Ok,
Now, the miners detect my card and work fine for bitcoin mining but scrypt still isn't working.
My cgminer output is stuck here:
Any ideas?
You need to change OSs as scrypt mining requires the SDK 2.6 or newer to work. Or you can do the financially in your best interest thing and mine BTC with that card until LTC is more profitable.
|
|
|
network terahash goes up, but not the price of btc What i'm wondering is, what timeframe are you looking at? When i look at the charts i only see the opposite, that the price of btc went up about 20% within the last month, but the network hashrate didn't change at all. We have had almost three difficulty increases in this last month, the hash rate has increased.
|
|
|
why is this? shouldn't the increase in terahash bandwidth cause an increase in price of btc?
i mean it means that more people are fighting over blocks and its rewards
That 50 (soon to be 25) BTC reward is a small piece of the bitcoin economy. Most people using and trading BTC are not fighting over the block reward. Anyway, your premise is wrong. The price of BTC is rising. It was 1/3 of the current price less than 1 year ago. And 15 months ago it was just about 3x the current price or if going back to last December just about 1/6, if you are choosing arbitrary point in time to try and look like you are making a relevant point at least pick the one that looks the best. Oh and to the OP as the difficulty has increased this summer the price has risen as well from $5 at the start to just about $13 now. More than making up for the increase in the difficulty.
|
|
|
I try with your confgs. 313 khash/s per device. Why you decrease stock engine and memory clocks?
Through many hours of messing about one day those are the settings I found give me the max hash rate possible. I did this on every card I have and every one without exception gives more hashes if you lower the gpu engine below stock speeds and raise the memory speed once you get to a certain point on the memory nothing above it will give you more speed in fact usually decreases it. You use --auto-gpu option. This option only decrease engine? No it automatically manages the gpu engine if it gets above the --temp-target you set the --auto-fan does the fan speed to keep it below the --temp-target. The second set of my posting is the speed settings I set for ltc mining the --od-setclocks lines. Thx for answers. What version of SDK you use? I cant understand why my hashrate is so low. AMD-APP-SDK-v2.7-RC-lnx64 is the directory I get from the tarball I downloaded and fglrx_8.980-0ubuntu1_amd64.deb is the package that was built from the .run file I downloaded I think the 12.6 version. You want to check the proper speeds are being set in my experience with cgminer unless you set them manually then it will use whatever is programmed in the cards BIOS ignoring the values put into your .cgminer.conf. You do this by using. DISPLAY=:0 aticonfig --od-enable --adapter=all DISPLAY=:0 aticonfig --odgc --adapter=all
In a terminal/console window.
|
|
|
I try with your confgs. 313 khash/s per device. Why you decrease stock engine and memory clocks?
Through many hours of messing about one day those are the settings I found give me the max hash rate possible. I did this on every card I have and every one without exception gives more hashes if you lower the gpu engine below stock speeds and raise the memory speed once you get to a certain point on the memory nothing above it will give you more speed in fact usually decreases it. You use --auto-gpu option. This option only decrease engine? No it automatically manages the gpu engine if it gets above the --temp-target you set the --auto-fan does the fan speed to keep it below the --temp-target. The second set of my posting is the speed settings I set for ltc mining the --od-setclocks lines.
|
|
|
I try with your confgs. 313 khash/s per device. Why you decrease stock engine and memory clocks?
Through many hours of messing about one day those are the settings I found give me the max hash rate possible. I did this on every card I have and every one without exception gives more hashes if you lower the gpu engine below stock speeds and raise the memory speed once you get to a certain point on the memory nothing above it will give you more speed in fact usually decreases it.
