620-680 mh/s Sapphire 7970 Ghz (1000 - 1170 Mhz)

Maybe stupidity is the greatest power in the Universe, but the greed does not fall far behind...

guys, can you imagine auction:
blessed 7970 +10% to luck

have similar problem, but I also have same problem under Windows.
will try the suggestions, thanks

Ubuntu 12.04 LTS
I liked 10.4 the best, but they stopped supporting it. Anything else is not so easy to use (although if you set up debian you are probably good to go for years without downtime)

Hi!
660 Mhash for ~ 1150 Mhz seems about right.
I think with most of the cards it is hard to reach 1200 Mhz without water cooling. I have 7970 Ghz edition and the most I can get is 1170 at 80 deg C (in open case). With higher clock the card crashes and I think even at this frequency it is sometimes throttling down.
Most of the higher rates reported are problematic to run for long term, so I would accept 660 Mhash. If you need to invest money into cooling, then maybe it is better just to buy another card than max out the one you have.

Sorry guys, nothing has worked. I suspect bad motherboard.

I will try to use some of your advices today, I will let you know guys, thanks!

tried that, but nothing

What I meant is that if there is only one card then PCIe runs with 16x speed. And if there are two then the bandwidth is shared and this is what I see. GPU-z sees one card on PCIe 8x, so I guess, it switched into sharing bandwidth. But the second card cannot be seen - nowhere, not in dev man, now in the GPU-z. I run Gigabyte, they were good so far.

Correct.
Also, only one device is detected in device manager and the same is true for linux (checked with lspci).

I posted this, because I really need fresh ideas how to solve this problem. Unfortunately, I don't have more video cards to experiment with, just those two. Also, I don't have another motherboard which I could use for crossfire.

Both have connectors plugged in. Only Sapphire shows up in Device manager when both are plugged in, though if only PCS is plugged in it shows up without problem. No option for Crossfire in CCC because only one card is detected.

One more thing to everyone. I will try out your suggestions after may 15th and will give out the bounty after that.
Thanks for reading!

Hi
I am banging my head against this problem for a while
I have Sapphire 7970 Ghz Edition Vapor-X
and Powercolor PCS+ Vortex II 7970
Each card mines fine on its own. However if I put two of them into my motherboard (GA990XA-UD3) only Sapphire is recognized.
Funny enough it does not matter which of the two PCIe slots it is in, always Sapphire is seen, not PCS+
Both cards spin fans. I tried a crossfire bridge, but no success. I am putting them into the 16x and 8x slot, so both are running at 8x and it is detected by gpu-z (this is how the mobo works - it is 16x for one card, 2x 8x for two).
I have also tried 1x to 16x riser for one of the card with second in the 16x slot and again, does not matter where, Sapphire is recognized.
That is both Windows and Linux can only see this card and so is cgminer.
I flashed motherboard bios to latest version.
I have latest drivers on Linux and Win. Though I think the other card should at least be seen by the system.
I have only 650W power supply (which itself is a high efficiency one - can output up to 54A of 12V - up to 648W), but I lowered voltage on the CPU and I think when the cards are just booting, they should boot up fine and be recognized and only at full load there may be a problem with throttling.

Thanks!

Any help is appreciated and there is small bounty for successful advice. (I am thinking about 0.2 btc - I am a late btc adopter and I did not manage to become filthy rich yet

I thought I am knowledgable about computers, but this is getting ridiculous

660 - 680 are fairly typical in my experience.
Thought I can throw some specs, so we have something to compare between cards.
I have:
Sapphire 7970 Ghz edition - goes to 1150/1150 ~ 680 Mhash/s temp 74C
Powercolor PCS+ 7970 mines 1100/1100 ~ 640 Mhash/s temp 80C
settings are: intensity 10, shaders 2048, threads 2, lookup gap 2
TRC and BTC mining
I find out that rising intensity past 10 does not do much - hashes go higher, but they fluctuate a lot and it is ~ 660 in the end
Voltage is locked on both cards and I cannot lower mem less than 950 Mhz, it does not seem to do much for me as well.

Funny thing, I cannot set up those cards to mine together - Sapphire always takes over and it is the card which is recognized under win7 or linux, PCS+ is not seen by the system.

If someone can help with this issue I have a small bounty to offer. Uncle google did not help.
The mobo is GA990XA-UD3

If sponsors wonder, this actually works. I frequent http://bitcoinity.org/markets few times daily. Link leading to this auction got me interested enough to brought me here. I didn't know about few of the services and clicked on the links. Well if even participating in the auction helps, think about the what link on mainpage can do. Disclaimer: I am not affiliated with the site in any means, I just find it useful.

Yes for silver and copper. To what extent I do not know (= I do not know what dose to use and how - sprinkle on surface?).

Noteworthy, copper is also used on intrauterine contraceptive devices as well. They are designed to have specific surface area to release copper cations in order to kill (or at least inhibit) sperm and therefore fertilization.

Citrus oils - seems so. Here is one link to abstract about it:
http://www.ncbi.nlm.nih.gov/pubmed/23381618
Though the concentration looks bit high compared to for example antibiotics.

As if to use colloidal silver by ingestion, I would say no. I do not use it. You can turn blue from high doses. Ok, seriously, that's true.
And in the end, all of it would sound differently if someone would try to sell colloidal lead and try to say its beneficial.

