Topic: [Guide] Dogie's Comprehensive ASICMiner Blade Setup - page 60. (Read 580761 times)

For what it's worth, I haven't experienced any significant efficiency difference between using native getwork on a pool with diff=8, or using mining_proxy.exe to proxy getwork → stratum. Using getwork directly on a pool with diff=8, I get about 96%-97% efficiency (measured at BTC Guild). Proxying getwork to stratum shows near 100% efficiency on the Blade counter, it also shows 98%-99% efficiency on BTC Guild *BUT* the total number of accepted shares per hour stays about the same in both cases. I think that is explained by the fact that the stratum proxy sometimes accepts getwork shares that aren't sent back to the pool using stratum, because they are detected to be stale already (just my thought, maybe it's another reason).

At any rate, I hope the next generation firmwares will have stratum (and/or GBT) available natively.

You can configure the proxy to use any Stratum pool, check the documentation.

are there other stratum mining proxy which blade can use it?

     Excellent. Way to go!

Started Slush but it was being strange. Pool luck was at like 70% consistently for over a week, which statistically is pretty unlikely. They kept saying there was no problem but I left anyway. BTCGuild gives slightly higher hash rates due to EU servers and better graphing/idle notification.

Too bad Which pool you're mining at?

Just to note: hashing speed is nothing to do with the voltage as far as all chips are working (displays OOOOOOOO in web-interface).

Found a block with 26GH within a week, then another with 52 after another 3 weeks. Final one after another week. I was mining on pool though, so I earn't nothing.

When I started the blades the voltages were all around 1.206V. Then I tried a little higher then higher still and now both blades are at 1.282 and 241Watts. The wattage increased by only 2W from the original 239 Watts measured at the brick plug into the killawhat . I now have one blade on btc guild getwork at ~12.7 GH/s and the other blade on slush stratum via proxy hashing at ~13 GH/s. My next step is to slowly reduce the voltages to see where is the max hashing at ~240 Watts.

J/

How long it was from the 1st block found and how much hashing power do you use in solo?

Found my 3rd block. Would be 280% my pool earnings :/

Glad to see you are running well finally. Enjoy securing the Bitcoin network

Update on the most annoying blade on the world:

Gave it it's own PSU so there is no more sharing of lines with my ESXi server, and received my Procurve switch from the bitcoinstore:



Vindication is mine! And wintame's as well for helping me troubleshoot this thing.

There is no reason to have the voltages so high, 1.2 is is fine. Any more is just wasting power, creating more heat and reducing life span.

Also get a stratum proxy set up.

All:

I thought you'd like to know the stats of these new blades. After 1 hour of operation (no proxy, will test proxies later on with one of the blades).
On both blades I have the voltage at 1.262 at full work load. Total brick amp at 2.08 A and 239 Watts stable. I have a large dual fan at the back (blades vertical) blowing on the chips and one 6" fan on half of each heat sink. Heat sinks are only slightly warm on the non ventilated side where I keep track of the temp by back of the hand feel. All seems to be in great working order. Please let me know if 1.262 is pushing too much.

Blade 1 (Margie):

Total MHS:    12833
Received:    0000014731
Accepted:    0000014625
Per Minute:    175.39
Efficiency:    099.28%
Up Time:    0d,01h,23m,23s

Current Server: btcguild.com:8332
Clock selected: High

Blade 2 (Lucy):

Total MHS:    13007
Received:    0000011295
Accepted:    0000011235
Per Minute:    177.72
Efficiency:    099.46%
Up Time:    0d,01h,03m,13s

Current Server: btcguild.com:8332
Clock selected: High

Total hashing at 25840 MH/s. Awesome!!!

Margie and Lucy ain't going nowhere anytime soon!

Many many thanks to forsetifox at irc #ozcoin,  Friedcat and Dogie for all your support and tips. I could not have overcome all of the issues without you all.

For those having problems between the blades and a local host, make sure the ip of the local host (say your pc) matches the ip of the blades to the second last portion of the ip. Example:

if the mask is 250.250.250.0  and the blade ip is aa.bb.cc.xx then the ip of the local host needs to be aa.bb.cc.yy. Only the last 2 entries (xx and yy) are different.


J/

Yeah, works fine for me.  I do use a /24 though, haven't tried any other length than that.

OK, thanks. I must have missed something.

I run it on other than default subnet with /24 mask.

Yes, I ran mine in 192.168.0/24 just fine - didn't try any other netmask though.

Has anyone successfully operated these blades in a subnet other than 192.168.1.0/24?

I was unable to switch to other subnets (tried 192.168.15.0/24 and 172.25.100.0/16), so I am just assuming that this is another one of the quirks of the communications firmware.