|
|
|
Hi im somewhat new to bitcoin. Ive been mining for a few months but havent really done anything with the bitcoins. What is the best/easiest/fastest way to exchange my BTC into a visa or my bank account. I live in Canada. Id prefer not to have a prepaid visa and just have it put on my regular visa that I use. Keep in mind im kind of basic here so if you could step by step me that would be great. Thanks
They have withdraw to bank account option. https://www.cavirtex.com/
|
|
|
Hello again. Some info about my rig: Xubuntu Live (on 4 gb flash usb); 4 GB RAM; HD 6950 unlocked @6970; HD 6970 I build cgminer with --enable-cpumining & --enable-scrypt. When i start cgminer i receive error message: sudo ./cgminer --scrypt -o http://ltc.kattare.com:9332 -u xxx -p xxx --shaders 1536 --intensity 19 --thread-concurrency 6144 -g 2 [2012-10-03 17:20:48] Started cgminer 2.7.6 [2012-10-03 17:20:49] Probing for an alive pool [2012-10-03 17:20:49] Long-polling activated for http://ltc.kattare.com:9332/LP [2012-10-03 17:20:49] Maximum buffer memory device 0 supports says 268435456, your scrypt setti ngs come to 402653184 [2012-10-03 17:20:49] Error -61: clCreateBuffer (padbuffer8), decrease CT or increase LG [2012-10-03 17:20:49] Failed to init GPU thread 0, disabling device 0 [2012-10-03 17:20:49] Restarting the GPU from the menu will not fix this. [2012-10-03 17:20:49] Try restarting cgminer. Press enter to continue:
[2012-10-03 17:21:11] Maximum buffer memory device 0 supports says 268435456, your scrypt setti ngs come to 402653184 [2012-10-03 17:21:11] Error -61: clCreateBuffer (padbuffer8), decrease CT or increase LG [2012-10-03 17:21:11] Failed to init GPU thread 1, disabling device 0 [2012-10-03 17:21:11] Maximum buffer memory device 1 supports says 268435456, your scrypt setti ngs come to 402653184 [2012-10-03 17:21:11] Error -61: clCreateBuffer (padbuffer8), decrease CT or increase LG [2012-10-03 17:21:11] Failed to init GPU thread 2, disabling device 1 [2012-10-03 17:21:11] Maximum buffer memory device 1 supports says 268435456, your scrypt setti ngs come to 402653184 [2012-10-03 17:21:11] Error -61: clCreateBuffer (padbuffer8), decrease CT or increase LG [2012-10-03 17:21:11] Failed to init GPU thread 3, disabling device 1
Before i export GPU_MAX_ALLOC_PERCENT=100 & GPU_USE_SYNC_OBJECTS=1. Maybe trouble with freespace on my usb flash drive. Without --thread-concurrency option maximum hashrate about 340 on each gpu khash/s. Try something like I used in a machine with two 6950s flashed to 70s. cat ltc.sh export GPU_USE_SYNC_OBJECTS=1 export GPU_MAX_ALLOC_PERCENT=100 export DISPLAY=:0
~/cgminer --scrypt --worksize 256,256 --thread-concurrency 7040,7040 --vectors 4,4 --gpu-threads 2 -I 18,18 -g 1 --auto-fan --auto-gpu --temp-target 81
And the card speeds.
cat vset.sh DISPLAY=:0 aticonfig --od-enable --adapter=all DISPLAY=:0 aticonfig --od-setclocks=770,1230 --adapter=0 DISPLAY=:0 aticonfig --od-setclocks=770,1230 --adapter=1 DISPLAY=:0 aticonfig --odgc --adapter=all
This gives me 410kh/s per card, leave out the --shaders like I do only one of them is needed so it is it or the --thread-concurrency option not both.
|
|
|
Quebecers have the signal honor of being one of only two nationalities who have defeated the US in war.
It was actually the British army who won that war. More like the Mohawks and assorted other tribes that were allied with the British won it for them. The french in France were working with the yankees against the British, the ones in Quebec had no army to fight anyone.
|
|
|
With 3 5970s, whouldn't you use "--shaders 9600"?
Each 5870 has 1600 shaders. Each 5970 has 2 x 1600 = 3200. You have 3 x 3200 = 9600.
No it is per core not total per card that is used by the program when making the .bin file used/loaded at runtime and I always use the --thread option as that seems to load the correct .bin every time the --shaders does not do so for me.
|
|
|
I got that when I had left over .bin files try deleting them, also you need only one of the --shaders or --thread-concurrency in there not both and the --thread-concurrency option should be a multiple of your shaders on the card less than the 8192 I have found to be the max that works for me so try 7168 for it from your example command used above. Also just noticed 5970s in your post they have 1600 shaders per core so 8000 is a good number to go with there, my script I use to start it with in a machine with a 5830 as first card in it, 5970 is second/third card. cat ltc.sh export GPU_USE_SYNC_OBJECTS=1 export GPU_MAX_ALLOC_PERCENT=100 export DISPLAY=:0 ~/cgminer-ltc --scrypt --worksize 256,256,256 --thread-concurrency 6720,8000,8000 --vectors 4,4,4 --gpu-threads 2 -I 18,18,18 -g 1 --auto-fan --auto-gpu --temp-target 83
Memory speeds I have found to give me the max hashing rate for each model of card. cat vset.sh DISPLAY=:0 aticonfig --od-enable --adapter=all DISPLAY=:0 aticonfig --od-setclocks=870,1170 --adapter=0 DISPLAY=:0 aticonfig --od-setclocks=750,1050 --adapter=1 DISPLAY=:0 aticonfig --od-setclocks=750,1050 --adapter=2 DISPLAY=:0 aticonfig --odgc --adapter=all
|
|
|
installed bzr with apt-get bzr and executed the commands as recommended, no success yet.
