One of my 7970s has developed a rather annoying trait, after a few hours it will drop to 1/2 to 1/3rd the normal hash rate. If I open up MSI AB, move the "power target" slider and hit apply the hash rate will go back to normal for a few hours. Doesn't matter if it's at +1%, +10% or +25%, after a while the card will throttle again. This is at 62C core temp at 0.98V undervolted by the way. Not like it matters, it does the same at any speed and voltage. The other cards in the same rig keep hashing on just fine without needing the power target reset.

It's a non-ref Sapphire card (blue PCB) anyone ever had something like this?

multipool works well here, not tried any others though.

83

So that's why all my miners switched to a backup LTC pool. ETA for a solution?

100 BTC would be a small sacrifice to convince people into your scam. Not saying it IS a scam...if they're legit they're likely to get a lot more orders by this kind of "advertising" than they'll lose.

I don't believe they can do it, but the resolution date is a bit too long, I'll probably make more trading my coins than waiting for the bet to resolve...especially if someone bets a bunch of coins shortly after I bet the return just isn't worth it. Even though I'm 99% convinced they'll be nowhere near their claimed efficiency.

Holidays in Europe and Russia, spring weather,...I'd expect it to pick up from monday onwards.

Many people have the same issue and there's been no response from the owner. Bit shady or at the very least not acceptable to have this going on for so long with no solution or even acknowledgement of a problem.

I bailed to another pool some time ago, suggest you do the same. Plenty of other pools out there.

Ditto.

And the issue predates the BTCe thing by almost a day for me.

Sent some FTC to BTC-e 1.5 hours ago, 0 confirmations

I'm getting like half the coins I should be getting...hashrate is stable on the graph at 3200 kh/s average, the past 3 hours my account has gained 9 coins. Should be making closer 20 coins every 3 hours. It's been that slow for at least 15 hours, the pool luck can't be that bad?

^ my 7970s do 350 kh/s with those settings lol....scrypt is weird there is some strange thing going on where individual components need different ideal settings even though they should be the same from a software level.

One of my cards has me puzzled. It's a 7950 on a 7970 PCB (Sapphire OC). I can mine 1150 core at 0.94V actual and memory goes to 1700+ no problem but I can't get the expected hashrate out of it.

I have no such problems with my 7970s...3 different brands all performing as expected (750kh at 1088/1875, 720 at 1015/1750 and one 660 at 956/1650)

The 7950 is sharing a mainboard with a 7970 but I've tried singling the card out with settings but it just won't do top hash rates. At 1150/1640 it manages about 580 kh/s but I've clocked it down slightly to 1130/1620 since the temps went up. Gives me 565 ish. I've tried lower clocks (for memory especially) any number I could think of but I don't get anywhere near as good as I've seen some people do. I'm using cgminer (latest and previous) with 8192 concurrency g2 i13. Higher concurrencies did not gain anything, most of them lost quite a bit.

Anyone have one of these particular cards with better results?

That just doesn't happen. A lot of hashpower jumped onto FTC simply because it's more profitable. I know I made 300% more LTC mining FTC than mining straight LTC. The supply of coins is the same whether they are expensive or cheap. How many people mine them (so how they get distributed) is determined by the price. But the price does not follow mining hashrate.

Obviously, but given the fact that a regular Jalapeno has 2 functioning chips and this one doesn't...they would either have to have no good chips left to use or have soldered on the dud and found out it didn't work only after that. It didn't appear to be an intentional move to create an inferior Jalapeno and ship that out. In fact, I don't really understand why they shipped that one at all. Why not spend a few minutes (based on the 400 units a day lol) to make a properly functioning one and ship that instead? They would've spent time changing the firmware on the semi dud one anyway.
It's pretty clear that they only discovered the issue when they powered it on the first time. Likely conclusion, their chips are not binned or tested.

Pretty sure I read that BFL themselves told the dev who got it that they had disabled a faulty chip and clocked the other higher. Ended up doing 4 GH/s. Other people have received Jalapenos with both chips functioning and doing the advertised hashrate of 5 or better.

edit: https://forums.butterflylabs.com/hardware-discussion/1983-grnbrgs-unboxing-teardown-new-bitforce-5-gh-s-sc-miner.html

Didn't one of the guys who posted about receiving a Jalapeno get one with lower hashing power because one chip turned out to be a dud so they disabled it and clocked the other one higher but falling short of target?
Nothing I've seen from BFL indicates they have any clue about their chips and simply solder them on and see what happens. Gives you a bit of a clue how "streamlined" their production process is at this stage. Hence the very low volume.

I don't think BFL knows which chips are good and which are poor until they solder them onto a board.

There's a million different scenarios, that's true. I have no clue on the middle term outlook at all. But short term...everybody is playing "guess the bottom" with $$$$ ready to jump onto those cheap coins.

They might all go broke (well it's a zero sum game obviously) in a month, but come today and tomorrow the above scenarios are most likely in my humble opinion.

The real question is what will happen when Gox resumes at 02:00 GMT.

a) since people can cancel open orders, it's fair to assume that most of the bids that were made will be cancelled. They were at much higher prices than what the other markets are at now. But there will be plenty of people who still want to sell at that higher price so they leave their orders in. Result? All sellers no buyers, very quick very deep crash with a big bounce up straight after.

b) Gox users have looked at other exchanges and noticed that the sky hasn't fallen, bitcoin isn't worthless they are still going for way more than they were a few weeks ago. Price will stabilize quickly and move upwards.


Going by what I've seen today, I'd put my money on option a. If Gox' servers couldn't handle a regular panic with bots fighting each other then this is going to be even worse. Data will be delayed creating a negative feedback loop.

Thing is if you're "losing" power somewhere, that must be converted into heat. If the device is pulling 190 Watts instead of 60, it should be fairly obvious exactly where that energy is being lost. You can't lose that much to resistance in traces, the whole thing would simply melt and break down after a second or 2 of running. What you do get is a voltage drop the more current you're trying to cram through, but most of that power has to be lost to leakage in the chip, the power circuitry and DC-DC converters are unable to dissipate that kind of heat.

Either BFL are stupid or they are playing stupid. Any engineer working on this would have known exactly where the power is going within 5 minutes of turning it on.

Reminds me of that GH/s vs power consumption of the BFL chip chart I saw posted once (not sure if it was authored by BFL)....it was LINEAR. I'm sorry but it doesn't work like that. There will be a sweet spot in terms of efficiency somewhere well below the maximum clock/hash rate, anything more than that will start to use exponentially more power (as anyone who uses GPUs knows).

(been lurking here a long time, just wasn't able to post)