Topic: EWBF's CUDA Zcash miner - page 250. (Read 2164327 times)

Happy New Years folks...

W7 and W8 have a different WDDM version then W10, it's entirely possible to get better hashrate in W8.1, that's the reason I didn't upgrade my rigs to W10. I had one on W10 and eventually reimaged it back to W8.1.

Newer isn't always better, especially when it comes to cryptos.

Why would he get better hashrate with win 8.1?

what consumption?

Win 10 is very good , i have 390 sol with nvidia gtx 1070 .

I think your os is win10. Try win 8.1

you should go to coinmine.pl and mine zcash, zcashclassic. and zdah, flypool.com . suprnova.com these works for sure!.
perhaps you should control better the batch files...............

cards not overclocked.on suprnova pool all works fine.but on f2pool after a while have error.really wanted to try f2pool.

Outstanding miner, keep working on it and don't listen to the haters.

I see people reporting around 400 S/s with overclocked 1070 on Win10. Best i can do is 340 S/s with 150W per card.

Samsung memory.
Drivers 376.33, already tried 376.09 and 375.50 - almost no change.



I am obviously missing something crucial. Any ideas?

410S/s stock EVGA FTW reporting in.
Can get up to 450 with +500MHz on mem, not sure about stability though.
Haven't checked power usage yet too.

@EWBF
Are you able to implement failover option in your next release , mate ?

Yeah i guess if you present it like this, it has merits.

For now, no source code was released, but it will add to their credibility as you said if they do. Also why EWBF should bother with AMD since there is already claymore, the pros of a new miner is to improve upon to be able to justify the fees, or not have any locking of pools. With nvidia he was able to offer a real alternative, for AMD it's pretty much locked on claymore that has cornered the AMD miner market, but as you said it could also be because claymore did a much better job to protect his code. Also he "could" be a cuda expert and less opencl... but again, as you said it's suspicious especially the timing between each optimisation that look to be too much "in-sync" with NH releases.

But what disturb me overall is what EWBF could still be accused of, his miner is obviously different, doesn't has the linux issues EQM has, and work a bit differently. So at the most it would be maybe "solvers cloning" ? I think in both cases the timing is weird, and may be coincidental as you said. I dunno how it goes when you only duplicate a "principle" he can't be considered code thief... maybe ideas thief then, but again can you really say that when it's all mathematical ? I guess there is some algos for trading that are kept very tight secret, the same idea could be applied here...

Waiting for more proof at any cases. What came to me @ first is that it was the same dev as EQM trying to get the best of both worlds... but it seems far-fetched now. Ah well wait & see...


HNY you all

bensam1231, if you weren't English I'd rip you a new asshole.

Sorry. LOL

New Years and I'm half way through a bottle of JD.

Have a Happy New Year - Nicehash and EWBF!

I don't believe shit anyone says, especially online. I look at what's available in terms of actions of participants in a conversation and what's happened. EWBF has suspiciously came out with very close hash wise miners every time Nicehash had a release a couple days later with the exception of where he apparently had virtually the same hash, but never 'fixed' the utilization issue until it was pointed out a couple different times in this thread. It was previously discussed in the Nicehash thread and fixed a few versions earlier.

I noticed the trend before Nicehash started pointing fingers, it just happens to coincide with what I already noticed so I'm giving Nicehash more cred then EWBF. On top of that Nicehash is a company and didn't just pop up out of no where and became a CUDA miner coding guru. Nicehash also already had two previously optimized Equihash miners before EQM, which also lends more credibility to Nicehash.

It may not be true and it could be this is a coincidence, but I kinda doubt that. All it takes is the rough idea and then a good coder can go to town on that. There are plenty of other miners where this hasn't happened, so this is a pretty rare occurrence if it is a coincidence. Take Claymores miners and a lack of a 'EWBF' competing with it. Claymore has also gone to a lot more extremes to prevent disassembly of his code. Considering Nicehash is new to the game, they may have been a bit too lax.

For instance, if it was really this easy EWBF should have a competing AMD miner. Since Nvidia is basically only 20% of the network hashrate, there are much greener pastures on the other side... But Nicehash's miner is in CUDA, not OCL.

As I mentioned this is all circumstantial, but I'm definitely more inclined to agree with Nicehash for various reasons and definitely am not going to hop on the poopoo train till this is sorted out. I'd rather not lose one of the few Nvidia developers. I will say with almost 100% certainty the relationship isn't reverse, Nicehash stealing code from EWBF.


Keep in mind Nicehash isn't trying to 'take out the competition', they basically flipped the board and said 'I'm done' by releasing their source code and exiting the game. That also lends more credibility to Nicehash. Like I said, we as miners are hurt any way this turns out.

Hey, try the following settings:

• power limit @ 80%
• core +185
• memory +640 (you can go further if Samsung memory)

It should give the same hashrate as the one you posted with only 95W.
Good luck!

I will reply to myself ... Newest drivers nedded
1060 does 267 Sols on stock setting.

Yeah, you not saying that back in the bad old days people weren't buying better optimised miners for BTC in underground deals?

That is not something, that anyone should wants too see again!

someone is noticing less gains on flypool?
I have the strange feeling that missing payments
 

I don't see how anyone is stepping on NiceHash toes, since NiceHash is a pool/hashrate selling entity... So no matter what is running they get their share.

regarding "badly optimised" this has no meaning at all. Software gets released whenever there is an improvement and that's it.
You will never get the fully optimize stuff at starts, because it doesn't even exist. Developpment proceed through step into solving issues and self improvement.

Personally, I'm all in favour of fee based miners - coders like yourself get decent rewards for sharing their improvements.

And, I've said, a lot of nice stuff about Nicehash in the past

Bought hash from the site when electricity was $0.30 in the UK.

But, the crypto-currency community needs to get away from having badly optimised mining software - that has no future support or development.

And, it needs to retain it coders for mining software and you know a 2% fee on miners is perfectly acceptable to the small rig owners.

Furthermore, Nicehash could have simply asked EWBF to make his software incompatible with Nicehash BTC mining pool. So, they weren't stepping on each other toes. Rather, than kick up all this dust in everyone's faces.

It makes no sense: closed source both (meaning they can't read the source of each other, have access to ptx or sass which may or may not help), second no licensing what so ever and definitely no patent... (and there is nothing to patent... lol a sorting/xor algo come on lol)

Quite frankly I don't like NiceHash behavior in that one, I understand they paid for their algo a dev, but that doesn't mean other dev must stop their own development... yeah right... they've been in that game for long enough to know that may and will happen...

ps: I hope Adobe open source their softwares because they have some pretty good competition at the moment