CCminer(SP-MOD) Modded GPU kernels.

[pallas]

Never had any issue with alexis mod on linux, 40gb total gpu ram and 4gb system ram.

[1713408086]

1713408086

[1713408086]

1713408086

[1713408086]

1713408086

[1713408086]

1713408086

[anatolikostis]

just increase swap-file (up to 20Gb) or memory (up to 16Gb)

I need a refund because I am too stupid to use this software properly.

So, there are no answers from you:
why does myr-gr disappear from start post?
why does 1070 revert to 1070 OC?
why does my 6*1080 rig work propertly?

[sp_]

why does myr-gr disappear from start post?

It is still there. My mod can mine groestl, diamond groestl and myriad groestl. All the kernels are are faster than ccminer 1.8.4 (wich the mod was based on) Myr-gr is around 15% faster than ccminer 1.8.4 same clocks and settings.

I have now changed it again to include diamond groestl:

9. 0.05BTC groestl sp-mod (1 release) Groestl/Diamond groestl speed: 40MHASH gtx 1070, Myriad groestl speed: (75MHASH gtx 1070, 40MHASH gtx 1060 3gb)

why does 1070 revert to 1070 OC?

There are many different models, with many different power configurations and clocksettings. I can reach this speed on my cards without any problems.

why does my 6*1080 rig work propertly?

You use windows 7.

Windows 8 and 10 could have problems. The workaround of adding alot of virtual mem might solve the problem.

[alexis78]

just increase swap-file (up to 20Gb) or memory (up to 16Gb)

I need a refund because I am too stupid to use this software properly.

If you paid for public source codes then yes, you need a refund.... (If you know what i mean )

edit: Oh, and there is no way you get 5% reject on pools. The code has been checked multiple times to eliminate such issues. If you find a bug in the source, i'm open to discuss it with you in public.

[sp_]

Do you submit stale shares?

I get alot rejects with "Job id not found"

Try to mine skein today on the yiimp.ccminer.org

[alexis78]

Do you submit stale shares?

I get alot rejects with "Job id not found"

Try to mine skein today on the yiimp.ccminer.org

submitting stale shares does not lower your actual hashrate since the kernels are restarting when a new job arrives. High intensity on the miner though is increasing the probability to submit a stale share since computation and the job fetching are asynchronous tasks.

You can choose to submit that stale share (as your public source is doing too) or not to submit it (as epsy's source is doing)
by not submitting it you wont get 5% rejects (or whatever percentage someone gets) but that does not mean that you will save your computational power for real work.

[sp_]

Well Yiimp is not working today, the stratum connection is lost all the time, lots of rejects and reconnects. Skein has fast blocktime, and the default intensities in your mod seems to be too high.

With the high intensity the miner doesn't exit properly and when a block is found, the gpu is working on  the old data instead of switching instantly.

Is the restart flag working properly?

} while (!work_restart[thr_id].restart && (uint64_t)max_nonce > (uint64_t)throughput + (uint64_t)pdata[19]);

[alexis78]

...yes the restart is fine. if it was not working then you would have the same 5% rejected issues on my fork you took and added a +1.6% fee increase on lbry

It's not skein that has fast block times, it's dgb, unless aur changed their block times too

[bensam1231]

Funny, you currently make more mining Ethereum with a 1070 or dual mining eth+sia then you do mining Equihash or Lbry... Just another example of how terrible a value proposition Nvidia is right now. Wonder when the tards mining Zcash will get the memo.

ha-ha, please, keep mining eth or zcash or lbry!!!

What makes you think I am? I'm talking about the 233,000 1070s on Equihash. You silly goose and your niche coins thinking it pertains to all Nvidia mining.

[sidog]

I'm trying to get the alexis miner to work on my 1070 Zotac Amp Extreme, mining Skein.

Keeping getting GPU#0:invalid argument - even in offline, benchmarking mode.

Anyone have any suggestions?

[sp_]

I managed to push the gtx 1070 above 600MHASH..

The GPL shitcode from alexis is down at 550MHASH

Skein SP-mod #2

[rednoW]

I managed to push the gtx 1070 above 600MHASH..

The GPL shitcode from alexis is down at 550MHASH

what clock and consumption?
My 1070 are doing 575@140w with alexis

[sp_]

Not done yet.

The 750ti does 115MHASH
1060 3gb 360MHASH

[Amph]

what about other gpu? my 1080ti can do 1gh but i think it should be more

[rednoW]

what about other gpu? my 1080ti can do 1gh but i think it should be more

Palit js 1080ti 1050mhs@250w, all that more 4mhs/watt is good

[Kompik]

so, if I show you 6*1080rig with no problem, will you return/refund in twice?

I have problems with 6 gtx 1070 and a celeron 1840 cpu. 2 cards are simply not starting, and never hash. The problem happens with all the algos in the alexis mod, and has been reported by a great number of users. It doesn't happen every time you run it though. Only sometimes.

There is a solution though, you can use --cuda-schedule 0 and force the cards to always start.

Also 360 skein on a 1060 3GB looks too good to be true? 45% price of the 1070 and 60% of the hash?

[antantti]

Guys, get your ethminers ready...

Tried alexis with 970, 47 on myr-gr and 335 on skein.

[rednoW]

Also 360 skein on a 1060 3GB looks too good to be true? 45% price of the 1070 and 60% of the hash?

I think it is some fake numbers again to receive 0.05btc "donation" from some fool ))

[sp_]

I don't need donations. I am harvesting.

here is the mod in action with lower tdp. been running steady for 12 hours 342.8MHASH.

[rednoW]

I don't need donations. I am harvesting.
here is the mod in action with lower tdp. been running steady for 12 hours 342.8MHASH.

Here gtx1080ti is the same price as 3xgtx1060, but with 1 gtx1080ti you'll get more hash with less power and higher density ))
1060 is shit ))