Quote from: DPoS on October 23, 2013, 01:38:35 AM

Quote from: soy on October 23, 2013, 12:18:59 AM

I see, you tried 0.96 and then 0.94 gave you the .7 volt VRM output voltage?

a hard reset and power cycle brought it back to high voltage again

it is screaming now holding ~525 at BTCGuild (just think if I had the other 59 cores)

10 min mark:

Block: ...fac0e16a #265429 Diff:268M ( 1.92Ph/s) Started: [01:10:24]
ST:2 F:0 NB:3 AS:0 BW:[ 50/ 34 B/s] E:2764.94 I:38.45mBTC/hr BS:141k
4/709 50.0C | 571.0/504.2/506.6Gh/s | A:167 R:10+0(5.6%) HW:6766/6.8%
--------------------------------------------------------------------------------
KNC 0: 46.0C | 142.5/136.8/133.8Gh/s | A: 45 R: 3+0(6.2%) HW:1140/4.4%
KNC 1: 43.0C | 142.2/135.6/133.4Gh/s | A: 41 R: 6+0( 13%) HW:3804/ 12%
KNC 2: 42.0C | 134.8/129.1/134.5Gh/s | A: 45 R: 0+0(none) HW:1112/4.5%
KNC 3: 50.0C | 106.2/102.7/104.3Gh/s | A: 36 R: 1+0(2.7%) HW: 712/3.7%
--------------------------------------------------------------------------------

20 min mark:

Block: ...fac0e16a #265429 Diff:268M ( 1.92Ph/s) Started: [01:10:24]
ST:2 F:0 NB:3 AS:0 BW:[ 48/ 34 B/s] E:2903.21 I:39.55mBTC/hr BS:141k
4/709 50.0C | 549.3/511.4/511.0Gh/s | A:272 R:15+0(5.2%) HW:10175/6.4%
--------------------------------------------------------------------------------
KNC 0: 47.0C | 144.0/138.7/136.6Gh/s | A: 75 R: 4+0(5.1%) HW: 1762/4.2%
KNC 1: 43.0C | 143.8/137.9/137.8Gh/s | A: 67 R: 8+0( 11%) HW: 5858/ 12%
KNC 2: 42.0C | 135.6/131.2/131.7Gh/s | A: 73 R: 2+0(2.7%) HW: 1557/4.0%
KNC 3: 50.0C | 107.7/104.6/105.6Gh/s | A: 58 R: 1+0(1.7%) HW: 1025/3.3%
--------------------------------------------------------------------------------

Observations:

temps are up about 7-10C each board (to be expected)
the one board (KNC 1) that was really low on HW errors when not overclocking, has the highest, while the others are all under 5% Huh

similar amount of cores as last test (gained that one back to 709)

we'll see how this goes after a few hours...since my temps are still lower than many people's non-overclocked jupiters it looks sustainable. (all 4 VRM boards) Dropping the fans to the crossbars does help in these situations

I should change the PSU though, it is only an 850watter

Update: close to an hour and running strong on BTCGuild - 533.87 GH/s 546304 (99.72%) 1536 / 23040 / 0

I'm not sure I'm running the same version of cgminer, 3.4.0 from the top line. My cgminer doesn't note the temp.

Yes, running the earlier firmware now sure makes more sense than trying to avoid the 100 watt difference I saw on my Mercury. It costs me $5.50/month for the extra 100 or so watts while the extra 37.8GH/s is like 3½ Blades. So, do you measure your power at the wall? What do you get at .7v and at .9v?

I don't understand what you mean by drop the fans to the crossbars - removing the front fans and grills completely?
KnC 0: | 140.4G/137.8Gh/s | A:3295472 R:133982 HW:173578 WU:2108.2/m