The Jamaicans have a common saying for this scenario...

"Soon come"

How is replacing "BTC" with "USD" any different than the status quo?  Unless you lie about it, which you could just as easily do with cash.

While you can certainly trade with people not using US Dollars (aka barter), and likely get away with that for a long time, eventually the IRS is going to want their cut of the dollar value of the bartering.

How is bitcoin any different?  Perhaps a little easier to conceal, and that's it.  If you are having problems concealing your giant phat piles of illicit cash from your illegal raw milk operation, then that market got to be a whole lot bigger than I remember.  I think the whole act of you giving product (milk) in exchange for consideration (bitcoin) makes it a sale.

If you want to pretend it's not an actual transaction - do so - but bitcoin really has nothing to do with it.

Had the same thing happen to me selling one out of two 6990's.  One sold fine, the other sold to a zero feedback guy who then tried to put in an ebay complaint stating I never shipped the item.  Still had to jump through hoops, but at least eBay got the resolution right...

-Phil

Hey,

Anyone happen to be around @ Defcon who can change $300 cash into the bitcoin equivalent?

Call it $9/coin, so $300 for 33.33 bitcoins.

I'll buy you a drink @ Masquerade Bar for your trouble!

please e-mail phil21@gmail.com or find me on AIM @ philtwoone if you can make the dream happen.


-Phil

How many days are you buying?  What are payment terms exactly?

I could possibly offer ~7Ghash if the term is long enough.  This would come with remote reboot and IP KVM access (e.g. install whatever you want on the boxes via remote USB media support), but I'm not interested in much less than 30-60 day "leases" at a time as it won't be worth the pita factor

I think everyone answered the theoretical part of this question sufficiently   The heat generated by a computer is going to be identical to a coil-based electric heater.

However.. What has not been touched on is perceived warmth.  You may be a bachelor, so this might not matter much to you... But those with wives and girlfriends will know how this works!  Assuming you have an identical power draw on both your furnace and mining rigs, I would still say you are very likely going to have to augment with the heater.

A furnace designed to create heat and move air (in theory in an efficient manner) is likely going to "feel" a lot warmer than an mining rig of the same wattage.  This is simply due to air movement and distribution throughout the house.

One thing I did was stuff my mining rigs in the furnace room, and open the cold-air return duct so they blow warm air into it.  I then set the furnace to just run the fan 24x7, even if the burner (gas here) isn't running.  End result was an extremely hot laundry room, and a slightly reduced heating bill.  I run around 6kw worth of gear, but honestly have no clue how many BTU my furnace is rated at.  I did ponder stuffing mining rigs into the cold air ducting itself (right before the burner intake) - but then realized the fire risk probably wasn't worth it!  Note to anyone contemplating this (or running wire through a heating duct) - smoke is the most hazardous part of a household fire, and putting things that burn toxic fumes into your air vent system to ensure all the fumes get quickly distributed throughout the house is an extremely poor idea.

Now, had I designed the house ground-up to be "mining rig heated" then this would be a different story!  It's very easy to neglect air distribution and simply focus on the raw math though, and modern HVAC systems are in general very efficient at taking created heat and distributing it to where it's needed.  Mining rigs not so much.

-Phil

And here I was trying to proactively head off someone being pedantic

What is it converted to?  Ignoring LED's and fans of course.

I can't remember who said it... But the quote below comes immediately to mind.

"A man may fail many times, but he is not a failure until he begins to blame someone else"

Probably interested in a Qty order should some projects finish up soon here.

Do you have access to any other Delta fan models?  I wouldn't mind a few of the 200+ CFM models to play around with for a current project

-Phil

Yep, pretty much.  Here is one new set of 6990's under burn-in, about 36 hours of data:
[ALL (30s):1686.7  (avg):1686.9 Mh/s] [Q:18011  A:41153  R:270  HW:0  E:228%  U:22.95/m]
GPU 0: [418.6 / 419.4 Mh/s] [Q:3558  A:10141  R:58  HW:0  E:285%  U:5.69/m]
GPU 1: [424.4 / 423.9 Mh/s] [Q:4674  A:10368  R:72  HW:0  E:222%  U:5.81/m]
GPU 2: [420.1 / 419.6 Mh/s] [Q:3602  A:10050  R:71  HW:0  E:279%  U:5.63/m]
GPU 3: [423.7 / 424.0 Mh/s] [Q:4639  A:10361  R:67  HW:0  E:223%  U:5.81/m]

The cores are clocked at 935, 945, 935, and 945 respectively.  So an average of 421Mhash/core I guess (the second number is the total average since it started up).  Once put into "production" it will be in a cooler place with more airflow, so I'd expect clocks to go up 5-10mhz a core.  These two cards are performing better than average so far, but nothing exceptional.  Memory clocks are 150Mhz across the board.

Temps are:
Adapter 0 - AMD Radeon HD 6990
            Sensor 0: Temperature - 78.50 C
Adapter 1 - AMD Radeon HD 6990
            Sensor 0: Temperature - 74.50 C
Adapter 2 - AMD Radeon HD 6990
            Sensor 0: Temperature - 80.00 C
Adapter 3 - AMD Radeon HD 6990
            Sensor 0: Temperature - 75.50 C

No special cooling here - just running in a cool (68F ambient) basement, inside a large full ATX case.  This is used for burn-in, then the cards are put in a caseless environment with far more airflow on them.  I do run the fans @ 100% 24/7 - I have had good luck so far, but others report early fan death doing so.  Either way, I expect to swap out fans every 9-12 months in these, but have yet to have to do so.

