You have to ask Windows 7 Pro X64, both are 6870s. 1st is a Reference Visiontek, 2ns is the XFX Black Edition (940 core 1150 mem stock). Yes Crossfired. GPU load dips juuuuust a little when the Mh/s dips. Since I set them to -f 120 to lower the dip, th 1st is a dead 98% constant, while the 2nd varies between 96 and 98, but stays at 97 most of the time. Thanks in advance for the help I know of someone who had a somehow similar problem: with crossfired card, the slave card wouldn't run at 99% load on -f1, but fluctuate instead. Taking off the crossfire bridge and running the 2nd card independently fixed it. I guess you could try and test both your cards running at -f1. If the slave won't run at max load then you know where the problem is coming from.
|
|
|
Thanks for the clarifications and explaining guys.... Now we can focus on my multi GPU problems? well give us details then. What OS are you on, what are the cards, are they crossfired, what are your gpu loads on both?
|
|
|
you need to reread my post smart guy....
Oh I got it. you need to reinterpret my amusement Comon man, can't you get angry n stuff, i was hoping for a flame war
|
|
|
yet all of their effort combined for a century could not drive the price of gold to $1400/oz.
LOL you need to reread my post smart guy....
|
|
|
What's this obsession people have with the concept that a commodity can't hold value unless some merchants endorse it? If the price of gold was determined by its actual use in the industry it would barely be worth over $50/oz. There are countless merchants that will buy and sell gold under various forms yet all of their effort combined for a century could not drive the price of gold to $1400/oz.
|
|
|
I think in the mid term a smart phone compatible Bitcoin client and the multiplication of public wifi nodes would be a first solution to this problem. It will certainly need a more secure system later on though.
|
|
|
That's kinda how pools work in paying their miners. If you don't want to set up such a system, I guess you could make an account for person B with his address to receive the rewards and run some of your mining power for it.
|
|
|
Ok Mansy! I didn't have to look very far to sense that BitCoin does not understand the management of an exchange medium. Considering the first thing you learn about Bitcoin is that it is decentralized, I guess this guy is a troll and I fed him. Mea culpa.
|
|
|
I agree with you goatpig. I said from the beginning that price-wise the 5870 is more economical.
Indeed. Anyways, do you have extra data to share on the 6990, like power consumption, how low can the ram be underclocked and what not?
|
|
|
Although single gpu cards o/c much better, so I wonder if a 6990 combo could beat four 5870.
Of course it would. Bitcoin mining scales linearly, so a pair of 6990 at 880MHz would achieve 1446 Mhash/s, compared to 1328 Mhash/s for four 5870. It is kind of pointless to argue which one would be faster when overclocked beyond the factory setting, because the overclocking room varies so much from card to card, and with the type of cooling (air, water, etc). Are you certain? I'm almost positive two 5870s will beat a single 6990 by a good margin at pure hashing power. Have you seen a 6990 hit over 700Mhash/s? a question on that subject was brought up, so i answered it. The point was that even though the 5870 is the worst Mhash/W ratio of the 3 choices mentionned, it's the best Mhash/$, partly cause it can overclock so well when compared to dual gpu cards. This is relevant to the choice of cards imo. Now, let's admit that four 5870's hash like two 6990, it would cost $700~800 to get the 5870s off of ebay, and $1400 for the cheapest 6990's out there. At worst this is a $600 difference. Assuming the 5870's run at 200W/h and the 6990 at 250W/h (im taking large figures here on purpose) you'd be at a 300W deficit per hour with the 5870's. At an expensive $0.20/kW(exagerating figures again), you'd be paying an extra $1.44 electricity per day compared to the 6990's. At this rate, the 6990's will close the gap in 416 days, or a little short of a year and 2 months. I'd take the 5870's, not to mention that you can build up to it slowly with the profit generated by cumulative card purchase.
|
|
|
5870's can squeeze 360mhash each with an average o/c on stock voltage.
