 Show Posts
|
|
Pages: « 1 2 [3] 4 5 »
|
|
Never dealt with Innosilicon, but I've heard from multiple sources to avoid their power supplies like the plague. Many DOA or shortly after, and they're allegedly louder than the miner itself.
|
|
|
|
Just got my batch 2 Mini yesterday. Ran it at 650 Mhz all night and had no HW errors and minimal pool rejects (using nanopool). Getting around 14.8Ksol/s. What's odd is I changed the cgminer.conf to set the speed to something higher but anything higher than 650 just gives me a really low hash rate, around 4.5Ksol/s, even 655 Mhz. No HW errors, just slow hashing. Anyone else experience this?
I noticed the Z9 mini seems to hash slowly unless you select speeds in specific increments. (usually 6 mhz) the only reason I know this is because after I experienced the same thing as you on my manually adjusted batch 2 mini, I looked at the OC options on my batch 1 mini and it looks like they usually let you select in increments of 6 or 7 mhz. I don't really know why but I'm guessing it's a memory timing thing. Try adjusting with the following exact increments until you get stability. 700, 693, 687, 681, 675, 668, 662, 650, etc. Is that starting at 700 and going downwards or starting at 600 and going up? or does not make a difference? It doesn't matter technically. but I would Just start from the highest and as soon as 1 hashboard crashes restart and drop it to the next highest and repeat so you can quickly find the maximum stable. It may take 12-24 hours until it faults out so don't assume it's stable after just a few minutes. |
|
|
|
Just got my batch 2 Mini yesterday. Ran it at 650 Mhz all night and had no HW errors and minimal pool rejects (using nanopool). Getting around 14.8Ksol/s. What's odd is I changed the cgminer.conf to set the speed to something higher but anything higher than 650 just gives me a really low hash rate, around 4.5Ksol/s, even 655 Mhz. No HW errors, just slow hashing. Anyone else experience this?
I noticed the Z9 mini seems to hash slowly unless you select speeds in specific increments. (usually 6 mhz) the only reason I know this is because after I experienced the same thing as you on my manually adjusted batch 2 mini, I looked at the OC options on my batch 1 mini and it looks like they usually let you select in increments of 6 or 7 mhz. I don't really know why but I'm guessing it's a memory timing thing. Try adjusting with the following exact increments until you get stability. 700, 693, 687, 681, 675, 668, 662, 650, etc. |
|
|
|
It's shady because of perceived value pricing.
"Shady practises" would include some of the following: - Selling the first batch at $1999 and then dropping them down to $1500 and $850 within a month's time.
- Issuing coupons to people who paid full price for it - that's like saying - come on suckers - buy more from us if you want to make up for the drop in price.
- Advertising as "limited to one unit per customer" and then changing that to 50 units per customer in a weeks time, and then making it unlimited after that.
- Rebranding and releasing old models as new with some miniscule adjustments to dump old stock. S9, S9i, S9j - all basically the same machines. L3+, L3++ - same story.
And the list could go on and on. Exactly, it's not like this is Bitmain's first sin. This just happens to be the one that finally pushed me over the edge. Also, how are batch 2s getting 14.5k sols/s? I haven't gotten mine yet but from what I've read if I understand correctly they locked them down so no more OC. |
|
|
|
Anyone opened up the miner to look at the boards? I'm curious if there are markings on the weak hashing boards to indicate some sort of difference and that bitmain knew something about them. Seems very coincidental that so many of these batch 2 miners have 1 weak board.