The kill-a-watt has held steady at 880w, using a 1000W Kingwin PSU.  Prior to the memory underclocking, ramping up all GPU's at default clockspeed read 1090w before it crashed, so memory underclocking = will save you money on power, and PSU's!  Do it!

These temps have held steady and are typical.  When in production, this is also targeted at 85C per core, vs. the 80C/core during burn-in.  Some cores simply underperform/run hot of course, especially some of the earlier run units.  I have yet to see higher clocks than 955Mhz be stable, but I also don't spend a whole lot of time tuning this area.

There is lots of random old information/mis-information about these cards out there.  Wanted to get some actual data put forth based on documented data.

Hope it helps someone!

BIOS Flash works perfectly.

You just can't boot the cards in Windows - the ATI driver will call a bluescreen on purpose, halting the machine.  Annoying.

When I get 6990's in, I flash them to 150mhz minimum RAM clock, and they tend to run anywhere from 915-950mhz at around 85C.  Some cores as mentioned by another poster in this thread just are horrible, such as one that maxes out at 910Mhz @ a running temp of 98C.  I'd say average is around 940mhz @ 89C - if you like pushing your gear that hard at least.  There is a lot of variation on these, but it seems cards of later manufacture do better than the earlier ones. 

I have not seen a difference clocking RAM higher than 150Mhz, but others report otherwise.  YMMV.

You absolutely *must* underclock RAM on these to get any form of thermal room for stable overclocks, I've found.  Also it saves about 50-75w per card.  I had issues keeping them stable prior to flashing them.

fwiw I expect 840Mhash out of each card, and tend to average 850ish it seems these days.

-Phil

Having operated hundreds of servers with the Netburst architecture (e.g. Pentium 4 based single cores, dual cores, and xeons) I can assure you load tdp is FAR different than idle tdp

Hey,

Great work on this!  Takes pretty much all the features I had hacked into/worked around/etc. w/ Phoenix, and integrated them nicely.  Currently in the process of rolling out cgminer across the whole cluster.

One quickie... Right now, I do graph mHash/sec per GPU, and have various alarms that trigger.  While options exist to get per-gpu stats as-is (e.g. --debug), they are not optimal.

Right now the interval status line logging works great for me - it would be perfect if an option existed to also list the latest GPU stats as well.  Something like:

[2011-08-18 02:12:17] [ALL (5s):508.0  (avg):536.9 Mh/s] [Q:5  A:0  R:0  HW:0  E:0%  U:0.00/m]
[2011-08-18 02:12:17] [GPU0 (5s):204.0  (avg):214.9 Mh/s] [Q:5  A:0  R:0  HW:0  E:0%  U:0.00/m]
[2011-08-18 02:12:17] [GPU1 (5s):204.0  (avg):217.9 Mh/s] [Q:5  A:0  R:0  HW:0  E:0%  U:0.00/m]

I figure this would save me an hour or two of hacking things together, so I'd be happy to "donate" 2btc for this feature

Also, with --verbose on, and logging stderr to a logfile, I'm getting a lot of:
[2011-08-18 02:20:49] X-Roll-Ntime found 

Does this log entry have a purpose/something I should be looking at?

Additionally, the periodic GPU stats could be removed from --verbose, replaced with the above feature - it would keep the log output consistent nicely I believe.

-Phil

I've been pondering this with some degree of seriousness.  Especially since I just had a 55 gal fish tank become free for this use.

It basically comes down to if I expect to be able to resell the components or not at this point.  Tough call!

-Phil

I run a dual 6990 box, running Ubuntu

It's generally stable (until I push them a little too hard of course...)

You will have to flash the bios on these cards in order to underclock the memory to any significant degree.  This helps.  A lot.  Your heat load will go down considerably.

Instructions are on the interwebs, but essentially create a DOS boot USB with atiflash.exe on it - back up all 4 GPU's (2 gpu's per card) and MAKE SURE you know which bios goes with which.  It matters.

Then move that USB stick over to a windows box, run the radeon bios editor, and set all clocks down to whatever you like for every power profile (I set mine to 150 myself, but you can do whatever).  Save these in a manner so you know which is which.

Back to the linux box.. reboot on dos USB stick, use atiflash to flash your new bios across each GPU.  Reboot into linux, run aticonfig --initial -f --adapter=all and you should now be able to set your memory clocks to whatever you like (after you reboot one last time).

I can PM you my simple .bat files and such if you like, and even give you my modified BIOS images (however, I would suggest you modify your own as there is no guarantee one bios from my card will work from yours).  Just let me know.  No guarantees you won't brick your cards though, of course!

-Phil

*looks* amazing, and I love the idea of a US-based exchange for those who do not wish or need to remain anonymous.  Perfect for some!

Now lets see if the trading engine and backend is up to the task

This   Nothing to worry about, you'll be safe until either you or bitcoin dies... Whichever comes first

Who is he supposed to return them to?  A currently completely owned MtGox system?  The original "hacked account owner" who has not been identified?  Someone else?  I know if I held them and wanted to give them back I'd have no clue who the rightful owner is at this time.

That seems pretty key to me

People completely lack reading comprehension.

Again.  For the zillionth time.

~250k bitcoins are with mtgox, they never left.
~650 bitcoins were transferred out by Kevin, and are now in a clearcoin account which is acting as escrow until this is all figured out


Why is this so hard to understand?  I really don't get it...