Apples vs. oranges. The 6990 too can be overclocked. Although single gpu cards o/c much better, so I wonder if a 6990 combo could beat four 5870.
|
|
|
Are you certain? I'm almost positive two 5870s will beat a single 6990 by a good margin at pure hashing power. Have you seen a 6990 hit over 700Mhash/s?
One 6990 beats two 5870 because the VLIW4 architecture is easier to fully exploit than VLIW5 arch. I benchmarked my miner, hdminer: - One 6990 (switch at position 2, ie. 830MHz): 682 Mhash/s
- One 6990 (switch at position 1, ie. 880MHz): 723 Mhash/s
- Two 5870: 664 Mhash/s (332 Mhash/s each)
However, price-wise it makes more sense to buy two 5870. 5870's can squeeze 360mhash each with an average o/c on stock voltage.
|
|
|
the only way to fill room is to schedule game. at least, for now.
name the blinds and the time, let's see what happens
|
|
|
There seems to be a lot of that going around lately. This is one of the reasons I believe it would be a great idea to have our own stackexchange site, specifically for answering questions in a dedicated environment. For one thing, it would clearly distinguish the consistent non-stompers for the newbies' benefits. To anyone who hasn't already subscribed to the proposal, please do that. Let's make a distinction here. There are brainwashed central banking and debt based economy lovers who come over this forum to "enlighten" us with whatever rationalization of this wretched system they came up with to make it all fit in their head. Those people don't intend to use Bitcoins nor do they understand it. Those you can bash. Then there are newbies, who have a genuine interest in Bitcoin and are trying to understand it. Those you should be nice to.
|
|
|
The 6990's Power consumption after mem clock reduction, is lower than that of a 5970 after reduction
2* 5870 takes 424W, 1* 5970 = 304W, 1* 6990 = 338W. This is without downclocking mem. The 6990, with it's massive 4GB and higher mem clock, has more benefits from the downclock, hence, less power usage than that of a 5970.
A 6990 outperforms a 5970, so I'm guessing it outperforms 2* 5870s too. If not, then it isn't far. Best $/Mh is still the 5970. I'd still get 6990s as I find investing in old tech for something that isn't for sure is not smart, at least the 6990 has resell value and better usability in other applications
care linking your source? I've got different numbers: http://benchmarkreviews.com/index.php?option=com_content&task=view&id=706&Itemid=72&limit=1&limitstart=165970 on load = 299W 6990 on load = 350W This indeed doesn't take in account the power saved through ram underclocking, and i think it is quite possible to get the 6990 to run in the lower 3xx, but I doubt it'll fetch below 300. I don't know how faster the 6990 is compared to the 5970, but I wonder if it justifies the $700 price tag considering you can get 5970s on ebay for less than $450 shipped. At any rate, you can get three 5970 for the price of two 6990. Better $/MHash for sure. As for four 5870, they'll definitely draw more power than two 6990. Although single gpu cards o/c much better, so I wonder if a 6990 combo could beat four 5870.
