Topic: [ANN] cudaMiner & ccMiner CUDA based mining applications [Windows/Linux/MacOSX] - page 180. (Read 3426932 times)

how many 780ti ? (actually I would go for the 880... ).
Actually, I have made so far 9.34btc (still highly dominated by maxcoin profit... I can almost get it already  )
I think I will go water for my next rig... do you know what kind of waterpump and stuff are needed for something like 4 cards and 1 cpu ?

wow.

This is the future of pricing. In BTC.

Try to look into the future and pick a coin which has a bright future and mine it. When POW ends hold and stake. Try to buy some if price is cheap. If price goes way up sell some.
Repeat procedure on next promising coin.  And think pumpkins not peanuts. 
 

ok, updated it :p

Looks like u are not a big fan of BTC. U should put that amount in BTC too. U are in crypto forum  

Long live fiat.

AMD is to me simply too power hungry, i had a top of the line FX not long back and sent it back due to power usage.
I need to buy a second machine anyway as my partner keeps stealing mine  , and even if i get last gen (same as i have now) 6 core xeons, with a proper mb, cpu, new memory etc. i am still looking at a good chunk of money.

Both machines are going to be high end gaming machines so the $4k was budgeting in 4 displays, another full system, plus a 780 Hydro Copper to go with my current 780 hydro copper, please another 750Ti, plus watercooling fittings.

This is the expense list (in the UK so inflated prices)

Displays: 4 x VN279QLB = 2.86BTC
EVGA 780 Hydro Copper = 1.42BTC
Watercooling Fittings = 0.43BTC
CPU: 1.714BTC
MB: 0.86BTC
Memory (16GB): 0.46BTC
Case: 0.37BTC
PSU: 0.26BTC
SSD: 0.514BTC

Total: 8.88BTC xD or 0.05BTC each if 178 people donate?  

how many coins per day? total hash?

Questions for developers:
1. why memory for hash is not allocated in GPU shared memory but in global ?
2. can OpenMP be used to parallelize tasks on CPU ? e.g. validation.
3. did anyone thought of splitting communication host<->GPU into 2 separate threads ? basically one thread constantly and asynchronously reads data from stratum/pool and copies it to an array located in GPU, GPU kernel does all calculations and save it to an array @GPU, second thread constantly and asynchronously reads calculated data from an array @GPU to an array @host, validates them @CPU and sends it over the network to the pool. the idea is to avoid GPU kernel from waiting for host.

what more experienced developers think about those ideas ? I am lousy programmer so do not expect proof of concept from my side soon :-(

Can't imagine that'll last long.

Homie, you're thinking about it in the entirely wrong way. You don't mine that one crypto that can make you cash now, it's a losing game. Even if you've got a Petahash of hashrates for BTC, to increase your hashrate with the difficulty, just to keep earning the same amount... It's guaranteed to lose all your money. Use whatever power you've got, but look into the different coins. How long have they been out, what's the network hashrate like? Does their announce thread have a lot of hits in a short period of time, or not much hits or replies but has been out 3 months? Do the devs stay active in threads where the coin is being talked about? Do any of them have a history in cryptos, and have they made previous Cryptos? How did they do, good and bad? Have they learned from their mistakes? Is there reports of fraudulent behaviour?
Base the coins you choose to mine on things like those examples. Pick a coin that looks like it's been looked over, and has a lot of support and promise in its future. People need to be rich in order to successfully mine bitcoins, and to efficiently profit, not talking about $0.70/day. The people that got rich off of bitcoin were the ones that thought they wouldn't mind sparing a bit of hashrate for this useless software hashing thing for fake "coins", or they heard about it, and when reading online, realized that they could buy like a few thousand for a few dollars and thought they were going to just piss their money away. Turns out it finally paid off for the people that just held it for fun, or invested a few bucks in it. It's unfortunate though, that a lot of people seemed to sell those coins that they bought for $0.00004 a piece for $10 each when it reached that, thinking they were the smartest people alive to have got out when they did. But then when it reached around $500, or even that one time it had hit $1200, well, people were just kicking themselves for it.

Personally, I'm always mining for different algos, whatever looks interesting. Some die before I can get on the exchange, however that just indicates that they were created by the devs for a cheap and easy way to get quick money. Charge people for

Anyone have any coins that are worth gpu mining anymore?