Weird I searched on how to see your installed python modules and found the python-pip which contains the pip command I found to use like so after I installed the python-jsonrpc. miner2@miner2:~/trunk/python-jsonrpc$ pip freeze | grep rpc jsonrpc==0.01
It clearly shows it as installed so that program you are trying to run has some kind of problem finding it, I have no clue how to track that down.
|
|
|
Im trying to install/running eloipool. Actualy i face an error as followed: root@j064:/opt/eloipool# ./eloipool.py Traceback (most recent call last): File "./eloipool.py", line 46, in <module> import jsonrpc ImportError: No module named jsonrpc root@j064:/opt/eloipool#
Any idea? Install bzr then. bzr checkout http://bzr.json-rpc.org/trunk ; cd trunk/python-jsonrpc sudo python setup.py install
|
|
|
Ah crap. I scrolled up and I see the problem: checking for working alloca.h... yes checking for alloca... yes checking for OpenCL... no checking for pthread_create in -lpthread... yes checking for json_loads in -ljansson... no checking for library containing addstr... -lncurses checking for yasm... /usr/bin/yasm checking if yasm version is greater than 1.0.1... yes checking libudev.h usability... no checking libudev.h presence... no checking for libudev.h... no configure: error: libudev not found root@raspberrypi:/usr/src/cgminer#
Fixing now... Edit: Or I'm trying to fix it but I'm clueless (apparently). I tried: apt-get install libudev
with no luck. If someone can help me with this I would appreciate it. When compiling you need the development package of the library when it is not found so to find it install apt-file. apt-file search libudev.h libudev-dev: /usr/include/libudev.h
|
|
|
To clarify: Do they appear in your wallet as soon are they are mined?
They do go to your wallet, assuming you put an address in that wallet in that block. For example, look at the first transaction in http://blockexplorer.com/block/00000000000000a4d73468e752d7eb54ac8de5c131188361a811676ecbd67228and you'll see it specified an address as the destination for the generation of the block. However, it's not immediate. You need to have 100 confirmations after you mine a block before the network recognizes that you own the coins, so it's going to take about 16 hours 40 minutes to show up. How do you "put an address in that wallet in that block"? I thought all I had to do was setup my .conf file with rpcuser, rpcpass, allowed ip's, port number and then run the client with -server Have I been doing this wrong for the past month? Never done it with BTC but any of the altcoins I have mined solo that is all that is needed for the coins to show up in your wallet as immature once you have found a block until confirmed by the network. And of course you need to have your miners using your local *coind as their source of getwork. Whew... Scared the crap outta me, thought I wasted a month's worth of hashing, lol... But yes, I have my .conf file setup for my "server" and my miners are all pointed at my local host, so I SHOULD be okay... I think... LOL... You are good there then although solo mining BTC is pretty much an iffy proposition with the diff being so high, you may want to give LTC or PPC a try those are lower rate chains where you have good chance to get a block within reasonable amount of time then you could trade them for BTC.
|
|
|
To clarify: Do they appear in your wallet as soon are they are mined?
They do go to your wallet, assuming you put an address in that wallet in that block. For example, look at the first transaction in http://blockexplorer.com/block/00000000000000a4d73468e752d7eb54ac8de5c131188361a811676ecbd67228and you'll see it specified an address as the destination for the generation of the block. However, it's not immediate. You need to have 100 confirmations after you mine a block before the network recognizes that you own the coins, so it's going to take about 16 hours 40 minutes to show up. How do you "put an address in that wallet in that block"? I thought all I had to do was setup my .conf file with rpcuser, rpcpass, allowed ip's, port number and then run the client with -server Have I been doing this wrong for the past month? Never done it with BTC but any of the altcoins I have mined solo that is all that is needed for the coins to show up in your wallet as immature once you have found a block until confirmed by the network. And of course you need to have your miners using your local *coind as their source of getwork.
|
|
|
They're 20 GH/s BTC setups won't be 20 GH/s setups on the LTC network. Have you ever mined LTC? If so, you should know that you don't get anywhere NEAR the same hash rates as on the BTC network even using the same hardware and exact same miners.
I have and have crunched the numbers. You can get ~90% of your rate in mh/s so your 20gh/s BTC gets 18mh/s LTC in your example here at a cost of 10% extra on the power consumed, if you were under clocking your memory if not then the power consumed stays roughly the same.
|
|
|
You mention 1-2 months, lets say 2 BTC per week (1 BTC to each lender) for 7 weeks? That is if BitcoinINV is okay with 14 BTC return.
So have you guys figured out how he is going to cover the other ~1.1btc per week needed for this repayment? A 5850 running at ~370mh/s the best I have been able to get out of any of mine gives you .13btc per day no fees taken off current difficulty on a mining pool so .13*7=.91 per week return.
|
|
|
you should be able to create symbolic links between the directory where Bitcoin-Qt looks for the files and the directory on the removable USB where you will actually store the files. This sounds like a relatively painless method for achieving cold storage.... can anyone point me towards how I can do this on Mac OS 10.6.8??? Once can can get this done, I will then look into wallet backups on wuala, and then I think I'll have a good system going. Thank you! ln -s /Users/username/Library/Application\ Support/Bitcoin /Volumes/usb_drive_name/Bitcoin in terminal or you could just use the --datadir=/Volumes/usb_drive_name/Bitcoin when starting the bitcoin to forgo the symbolic link all together. You would do this in a terminal window like so /Applications/Bitcoin-Qt.app/Contents/MacOS/Bitcoin-Qt --datadir=/Volumes/usb_drive_name/Bitcoin.
|
|
|
|