--
The answers are for free. However I accept donations, since grad students are not paid so well...
16a1YmEJwR3vZXdKAq65QANMYBdzTGCgiE

I had to check it out. It works! I got elevated counts, nothing spectacular though. Potassium 40K, which is just 0.012% is a long-term radioactive nuclid. But I doubt that it is high enough to kill any bacteria. Those guys are very radiation resistant.
Also, rubidium chloride, which used to be commonly used in molecular biology is even more radioactive.

* toxicity of the salt itself - salts of high molecular weigh metals (copper, silver, gold, cadmium) tend to kill or inhibit microbes, there are also some salts where anions are toxic, like azides or cyanides
* concentration of the salt, also by putting the cell in osmotic stress if the conc. is high enough
* pH of the solution - so how acidic or basic the solution is after dissolving the salt. This effect rather just helps main toxicity of the salt by enforcing additional burden to the microbial cell, which is now outside of its optimal pH range. Also it may help in uptake of the salt.
* specificity of the toxic action of the salt - for example, some salts, just have broad toxic action, like blocking active sites of many enzymes, but with small affinity. So they just bind and then unbind in a form of competitive inhibition. But if you have something which is specific to a certain enzyme, then usually has high affinity and in some cases form covalent bond, then inhibition is permanent and the enzyme is dead. Cyanides for example inhibit an enzyme in the electron transfer chain of mitochondria. But this is only for eukaryotic microbes (am not sure if it is right, you probably meant microbes=bacteria), like amoeba, since those have mitochondria.
But I guess you can put this point in the toxicity point.
*Whether exposure is acute or milder, but prolonged. It is probably better to use high dose and not to increase it gradually, because then microbes do not have time to switch on the compensatory mechanisms (express genes which will regulate uptake of salts).
*Temperature - this may pose an additional burden to the bacterial cell or help in the uptake (I think).
*Presence of other salts or antimicrobial compounds = additional burden to the cell.
*Lack of the food source = no energy for compensatory mechanisms, like active transport of the salts out of the cell.

I can't think of any more, but I bet that with more details about the situation it is possible to find some more factors affecting the antimicrobial action of salts.

Unless someone will find a way to make a somewhat programmable computer, I doubt it will have any impact. And this experiment takes time plus few days of work. If yours steps are enzymatic reactions, then it is going to stay like this. There is no nice scheme yet, where you would just mix everything together wait a bit and readout with sequencing. I saw some papers on DNA cryptography, some interesting ideas, but nothing too practical, nothing which beats computers yet.
I have heard someone is experimenting with DNA for breaking of some form of cryptography, but I did not have time to track this and get details. It may be bogus, or just unsuccessful attempt. If something like this works, then it will come out big.

But I think we are tantalizingly close to storing information in DNA. We are already doing that in our experiments. It just simply costs a bit and you need a company for readout.

For example, I can order a 500 base pair long DNA fragment (1000 bits) of arbitrary sequence. It costs 99$. So I order it, take the tiny drop and put on a piece of paper. I let it to dry and encircle the place where a drop was. I send it to you in a letter, you cut this out, soak in water and send this water to a sequencing company. It costs about 10$ to read out. You will get results in few days. I guess that's a nice example of steganography.
We can also synthesize an oligo library, which is 22 000 fragments of 36 nucleotides each. It costs much more, but it is 1584000 bits, 198000 bytes. Well, close enought to 5.25" floppy disk (320kB). Reading is also pretty costly too.

DNA computing.

Since the DNA and enzymes which work on it are so well studied, there can be used for many interesting manipulations.

DNA does not offer anything better for computing (it is definitely harder to program), except for the numbers. If I get a typical amount of 0.5 ml of 100 micromolar dilution of the oligonucleotide from commercial synthesis company for $10 that's about:

500 microliter * 100 picomol/microliter = 50000 picomol. 1 mol is 6.023*10^23, so: 50000 * 10^-12 * 6.023 * 10^23 = 5 * 6.023 * 10^4 * 10^11 =~ 3 * 10^16 molecules.

First implementation of DNA computing was for a travelling salesman problem, which is NP hard. With growing n=number of cities, the number of possible solutions explodes. For even 7 cities the combination number is 360, it grows faster and faster past that.
Leonard Adleman has used DNA fragments for all possible edges of the graph, with their length being proportional. The nodes were ends of those DNA molecules. DNA is a double helix - has two strands, it is possible at the end for one strand to be shorter than other, like this:

ACTGA
TGACTACGAGA

this part will have interesting property of forming relatively stable interaction with part with compatible end:

     TGCTCTCGACGC
ACTGA::::::GCTGCG
TGACTACGAGA

And this interaction can be later fixed by the reaction called ligation which makes 1 DNA molecule out of those two:

ACTGATGCTCTCGACGC
TGACTACGAGAGCTGCG

Adelman took DNA fragments representing all the edges of the graphs for the 7 cities, mixed them in the tube and they all started to link via compatible ends (which represent cities). Then he did ligation on those fragments and he got DNA fragments representing all the routes possible between cities.

Of course, some of them where much longer, some much shorter. So he had to select ones having all 7 cities.
Now, I mentioned that this link between fragments of DNA represents a city. DNA also has the property that if we separate strands of the double helix, then this strand will look for a compatible (complementary) sequence. If we have small fragments of single stranded DNA complementary to a city attached to a small polystyrene bead, then we can select all the DNA fragments having this city in path (they will stick to the DNA attached to those beads).
So if we do a step by step selection for all seven cities, we will have only fragments with those 7 cities in the path. Now we simply need the shortest one, but that's easy, because there are size selection methods.
So Adelman did all of that in a couple of days and sequenced the shortest strand, and voila, he had the optimal solution to the problem.

Of course it is not so practical, but that was first proof of principle demonstration.