Ok, got a 5th card. Since I am maxed out at 4 cards with a 1200 watt power supply I had to get another power supply. I plan on eventually having two 4 card rigs but until I get more GPUs I will use the second rigs PS and GPUs in the first one. Will also be running up against the max of 16 amps on the 20 amp circuit I have. May or may not be able to run 6 GPUs until I can get a higher amp circuit. At that point I will be switching to 208 volt from 120 I'm using now. The second power supply is also a Corsair AX1200i. This picture shows the two CorsairLink interface units. Since CorsairLink only supports reporting on one PS at a time I have to disable one and then enable the other (in device manager) to see it's individual power usage. A nice surprise was that even though some of the GPUs are being used on the second power supply, CorsairLink still see's them for monitoring.  Here is a pic of the case with the GPUs now facing the front of the rack (cold air isle). You can also see the temporary orientation of the two power supplies. You might notice that I am only using powered risers on the first 4 GPUs. Think I'm okay not powering the 5th one. If I had 6 or 7 GPUs then I would break out another molex cable to the power supply. I may still do that just for the 5th GPU.  Close up of the connections. 1st power supply has 3 GPUs + 8pin MB + SATA power. 2nd power supply has 2 GPUs + 4 connector molex cable. I can now see that if this was going to be a permanent setup the case should of been made longer so there could be more space between the GPUs and power supplies. The additional space would facilitate easier cable management and allow some distance between the GPUs and the PS since these particular GPUs expel the majority of hot air towards the power supplies. Even with one PS a longer case to distance the PS from the GPUs would (I think) have been beneficial.  |
|
|
|
There's no way you can run 7 7970's on a standard 16amp circuit.
Even if you could undervolt enough to make it fit, it's probably not safe to run that 24/7 for months anyway.
It's a 20 amp circuit. Don't want to go over 16 amps (80%) on it. |
|
|
|
|
Prelude,
Can you please post your power consumption on your (7) 7970s rig?
Watts and amps if you have them.
I am at (4) 7970s and am hitting over 1000w. and 9.4 amps with each GPU at 1.125 volts.
Watts Amps # of GPUs Notes
291 2.45 1
539 4.55 2
804 6.906 3 verified at the wall with Kill-a-watt
1072 9.422 4 Started using power-in for watts in CorsairLink as efficiency in/out started to go down. Have not verified with Kill-a-watt yet.
Trying to see if I can fit 7 GPUs inside of 16amps. Looks like I'll be just over :/
Note: worst case watts/amps
|
|
|
|
nice work!
where you get the powered riser ?
Thanks! buyahash.com They have standard length and extended length. Also got the PS splitter from them. The extended length risers are really long though. Would be happy with just an extra 2" from the standard length version. |
|
|
|
Where's the source for DGC-P2Pool? This seems like it should be impossible with vanilla P2Pool.
https://github.com/CartmanSPC/p2poolFiles modified: p2pool/networks.py p2pool/bitcoin/networks.py |
|
|
|
Based off of p2pool 13.3. 6 nodes with one having 98% of the hashrate. Any help would be appreciated. I wouldn't have a clue how to fix it, but it'd probably be helpful to (someone) to know if you're the node that's contributing the 98% Yes. The node contributing 98% is the one that got the error. |
|
|
|
Have received this error twice now. This is on a DGC p2pool with 30 Mh/s. When the error occurs it stops all mining until the pool is restarted.  Based off of p2pool 13.3. 6 nodes with one having 98% of the hashrate. Any help would be appreciated. |
|
|
|
With WDC changing it's specifications p2pool code will need to be updated. The two parts I have identified that need to change are: BLOCK_PERIOD=15, # s targetspacing
change to:
BLOCK_PERIOD=30, # s targetspacing
and... SUBSIDY_FUNC=lambda height: 32*10000000 >> (height + 1)//4147200,
SUBSIDY_FUNC is a bit tricky. Anyone have some code examples from another coin that reduce subsidy similar to how WDC will? New block reward details: Block Reward: 64 coins, reduced 1% every week (20,160 blocks) I suppose it could just be changed to SUBSIDY_FUNC=lambda height: 64*10000000, but that is not optimal. Help? |
|
|
|
beautiful...
but...
=/ never will reach ROI.. free electricity? still difficulty is gonna kill this rig
atlcoins only other avenue..
I agree that GPUs will soon be obsolete for Bitcoin mining but these cases may still be useful for the new series of ASICs coming out. Don't know what the cooling requirements are for this thing but the server enclosures that can house multiple units are prohibitively expensive. Introducing The Monarch Butterfly Labs 28nm Technology Bitcoin Mining Card  The fastest and most power-efficient Bitcoin miner yet 600 GH/s (if you believe their marketing hype) |
|
|
|
That address is coming from a connected node in the WDC p2pool. Currently there are 5 nodes combing their local hash rates to make up the p2pool rate of almost 3 Mh/s. http://wdc.xpool.net:8820/ is contributing about 600 Kh/s to that 3 Mh/s. It's working as it should.  |
|
|
|
Wow, that's beautiful! Make sure to keep us updated on your progress! I ended up reversing the placement of the case in the rack since these MSI GPUs blow most of the air out the front of the cards instead of the back exhaust vents. It does not look like I'll need any additional fans to keep them cool. Thank you for everyone's encouragement. It really helped and pushed me to keep going. If you’re thinking of building your own case go for it! Although it was frustrating at times it turned out to be a fun project.…just make sure you have the right tools |
|
|
|
I have decided to only use 4 GPUs (7970s) per rig. Is there a preferred order of slots that should be used on the Z77A-GD65? I currently have following slots populated all with 1x to 16x powered risers:  The first two GPUs (closest to the CPU) run much hotter than the other two GPU1 73.0C GPU2 75.0C GPU3 65.0C GPU4 62.0C Video is plugged into GPU2. |
|
|
|
|
Q: How do I tune for p2pool?
A: p2pool has very rapid expiration of work and new blocks, it is suggested you decrease your intensity by 1 from your optimal value, and [possibly] set GPU threads to 1. If you have non-p2pool backup pools it is recommended to use --failover-only since the work is effectively like a different block chain.
|
|
|
|
Unlike the previous P2Pool Server List, I will maintain this to my best ability and update this daily if possible at most weekly with new nodes, or any updated changes made by the pool operators.
Thank you for this! There are a few web sites that list p2pools. Here is one of them: http://p2pool.hostv.pl/(...also xpool.net is actually in Southern California) |
|
|
|
A p2pool at this stage is only really for scrypt coins. P2pool doesn't really work well with ASICs such as avalons arm...I have no problem with a p2pool as I've set many up but its only useful for GPUs or CPUs
Try p2pool 13.3 for ASICs. The BTC p2pool quickly increased over 4x last I checked (2-3 weeks ago) after they added ASIC support. |
|
|
|
|
Show Posts