No, I don't think numbering has anything to do with connected display, GPU#0 is usually the card installed into the "main" pci-e x16 slot (in most cases the top slot), but sometimes the order changes in multi-gpu rigs. Install some monitoring app (like MSI AB), watch for "gpu tachometer" graph while stopping the fans on the cards with your finger.. that's what I do to find out gpu order in my rigs, then I label the cards and write down which card corresponds to which index in monitoring and mining software.

i bought 16p to 16p for € 45 and the others 1p to 16p for € 45.. molex versione 'cause no-molex version seemed extreme China version anyway, i don't think that the cause is one riser 'cause gpu error goes from GPU#0, that is into MoBo, the only one into MoBo.. at least this is what i think.. the error comes from the only card that is into MoBo, that is the same card where i've plugged in the HDMI.. so if that is the card connected to monitor, that must be GPU#0..right? in other words, GPU#0 is the card you use to connect monitor, is it true? if i see video output on monitor it means that card is the GPU#0

Oh, I see. Been there, done that. Made out alive, though wasted a lot of money in the process. It's just consumer-grade hardware does surprisingly well these days, a simple AMD FX-8350 gives all the power that vast majority of hobbyists need, supports up to 32GB of ECC RAM, and there's enough 990FX boards out there that work great as hypervisors with Xen/Vmware.. There's also a lot of previous gen Xeons on ebay, 6 core / 12 threads 1366 cpus go for like less than $150 and are still very capable. One can build a very powerful server for under $1k and in real-life applications most would never even notice the difference in performance compared to the latest/greatest top xeon.

Well, if you want that latest stuff and can afford it — then sure. For me it's just never worth it, I do like to geek out and play with new toys, but most fun is gone in like a week after the purchase and then I start to feel like shit, realizing that I could get all the performance I needed for a fraction of the price paid.

You might have a problem with power, afaik your EVGAs don't have 6-pin connectors so they draw all the power through the motherboard, and that's a lot of current. The easiest way to check whether not enough power is the issue is to use a powered riser (usb, or powered classic). Edit: oh, sorry, they do have 6-pin connectors, my bad (cool-looking cards by the way, might be a better option than my gigabytes). Not sure then what's the problem. Try to swap risers/cards and see whether there's a bad riser or something. I've had a problem exactly like this (connected the 6th 750ti to a rig that previously had 5 cards installed, and also had crashes, although I don't remember whether those were ccminer crashes or OS crashes) and in my case it turned out that one of the risers went bad. Those cheap "classic" risers have really bad soldering sometimes, I've thrown away like 2 or 3 out of ~30 I've bought on ebay.

community, i have one big problem.. i run windows 7 Home Premium 64 bit on a MSI Z97 G7 (Gaming series), an i5 4690k, 8 gb RAM Kingston 1600mhz, a 850watt PSU and 6 evga 750ti ACX FTW.. until yesterday i've ran my rig only with 5 cards 'cause i had only 4 PCI-e risers.. and all ran fine..today ebay delivers me the missing riser, so i've push it into the PCI-e and i've attached the 6th vga card.. now ccminer don't work.. it runs only for 30-40seconds and then crash.. i don't understand why.. with 5 card it works, with 6 cards it doesn't work..
risers aren't usb risers, they are classic riser.. i've one card attached into MoBo into the 16pin PCI-e, 2 cards attached with 16pin to 16pin risers (obvioulsy into 16pin PCI-e) and in the end the other 3 cards are attached with 1pin to 16pin risers.. what's my problem? it's caused by my rig or the problem may be ccminer? have sense to try with ubuntu?

many thanks for your attention..

Shhh I neeeed it :p
I'm waiting for the new 8 core 16 thread i7's to come out. CPU, mb, ddr4 memory. Roughly $1700 or so

My current Xeon is going as a spare so the website will go on that. Plus its an excuse to get a second PC set up. I also need to sort out my new displays to code properly. Coding on less then 3 screens sucks. And I only have 1 right now

So yes, it will be my new toy haha. I have a list of water cooling fittings that are going to cost $500 or more just to fit everything again lmao

Why on earth would you need a $4000 server? There are many ways today to build a powerful "hobbyist" server for under $1k.. I mean, if you're going to host as simple a web site as cudamining.cc there (obviously among other things, but still) — are you sure you need the latest and greatest enterprise level hardware? Sounds more like someone just wants to buy new cool toys to play with.

Thanks
I brought the new domain for 2 years so we don't have to worry about it for a while.
Once my new server is built the website will get its own dedicated server to keep things running nicely. I just need to save the money up for it. $4000 is a tough number to swallow

I will send some JPC to help with your troubles 

Right everyone. As I was taking so long building the new cudamining website, and found out yesterday I couldn't edit the domain. The website will soon be moving to a new domain and the old website will forward you to it.

The new domain name will be cudamining.co.uk so i am keeping the name just changing the extension.
It is not live yet so don't bother checking

Just a heads up to everyone