GTX1070 ZOTAC Graphics Card Issues: Warning about Excavator-Origin & Damaged Cards!

What are you talking about ...

2konrafal1993 wrote:

are exploited to the limit

I wonder how. Typically, excavator cards run at reduced clock speeds (and hence reduced power consumption) to optimize the power-to-performance ratio. Besides, the cards don't wear out! The fans may wear out, although those in excavators usually work in very good conditions because they are much more distant from each other than if they were installed in a standard PC. This is ordinary electronics. If they are not working at cosmically elevated voltages, or in an environment with high ionizing radiation, then nothing happens to them. Silicon does not wear out!

Robert B wrote:

They're going to give you marbles under the GPU in a moment and you'll be crying.

Why is that? The soldering under the elements oxidizes even when the card is left unused, although the process progresses faster at higher temperatures. In practice, you have a lottery or the card will fail after 3 years of use, will it survive 10 years and still work. Simply the element of randomness plays a much greater role here than the temperature, and certainly where the card worked.

@misterbociek The problem with the cards you bought isn't that they were excavated. You just came across a dishonest seller who gave false information in the descriptions. That's the whole philosophy.

misterbociek wrote:

hahahahahaha I can already see it when they buy an MSI gaminX card that can be easily turned up to 2000Mhz and they break it to 1300MHz hahahhaha

You have no idea about digging. Yes, the clock speed is reduced, because although the performance drops only a little, the power consumption can be cut by as much as half. So you dig a little less, but only pay half for the electricity. Get it?

The other way around - overclocking the card makes the cards "a little" more efficient, but the power consumption is much higher than the base one. In this case, so what if the card copies faster, as the excess that it digs will go to electricity and you have to add electricity to it.

Added after 1 [minutes]:

Why do you think that why special BIOSes set for optimal mining are loaded into the mining cards?

I don't know anything about filming memories, although it sounds logical. Just note that the memories do not fall from too high clock speed, but from degradation due to high voltage.
Here again we come back to the same conclusions - the fact that the post-excavation card died shortly after the purchase is not due to the fact that it comes from the excavator. I would even say that it has nothing to do with it. Nobody will forbid me to turn my cards into space and use them for games. The consumption of such cards is only due to improper handling.

If the memories on the card in the excavator (while digging) work at very high voltages, but do not degrade, then after selling such a card and using it at the correct voltages, the degradation does not continue. So if you buy a working card and make sure that it works with the factory parameters and not overclocked for the excavator, then nothing will happen to the memory.

Who even invented such stupidity as graphics cards that wear out. It can wear out, the hard drive, fan, SSD can also be said to wear out, although in practice they are simply less unstable, and then such @ 2konrafal1993 he duplicates them on the Internet without knowing what he is writing about.

Robert B wrote:

Kraniec_Internetów wrote:

Who at all he invented such stupidity as graphics cards that wear out.

Oh ... technician, you still don't know much.

I corrected. And if you have a different opinion about it, I'd love to hear it. Because your above post alone does not add anything to the topic

Robert B wrote:

I edited the post. Look up.

Really.

Robert B wrote:

In the now fashionable "Pb free" solders, degradation from elevated temperature progresses faster than in the old ones with lead admixture.

Yes it's true. However, what does this have to do with excavators? No. It is precisely because of the type of solder that the new cards will last less than the old ten years, but if you were to check the effect of digging on the "card wear" on the same specimens, putting a few cards into the excavator and kicking, and a few using in desktop computers for more commercial applications, the service life would be Both groups will be very similar and the first violin will be played by luck in hitting a better or worse specimen.

rzymo wrote:

Only the memory voltage cannot be moved without physically modifying the card. So after all, this high clock speed and / or temperatures must have an effect and they are doing the bone a disservice ...

Not completely. While conventional overclocking programs do not allow you to change memory voltages, you can do so via the BIOS. It is true that I did not rummage in the BIOS of modern cards, but only in the ten-year-old EAH5550, but I assume that if that card had such an option, modern ones also have it. And I have already written about BIOS modifications several times here. I also know that the BIOS is modified in 90% of cases in the cards intended for mining.
Consider whether excessive clocking would only harm graphics memories and not other chips? Because it does not hurt other systems, at most they simply do not work properly, they fall apart or do not get up at all, but nothing is damaged.

Robert B wrote:

You're contradicting yourself.

Why?

2konrafal1993 wrote:

Nowhere did I say that I should not buy cards only from excavators, but in general used cards, a used card, the more of an unknown origin, is a waste of time and money, not to mention nerves.

Okay, then I can partially agree with you, although not entirely. As I said, the issue of solder under the systems is a lottery - you can buy a 5-year card that will work for the next 5 years without any problems, and you can buy a new card that will win after 3 years.

Robert B wrote:

Of course, but the future engineer slept in chemistry and physics classes

He was awake, he was awake, and I would like to ask you not to catch my words. You base your opinion on several laws of physics and formulas, and I base your opinion on practice and statistics, because if I wanted to describe the phenomenon in some way with laws and formulas, you would have a whole bucket of variables and China would not be one with the other. I admitted you were right that lead-free solder is worse and is certainly more likely to cause malfunctions. However, returning to the original topic, we do not associate excavators with excessive wear of cards. True, the card in an excavator usually works non-stop for several months. The fan still rotates with a considerable rotational frequency, and it will probably wear out faster. But I wrote about the wear of the fans earlier. I currently have a GTX 1060 in my computer, and I haven't turned it off since the end of September. Yes, September. Not counting the quick restarts, it has been working for almost half a year. On one of the monitors, I usually have a game running, which causes the fan on the card to turn on for a few seconds every few seconds. And I bought the card itself, perhaps used from an excavator, because the seller was selling 6 of the same. The paste on the core turned to stone, now I have liquid metal in there and peace of mind with maintenance for the rest of the card's life.

@ 2konrafal1993 You don't want to buy used cards, ok. I prefer to buy a higher but used model for the same price. I will definitely never buy a used hard drive or fans. Because these are the elements that actually wear out. They wear out, i.e. they will die faster, the more they are used. Graphics cards won't crash from constant work, and neither does processors or motherboards.

2konrafal1993 wrote:

take into account that there is an elevated temperature on the GPU

This is bullshit. As I said, the cards in excavators usually work at a significantly reduced power, secondly, digging does not use all elements of the graphics core, so in total the power dissipated on the core of the kicking card oscillates around half the power that would be released at full load, let's say Furmark, at factory timings and tensions.

Robert B wrote:

You think so? Then throw that bucket

I am not able, and neither are you. Nobody in this forum can.

Robert B wrote:

interesting, because I have over 40 years of professional practice. How many, kid?

And I dug almost a year in total.

Robert B wrote:

You sound like an oracle normally

I just know it's very complicated. Much more than you think, and only research and tests are capable of determining it, not a set of laws and calculations.

Robert B wrote:

They're going to give you marbles under the GPU in a moment and you'll be crying.

And I am the oracle here? How do you know in a moment and not in 3 years? Or maybe in 5?

What did I tell you about catching my words?
How do you know it's soon? Why not in 3 or 5 years? And how much is that soon for you? Month? Year?

rzymo wrote:

No, you can't and I'm writing about it. You can modify the clock and timings. In modern cards you will not touch the tension of memory.

I take your word for it. Could you drop off an article on memory chip damage in excavators? As long as they were not GDDR6 memories, because they broke by themselves, regardless of how they worked.

Unfortunately, Robert B is no discussion partner, because apart from personal trips and picking on words and spelling, he did not write anything constructive on this topic, so another mockery of me will end with a report. And we will see if my colleague Robert will be so happy.