stop your insane one page spamming...
oh and again your numbers, as usual, do not make sense (where did you see 1500$ for a week ? 40 hours... ) you are clearly not awake... still dreaming
so it isn't last time, it is all the time
10% of the bounty is most likely 0.2-0.4btc, ok you are still in unicorn world where monero team proposed a 15k $ bounty (not sure what you used that time for btc either...) may-be one day, you will make a post when you are awake...
and for your info because you seem to invent really a lot of stuff this is how I work:
someone wants to get some work done on something, he contacts me (or I contact him) we agree on a price and I work on it... that's it...
anyhow, might have a look it, but not now, because I don't have time and the bounty system obliged me (since I am not hired) to do that only on my free time... so long for your business model (as usual).
ps: Have you consider making of your trolling a business model ?
Type page long posts and being paid on a word basis... you will make 10k $ per post (I might not be awake, but that's ok
)
3000~ XMR bounty. 3000*0.0019=5.7btc That's $1425@250 in USD for the full bounty. You'd know that if you checked the thread before turning the job down that he brought you because 'it didn't pay enough'.
I never said $1500 for 10% of the bounty.
Humoring you, I find it extremely hard to believe you need to work 40 hours for a 10% improvement in something like Cryptonote which has a miner from over a year ago that was thrown together for a bounty and really hasn't been updated since.
If you want to convey thoughts instead of showing of your childish sense of humor and maturity, it takes more then a couple run on sentences to do it. You know, the whole being professional thing.
Linux?
Works on linux, but you don't get the sourcecode and I won't make a build. Windows for now.
The miner needs some more work. As you can see the performance on 960 is worse than on the 750ti. The gtx 980 is above 500H/s
There is a pretty sizeable bounty for work on it dude. I guess it's not good enough for DJM, but perhaps you'd be interested. I don't know much about coding, but that is one of the bigger bounties I've seen on the BCT forums if it's legit.
(a) I think you're missing some of djm's points: In a bounty, the developer takes on all of the risk. There's a decent chance of failing to get a 20% speedup after a month of work. That's kinda sucky, particularly when contract programming work pays better anyway.
(b) $1500/week is low for a good US programmer.
http://money.usnews.com/careers/best-jobs/software-developer/salary2013 average software dev salary was over $92k/year ($1769/week), and the ones who can crank out a 50% improvement on the Monero miner are closer to that top 10% figure -- think $150k/year and up. The bounty also doesn't include benefits and the other nice things offered by a standard employer.
I base this upon knowing personally a few of the people who ship some of the highest performance mining code. Many are not US based, but I know what their day jobs would pay if they were here. (It's probably more than $150k/year.)
Freelancing generally needs to quote double the hourly rate of a salaried position when you take everything into account, which means that a US programmer earning $150k per year needs to charge about $150/hour -- or about $6000 for a full 40 hour week, not the $1500 you mentioned.
I ran the numbers on doing this as a business a while ago, and it's not clear that it's worth it given the other things you could do with the software talent you need.
-Dave
a. Perhaps if you're trying to optimize and make sha256 competitive with ASICs. A year old miner in something like Cryptonote that no one touches or even competes with and isn't in any more of a state then a working one is something I hardly consider you'd spend two months on for a 20% increase. You're trying to overstate your position.
While I don't program in CUDA or OCL, I can put two and two together. When Wolf0 brags about 15% increases on things like scryptn for fun and he and DJM like to whip their epeens all around, if you listen, you realize they're full of shit when they try to pretend how much time they spend on things.
b. OCL and CUDA are just like every other language. There is nothing super specialized this. I offer you less cherry picked results:
http://www.bls.gov/oes/current/oes151131.htm$40 a hour isn't unheard of and that's for people with credentials in a actual position, not backwater bumpkins posting on BCT. Either way, we aren't talking about hiring a 10+ year experienced software developer that's sitting comfy in a position at a financial investment firm.
I ran the numbers on doing this as a business a while ago, and it's not clear that it's worth it given the other things you could do with the software talent you need.
And there was never anyone on BCT with a ulterior motive.
We aren't talking about competing with ASICs here, it just needs to be better then the competition, which is basically just DJM and Wolf0 right now.
I think I can make the monero miner 100% faster. but the AMD miner from claymore is too good to make it worth the effort. I work fulltime in a company that pays $xxx,xxx dollars a year. Optimizing CCminer is just a hobby I do in my sparetime after work.
Donate some beers, and I am happy.
Considering the power efficiency of Maxwell it'd be inherently more power efficient regardless of being that much better then Claymore. Towards the end of mining I basically just mined Monero, but there is still profit to be made there. A 290x gets you 710h/s... A 970 currently gets you 400h/s with TSIV Cryp miner... Double the hash would be on par with 290x's with a 970 and more power efficient to boot.
About 810 h/s on stock 290X (Hynix memory).
About 710 h/s on stock 290X (Elpida memory).
About 690 h/s on stock 290 (Elpida memory).
About 550 h/s on stock 280X (Hynix memory).
About 440 h/s on stock 270X (Elpida memory).
About 410 h/s on stock 270 (Elpida memory).
I know you have a great job, but think of the children...
