Is there any significant (>1%) performance difference between 20130703 and 20130819 (or 20)? I'm trying to keep the reflashing to a minimum.
Are there changelogs somewhere for ckolivas' firmwares?

I finally received my first unit. It took one week to get here from Hong Kong via DHL (with 4 days in customs...) and the next difficulty jump is going to be quite small, I might even earn a buck or two.

I just got a tracking code for my Feb 2nd order. The status of the second order (Feb 18th) is still processing. It has been a long wait. Thanks for the awesome peer support in this thread!

[/Yes, but that poll has been there like six months and the situation was very different back then.

With the recent exchange rate rise, the 0.25 BTC ticket is becoming very expensive. Many new adopters do no want to bet with such a large amount and I think it will start affecting the jackpot (even though historically this hasn't been the case). Cheaper tickets might also bring new players to this game and establishing a large customer base is the key factor in this type of game.

http://i.imgur.com/3JlGV.png

High resolution version is available.

1GcPZVP5LduGnipnbnAauJ4k9BTgvrBRtS

Thanks! 

How long does it take get this fixed? I just saw over 5000 shares disappear from my worker summary.

Same problem here. JSON prints out the same info too. Miners show no error.

Sure there are cheaper boards if x1 is enough. If 3 PCIe paths are enough, the cheapest I could find was ASRock I-H61 S1155 H61M-S which has one x16 and two x1 and that costs under 45€. For 4 PCIe paths the cheapest was Asus P8H61-M PRO Intel H61 Rev.3 (two x16 and two x1) for under 70€. The prices are from www.mindfactory.de. You might find them even cheaper on eBay or something.

EDIT: Though, they are both on H61 chipset which doesn't have SATA3, USB3 or for that matter anything else interesting.

There is no general rule. It mostly depends on the architecture and memory technology used.  In heavy scientific calculations best worksize is usually the one that the card can process natively but in mining where a single loop is very simple and fast the optimal worksize can vary. In mining lowering memory clocks saves power and therefore may allow for extra OC on the core thus speeding up computation. If you lower your memory clocks too much it can lower your processing power but this kind of loss can be compensated by lowering worksize.

So without solid background in high speed computation architectures the fastest way to know is trying out all possible combinations.

Works very well. I am getting over 440 Mhash/s on HD 5870 (1000/375) with -k phatk AGGRESSION=13 WORKSIZE=256 VECTORS BFI_INT and about 416 Mhash/s on poclbm. However my other HD 5870 running at 950/375 with same switches only hashes about 410 MHash/s with phatk while poclbm gives about 400MHash/s.

Running RealVNC on Windows Server 2008R2 and RHEL. Viewer works great on WinMo too.

How many GPUs are you running? Are they all mining? How much CPU are they hogging? You could also try putting the miner in verbose mode (with -v flag) to see any possible errors.

I think vectors are defined just "VECTORS" not "VECTORS=on". That might solve the performance issue.