Hey everyone, back from my little vacation.
I'm still a few pages back in this thread, so not fully up to date on the discussion but I'll chime in quickly here, lunch is over and I'm not going to get caught up till this evening.
First thing I'd like to note, in regards to 2600k's capability.. One of the reason I'm always near the top of the cat 1 is not only a fairly optimized setup, but consistency. I'm not sure how much that has been taken into account. I came home from my holiday to see my internet connection down, and has been for about 36 hours now, I've missed the deadline on the hugeadv I had so this is going to be a major hit for this month. I spoke to the ISP a little while ago, they don't know what's up and can't help me while at work.. Now, until this mishap my 2600k TC rig has been on, and folding uninterrupted for over 3 months now. No restarts, not even hooking a keyboard/mouse back up to check temps. My rig is 100% rock solid stable, and consistent. I like to lurk the other TC competitors HFM's from time to time, and occasionally you will see their rigs off, or have been restarted, whatever... they aren't running as consistently which costs points over the course of a month. Just something to keep in mind.
On the bloomie situation, yes I managed 51K out of both of mine, which was native. IIRC in a VM they topped out around 49K. For clarification, that was with p6903 in both cases. Now, 6904 is a different story, I do not believe they can reliably fold 6904's within deadline requirements to allow leaving the corefix in place. I picked up one and ended up folding it through without receiving any bonus points. I don't think it would be wise to assume bloomfield folders can keep the corefix active and risk pulling 6904's since it does not appear the memory allocation trick is effective in blocking them. Maybe, just maybe with a very strong OC, high end W/C you might be able to pull it off, but as I have seen being debated here I'm not sure that's a reasonable expectation for all of us. Finally, on the bloomfield situation, to run at my current levels with them is requiring almost double the wattage. One of the key reasons I was keen on switching from my 930 to 2600K for the TC was to save on electric costs. If the decision is made and I am asked to take my 2600K offline I'm not too sure I would want to go back to the 930 on a 24/7 basis.
That's all I can manage right now, I do plan on doing some more benchmarks on the 930's for you to see if I can find a better way of doing things.
Well, that means the only Bloomfield numbers we have are deemed unreasonable by the person who got them. I guess we'll wait and see what Lutro0 gets, and if he can actually complete a -hugeadv. If not - or if it's marginal - I think we might have to base Bloomfield off -bigadv only.