So because you cant OVERCLOCK one board as much as the others you are jumping right to a conspiracy theory that bitmain is doing things on purpose? The boards are not weak, they hash just fine at the factory speeds. There is NO guarantee of overclockability, you should be thanking bitmain for putting out a unit that can overclock by 40-60% in the first place. They only promised 10k sols out of the device.... I agree that technically they were honest with them providing their product to the advertised spec. But you can't say it isn't shady or misleading to give first batch users some extra unexpected hashing power as a silent marketing strategy just to sell batch 2s to folks expecting the same. I'm not going to bitch & complain about it because yes, I'm getting no less hashing power than what I paid for, but also I personally do not support this marketing practice and as a result I will never be purchasing another Bitmain product or recommend them. Edit: Also, this is neither a conspiracy nor a coincidence. It's simply silicon binning which is a standard practice throughout the entire semiconducter industry because all chips from a batch are not created equally. The part where it gets shady is that it's likely the highest quality chips were reserved for the batch 1s so everyone could hype over a 50% OC and then pre-order batch 2s. Why is this shady? They aren't going to redesign the chips from batch 1 to batch two just so YOU can't overclock. Get over yourself, the world doesn't revolve around you. It's shady because of perceived value pricing. Bitmain sold batch 2s at the price they did knowing the value was perceived higher than reality. I don't even know what you're talking about with chip redesigns, they're the same chips obviously. I really don't know what I said that makes you think i'm selfish either, I'm just saying how I feel about this behavior. Sorry that some random guy on the internet choosing to no longer support a specific company offends you. |
|
|
|
All A9 and Z9,
US Customs Duty / Tariff has increased on all miners coming into the USA.
26.5% of the value listed on the import documents.
This really sux.
I had to pay a total of $1223.25 to get five Z9 mini's out of US Customs
Follow the news much? Trump tariffs - 25% on goods made in China. (aka Make America Great again.) Round 1 was a few months ago and didn't affect us much. Round 2 started Aug 23 added some high tech items - not all but a lot of it is unclear. Round 3 expected to start in October and covers everything else made in China. If you think this is bad, wait until you see what it does to WalMart after Round 3. BTW Not all US Antminer importers are getting hit with the tariff. Perhaps it depends on the mood of the customs agent.  Looks like I lost that lottery. Got hit today $641 in import duties for 4 Z9 minis (used coupons) |
|
|
|
|
Do you have an example website that's using this? I've been following JSE for awhile now but haven't actually seen what it's actually like.
|
|
|
|
Anyone opened up the miner to look at the boards? I'm curious if there are markings on the weak hashing boards to indicate some sort of difference and that bitmain knew something about them. Seems very coincidental that so many of these batch 2 miners have 1 weak board.
So because you cant OVERCLOCK one board as much as the others you are jumping right to a conspiracy theory that bitmain is doing things on purpose? The boards are not weak, they hash just fine at the factory speeds. There is NO guarantee of overclockability, you should be thanking bitmain for putting out a unit that can overclock by 40-60% in the first place. They only promised 10k sols out of the device.... I agree that technically they were honest with them providing their product to the advertised spec. But you can't say it isn't shady or misleading to give first batch users some extra unexpected hashing power as a silent marketing strategy just to sell batch 2s to folks expecting the same. I'm not going to bitch & complain about it because yes, I'm getting no less hashing power than what I paid for, but also I personally do not support this marketing practice and as a result I will never be purchasing another Bitmain product or recommend them. Edit: Also, this is neither a conspiracy nor a coincidence. It's simply silicon binning which is a standard practice throughout the entire semiconducter industry because all chips from a batch are not created equally. The part where it gets shady is that it's likely the highest quality chips were reserved for the batch 1s so everyone could hype over a 50% OC and then pre-order batch 2s. |
|
|
|
I got my 2nd batch of z9 mini today. None of the 4 miner from this batch can hold 700 mhz clock, they can do 650mhz at 14.1k sol/s. Still fine tuning them for max hash rate.
2 miner from the first batch easily does 750mhz, 16k sol/s.
They are probably sorting chips for mini vs full size z9. That is common for other chip makers and helps explain variation in batch one hashrate. Unfortunate for batch two buyers on all fronts, it seems. The Mini and full have same chips, it appears, and same rated MHz/performance. So not much need to sort the chips for either. Not all chips in a batch are created equally. They pick the best ones for higher end products like the non mini which require a higher stock clock speed. This is called silicon binning. Only that the manual of the non mini says it’s also clocked at 500. Interesting. Does this mean the non mini has 4-5x the amount of hashboards then? Based on how well the batch 1 minis OC I would think the non mini should be able to hit 50k with just 2-3x the hashboards. |
|
|
|
I put coinhive on one of my websites to test it out, but some browsers are starting to flag it, causing trust issues with users.
