|
4x 6950 --> 1.80 MH/s
2x 6970 --> 0.95 MH/s
3x 7950 --> 1.60 MH/s
3x 7970 --> 2.10 MH/s (coming soon)
--> 6.45 MH/s
Also coming soon: summer heat underclocking panic at 5.00 MH/s
|
|
|
|
Your settings are all messed up because all you're throwing is hardware errors Please use the 7950 (high usage) defaults specified in GUIminer-scrypt http://bitcointalk.org/index.php?topic=150331.0Also please cool your cards better or turn down the voltages, you're going to kill them very quickly |
|
|
|
This is just mad. You don't want an expensive 51% attack with only temporary results. You want a re-org attack, a blockchain fork, or hack a client to remove the requirement for transaction fees and start throwing massive blocks at the network.
LiteCoin is based on a very old version of BitCoin that has a few known problems. These problems can be attacked for less than what a 51% attack would cost.
Coblee modified the client already to reject spam transactions by making the fee mandatory in the client/litecoind. Pretty much any attack that forks the chain will be corrected within a few hours, as coblee, pooler, and laSeek keep a close eye on it, despite not posting here much. When the spam attacks started, coblee fixed them within a day or two. When the previous 51% attack was threatened, he checkpointed the chain. |
|
|
|
|
I'm starting to think this is actually just a scam.
|
|
|
|
Okay, yeah. Even if you do get this to work, you may only see 50 kh/s: http://forum.litecoin.net/index.php?topic=1493.0My advice would be at this point to upgrade the card. You can try updating the display drivers and seeing if that helps, though. edit: You can also try running this from cmd.exe (command console) and seeing if it works: C:\guiminer\cgminer\cgminer.exe --scrypt -u user -p password -o stratum+tcp://ltcmine.ru:3333 --shaders 800 --worksize 256 --intensity 10 --no-adl You will need to edit the path above to point it to the correct directory, and edit the user/pass/server info |
|
|
|
|
Sorry, it slipped my mind. It's starting to look like a driver side issue. What version of windows are you running, what AMD Driver version, and what AMD APP SDK version? Also what card and motherboard are you running?
|
|
|
|
There is a severe security flaw using Berkeley DB and the default settings given in older versions of Bitcoin. https://en.bitcoin.it/wiki/BIP_0050http://bitcoin.org/may15.htmlThis needs to be addressed within the next quarter for Litecoin. [00:44] <fuzzy> uh guys
[00:44] <fuzzy> https://en.bitcoin.it/wiki/BIP_0050
[00:44] <fuzzy> is that serious?
[00:44] <fuzzy> that article specifies that the bug persists for all versions of bitcoin prior to 0.8.1 and that implicit forks of the blockchain can be performed at any time
[00:49] <warren> fuzzy: yes, it's possible, just very difficult
[00:49] <warren> fuzzy: easier than 51% difficult
[00:52] <warren> fuzzy: I'm posting on the forum about this now.
[00:53] <fuzzy> Thanks!
[00:55] <warren> coblee: gavin is raising awareness of the self-consistency hard fork risk and users are now beginning to ask about it. https://bitcointalk.org/index.php?topic=159238.0;all
[00:56] <warren> coblee: IMHO, you should issue a patched 0.6.3 with a hard-fork scheduled on a future date so litecoin users don't have to worry about this.
[00:56] <warren> coblee: on second thought, I was not cautious enough in accepting waiting for May/June to decide what to do.
[00:57] <warren> coblee: https://en.bitcoin.it/wiki/BIP_0050
[00:58] <warren> coblee: your 0.6.3 hardfork would be modeled like this: http://bitcoin.org/may15.html
[00:59] <warren> coblee: Please understand that the DB_CONFIG workaround is NOT safe. That *does* the hard fork. Rather than have users do it haphazardly, you should schedule it in the future.
[01:00] <warren> coblee: I have more bad news.
[01:00] <warren> hmm. I'll explain later.
[01:08] <warren> fuzzy: to be clear, it is my personal opinion having studied this issue that the risk is REALLY SMALL. Bitcoin itself is vulnerable to this risk until May 15th and folks aren't worried about it.