|
|
|
Roger Mitchell claims that the United States is Monetarily Sovereign and therefore can fund unlimited programs because it can always print more money. http://rodgermmitchell.wordpress.com/2010/08/13/monetarily-sovereign-the-key-to-understanding-economics/My understanding of the way the US "prints" money is that it actually sells US Treasury bonds to the Federal Reserve, and therefore at some point has to pay back those bonds with interest. If this is true, this guy is completely nuts or just doesn't understand basic economics. Thoughts? US Treasury bonds that the congress "sells" to the Fed don't hold a debt, since even though there is interest on it, the Fed returns that interest back to the Treasury. If the congress wants to reduce the outstanding amount of bonds in the economy, it buys a few back from the Fed, but interests aren't involved. There is an ongoing scam on the selling and buying of US Treasury bonds to individuals though, on which the government has to pay interest. The Fed can't just print money on its own accord, but they get to control the interest rate of the (fake) money they loan out to the financial institutions. "Monetary sovereignty" holds no power. The government has 3 ways of raising funds: tax money, outright printing money through the fed, or borrowing on the market, by selling Treasury bonds to whoever wishes to buy them and paying interests on it to the holder. This "monetary sovereignty" delusion is just a scam when you understand those points: Any interests collected by the Fed is returned to the treasury, and taxes are an interest free supply of money, which means the only source of outstanding Federal debt is from loans taken on the market. Why would a country with the power to print as much money as it wants borrow so much money on the market that the interest on it will exceed its GDP? Certainly if they can print money, they don't need to borrow this much, do they? Hell, why would the government even need taxes if it can just make all the dollars it needs out of thin air? Simply because if the government printed every dollar it needed, the inflation would destroy the economy. To prevent that, they borrow already existing money. The problem is, when the interest becomes too big, they start inflating the monetary mass to ease up the payments. That so called monetary sovereignty is not only powerless at keeping the economy healthy, it gives power to the government to largely exceed its outstanding funds by taking a huge dump on the economy.
|
|
|
Credit Unions ... Nice .. FRACTIONAL BANKING ALL OVER AGAIN .. the bubble begins, slavery and death to all
Stupid idea .. that's why we use BTC, Gold, Silver ... if you don't like it ... go back to Fiat.
You don't understand what fractional banking is... Nor do you understand why it's a problem. It's common sense to take a collateral for a loan. This may require the transaction to be done in the open. From my stand point, a Bitcoin CU would operate the exact same as a regular CU, only that it lends Bitcoin. Certain financial practices require for the parties involved to be known. The anonymous aspect of Bitcoin is a plus, but not a necessity in itself. And frankly, the anonymous part is really icing on the cake. As long as your funds can't be ceased or frozen, you're exposing yourself to much less risks by dealing in the open.
|
|
|
If I was a pool operator and another operator came into my thread and offered this up I would see it as hostile and be apprehensive also. He has his own pool. Even if is stats aren't constantly updated, anyone can calculate the sweet spot with some accuracy and try it out. There is no reason to repeatedly challenge another operator to exploiting their pool at the detriment of their users either way. That was adding fuel to the fire at a time when the whole conversation was going downhill.
slush didn't just walk in there saying "hey ima assault your pool". He defended his score system, that FairUser and Geebus presented as unfair. He exposed how it was mathematically yielding the same result as a share based pool, and how it protected against Raulo's attack. Now you don't see slush saying how he'll attack Tycho's pool to prove share based distribution vulnerable, simply because you don't see Tycho pretending score based systems are unfair. Now, it is up to one's opinion to disregard mathematical evidence. If you manage to build a successful pool while remaining exposed to an exploit, the more power to you. It is a different matter if such opinion is the basis of one's criticism of the mathematical evidence. In a public discussion, your opinion doesn't weight the same as a mathematical theory. Ill weigh in here. Most of the above is true however towards the end they did actually accept slush's challenge right before it was locked out. Also much of the confusion in the thread came from members AND non-members adding nothing constructive to the conversation mostly complaining about long rounds even though the pool was doing fine for its hash rate. There was an awful lot of critisim going on, much more than any other pool and im sure that was frustrating for two pool operators who weren't making a mint off running it like Slush
For information I tried the pool with the moded miner. I related that part of the discussion because it led to the locking of the thread. To the question "what are the downsides of BitcoinPool.com", i answered "the pool hopping", which is also relevant to the discussion prior to the lock down. And however frustrating it might have been FU and G, you might consider the fact that they're the ones who accused slush of cheating his pool. You should be prepared for a flame war when you go around flaming others. I mean, this is the internet. I think this is a perfectly fair and open challenge so long as he doesn't try and hide it from everyone.
Maybe yes, maybe no. I don't speak for slush but i can come up with a few arguments as to why he has little interest in performing that test under all these conditions. I'll elaborate if you so desire. Note also that slush offered to perform the attack under his name before Geebus gave in.
|
|
|
|