you need to be very careful of HW errors as (unlike the ztex above) these are counted per share, so you want to keep these very low (a few are unavoidable due to timing issues as new work is sent). Just note the frequency where the errors start to climb rapidly, and dial it back slightly, you may want to set this on a per-device basis. See https://github.com/kramble/FPGA-Blakecoin-Miner/blob/master/cgminer/cgminer-3.1.1/README.txt for notes and a link to the new windows binary.

PS It you do set the clock speed too high, you may find the devices are not detected the next time you run cgminer. Don't panic! Just run it again with a lower clock speed, perhaps twice, and the devices should be detected OK. If you have not yet already done so, you may want to flash the bitstream into ROM so it survives a power cycle as this is the ultimate reset.

Any progress on pool?

Kramble, you're a hero for getting this working. One overvolted Ztex 1.15y hashing away happily at 204 MHz now (could probably push it further) - no sign of magic smoke yet . Do you have a blakecoin tipjar?

What kind of hardware error rate do you feel would be reasonable for this coin? I've got around 300 HW errors on most of the chips so far (although one has substantially fewer than the others for some reason) after 9 hours of mining. The HW errors seem to come in groups rather than in a regular pattern. None of the heatsinks feel particularly hot (they actually seem to be a little cooler than when bitcoin mining).

Great, thanks! Let me know if I can support you with the setup.

Hi All, has anybody had any luck compiling the cgminer for Blakecoin in linux from the source linked in the OP? I'm getting the following error:

Edit: Just to clarify, I'm talking about the gpu-mining version of cgminer, not kramble's FPGA version (which works fine).


I'm running ubuntu 12.04 LTS 64 bit and have all of the dependencies to compile the regular cgminer. With the Blakecoin version I had to copy the contents of compat/jansson/.deps/ from a "normal" version of cgminer to the Blakecoin version to get rid of the windows-specific file locations in these files (I'm afraid that I was too lazy to edit all of them by hand). Is this the likely source of the problem? I realise that I'm probably missing something completely obvious, but would certainly appreciate a hand.