[01:08] <warren> fuzzy: the real risk is if Litecoin doesn't hardfork to eliminate the risk in a timely manner.
#litecoin-dev @ http://webchat.freenode.net/Details on how to temporarily fix this for the Litecoin client version 0.6.3 are here: https://bitcointalk.org/index.php?topic=159238.msg1698182#msg1698182Eventually the client will need to move to 0.8.1+ versions. Coblee has stated he will do this: Litecoin will be updated to 0.8.1 soon. Like wtogami said, I had originally planned to stay 1 major version behind Bitcoin so that we don't get bit by a bad release. But as it turned out, the 0.8 hard fork just made it clear that there was an issue with old releases that could cause a hard fork even without 0.8. So wtogami and pooler will be helping me with rebasing the Litecoin code to 0.8.1. I do want to wait til May 15 to make sure that Bitcoin is able to handle the potential hard fork. Since the Litecoin community is much smaller, we should have even less of a problem. But it's better to be safe.
https://bitcointalk.org/index.php?topic=159238.msg1699934#msg1699934However, most other major altchains operating on forks of bitcoin versions before 0.8.1 will also need to implement this, so please update your sources accordingly! |
|
|
|
this is not best hashes per money, but absolutely perfect for first time rig builder, do that. and remember "cheap" mean the cheapest you can find. I would only get more power for PSU, but i know thats not a popular view to have a 30% power reserve, but thats my way.
No need, this rig only pulls 525w from the wall. This is my major gripe against 69xx cards, my 4x 6950 rig pulls 850w and 1.6 MH/s while my 3x 7950 rig pulls 1.6 MH/s and 525w. |
|
|
|
Ummm.... speaking of the hardfork, has a version of Litecoin been released yet that fixes the problem?
It was my understanding that we avoided the problem by simply never upgrading the software beyond 0.6.3c. Your userbase makes for a better testnet than our userbase because it's larger. My impression of the bug leading to the fork was that 0.8.0 allowed the mining of blocks greater than the allowable size in the older clients, and that the adoption of 0.8.0 by half the network was what led to the fork. If I'm wrong, please correct me (you'd be the one to know). If 0.8.1 works well, we will probably see an update to the client for it later this year. Edit: Reading https://en.bitcoin.it/wiki/BIP_0050 , I see this may affect Litecoin Can we use a file called DB_CONFIG in the Litecoin data dir that contains set_lg_dir database
set_lk_max_locks 120000 To temporarily fix this until we update to 0.8.1? If so I will give the heads up to pool ops |
|
|
|
|
Intensity is too low, increase intensity or try the low usage profile (which is sometimes strangely faster)
|
|
|
|
|
Any good 500w gold or platinum psu can power that. You will never pull more than 400w with 2x 6770s on that setup.
|
|
|
|
|
No, get a 3x pcie x16 mobo, 3 7950s, a cheap cpu, a gold 850w power supply, 8 gb of ram, and a cheap hdd and you're good to go
This was my first rig, it paid for itself in about 4 months.
|
|
|
|
Interestingly if I downclocked the core to 725mhz it increased the hash rate to 315kh/s... Downclocking memory had detrimental effect after hitting below 1400mhz
Try 725 MHz and 1500 MHz RAM, intensity 14, 16, and 18 There may just not be enough memory bandwidth for that many shaders |
|
|
|
The difficulty will only go up from here and it's already just barely profitable I know, it's awful making $40 a day right now on 6 mh/s. OP: 7950s. |
|
|
|
In afterburner, set core to 950 MHz and memory to 1555 MHz use reaper and this litecoin.conf host yourpool.com ### Do not add http://!!
port 8344
user username
pass password
protocol litecoin
worksize 256
vectors 1
aggression 20
threads_per_gpu 1
sharethreads 32
lookup_gap 2
gpu_thread_concurrency 18610 Tell me what you get Try also aggression 18 |
|
|
|
|
I've been mining on 7950s for six months, cards are fine, no artifacting.
|
|
|
|
|
I want all of them. Shipping to Canada. Payment in BTC. PM me if interested with what you want for shipping. I would guess it's about $40-50 extra.
If you respond quickly I'll get the payment out today.
|
|
|
|
|
Show Posts