It didn't make very much, like $20 over the course of a month, and this is on a website with thousands of daily visitors.
I think the only way it could make decent money is if you have a ton of traffic, and I am talking hundreds of thousands a visitors per day.
Else its not worth it.
You might want to think about giving JSE coin a try, it's opt in javascript browser mining which means that when a user is presented with your website a banner at the bottom appears prompting the user if they would like to participate in light mining (10% CPU). The JSEcoin team has already worked with many antivirus providers to whitelist the mining script. Basically it's similar to coinhive except not shady. |
|
|
|
I got my 2nd batch of z9 mini today. None of the 4 miner from this batch can hold 700 mhz clock, they can do 650mhz at 14.1k sol/s. Still fine tuning them for max hash rate.
2 miner from the first batch easily does 750mhz, 16k sol/s.
They are probably sorting chips for mini vs full size z9. That is common for other chip makers and helps explain variation in batch one hashrate. Unfortunate for batch two buyers on all fronts, it seems. The Mini and full have same chips, it appears, and same rated MHz/performance. So not much need to sort the chips for either. Not all chips in a batch are created equally. They pick the best ones for higher end products like the non mini which require a higher stock clock speed. This is called silicon binning. |
|
|
|
Who wants to buy 3000 jsecoin for 10000 doge send me pm
But I don't think you can even transfer the coins yet. |
|
|
|
The most I can get out of my two Batch 2 machines is 687 (although 681 produces a higher hashrate) and 625. Anything over that and the third hashboard on both machines goes to 0 hashrate.
i just recived my 2nd batch. mine too. 3rd chain die at 681 or up.  What hash rate are you getting at 681 or close. Do you get hw errors on any chains or does it just hash at 0 when you go at or over 681 same here... what about flashing the may 2018 firmware on to it?.... my batch 1 runs fine at 750..17 sols batch 2 one board dies above 675 freq.. 14 sols... I'm wondering if bitmain hasn't cherry picked hashboards for the Z9s over the minis in this batch? In theory I'll have a Z9 next week to see how (if?) it overclocks. I wouldn't imagine their yield suddenly going down across the board on OC stability. This makes sense. I'm at least 95% sure that the Z9 is just 2 minis combined with a much higher stock clock frequency. Classic silicon binning. |
|
|
|
My 2 latest z9s will not do 750. faults
I have the batch one at 756 and it's getting 16.4 ksol. How did you get up to 756? my batch 1 only lists a maximum of 750. |
|
|
|
|
JSE coin is the only one that looks promising to me. can either mine it directly in your browser or add the java extension to your website and visitors can opt in to mine for you. Mining uses about 10% CPU so browsing performance is not impacted much if any. Plus there's no advantage for having better hardware/higher hashrate as the system only accepts a maximum of 1 share per miner per 30 second block.
|
|
|
|
The APW3++ is supposed to be rated for 1200w on 110v. You can't run 3 minis on 1 APW without 6 pin splitters which I believe you're not supposed to do. However I am going to try to run 5 of them on 2 of these PSUs
The mini draws like 350W overclocked. You can likely run three of them on one APW3++ (though it's rated for only 1000W @110, I've never had a problem going over with any of mine--not suggesting others do it, just saying).
The APW3++ is rated for 1200W on 110v. You can't run 3 minis on a single APW unless you use 6 pin splitters which I believe you're not supposed to do. However I am going to try running 5 of them on 2 of these PSUs |
|
|
|
|
Just look at Siacoin. Difficulty is not dropping despite losing $2-3/day net. I think plenty of people (myself included) would be more than willing to mine at a loss gambling that the price will go back up... Remember, it's not a loss until you sell.
|
|
|
|
|
Why are there no other Blake(2b) coins out there anyway? SIA probably wouldn't be so overmined if there were other options available.
|
|
|
|
|
Are there any examples of websites that currently have this visitor mining system implemented?
|
|
|
|
I feel like zencash returns dropped because they changed the block reward for miners recently
Yes, last week the block reward dropped to 8.75, so now I believe miners are only getting like 70% of the coins and the masternodes are getting the rest. |
|
|
|
|
