|
IMO, what has kept ETH mining profitable for so long is it's unique aspect of not just being a Cryptocurrency, but a development ecosystem which has evolved in to an entire market sector. I very much doubt we will see another POW coin that can support the amount of GPU's and replace ETH's profitability. But I also believe that the millions of GPU's that are currently mining ETH are not just going to sit idly without some project looking to tap in to that enormous computational resource. If ETH's transition to POS is successful, instead of GPU's doing meaningless POW computations to secure a network it may evolve in to a distributed, decentralized network of on-demand computing power such as with the Golem Project.
|
|
|
|
Perhaps I should go back to my roots then and only buy amd from now on  Edit: oohhh, it does not concern the 3090 apparently, at least according to that statement. I Agree. AMD has taken the opposite approach as Nvidia in their support of the mining community. While Nvidia is looking to demonize and alienate casual miners for Nvidia's shortcomings, AMD supports them. My Polaris RX 580's hashing at 31-32 MH/s @ 115W each make them just as relevant for mining today as they were in 2017 when they were released. When it became apparent there was an issue with the Polaris line up losing ~10% hashrate every month from the growing DAG file size, AMD quickly developed the blockchain drivers to address the issue before it escalated and continues to implement it in their current drivers with compute mode. Now that it's apparent that Nvidia's Pascal lineup from back then also has the same problem, nothing but crickets from Nvidia. My few Nvidia Pascal cards are no where near as efficient today as the Polaris cards despite pulling almost twice the power and running much hotter., so I have switched to all AMD cards. Many large mining farms that were started during the 2017 bull run still in operation are still using Polaris cards. |
|
|
|
There are NO differentiation on the packaging at this point in time. They did say they are goina be sneaky about it and so far they are using the old boxes. There is a slightly different serial number on the card. So if you buy new cards you will not know until u open the box up. For cards being sold at retail that have a LHR and non-LHR varient the LHR 3060 Ti, 3070, 3080 etc. will definitely have a LHR indication on the product listings and packaging. The 3070 Ti and 3080 Ti will only be produced in LHR versions. https://youtu.be/T4R5GN60CBo?t=309 |
|
|
|
|
The worker name is just for you to keep track of your rigs, so you know which one is being reported on the pool. Most ETH pools don't require an account. You can do all the configuration in the bat file and start mining. Some pools will require you to set up an account with a username/password and worker names which you then use in the miner configuration. Every pool has a 'getting started' page with the server addresses, port numbers and setup needed to start mining on the pool.
|
|
|
|
I'm with you. Plotting 7TB a day with the expectation of solo mining a block is not sustainable, even if hard drives were available. In reality for most farmers, solo mining was a realistic option for the first month after launch. Once HPOOL came online the net space growth rate was unmanageable in trying to just keep up. We really need the official pool ASAP. https://www.chiaexplorer.com/charts/netspace |
|
|
|
It depends on the coin. If you are talking about one of the top 25 market cap minable coins, then there is some miner capitulation during a prolonged bear market, but it takes a while. Much of the mining is now commercial large scale farms concentrated where they can operate at a very low overhead. During the last bear market from 01/2018 to 06/2020 the ETH and BTC difficultly actually went up. Smaller cap coins are much more sensitive to market fluctuations. When the price goes down miners move to other coins that are more profitable. https://bitinfocharts.com/comparison/bitcoin-difficulty.htmlhttps://bitinfocharts.com/comparison/ethereum-difficulty.html |
|
|
|
|
I have a 24-core 3960X Threadripper build with 75TB across 8 HDD, 3 plotting SSD's, 128GB RAM and 9 case fans farming Chia. The whole system uses 450-470W at the wall and that is with 2 RX 580's mining ETH that use at least half of that.
|
|
|
|
one other thing I haven't been able to figure out... do we know what a "block" is? Is one block equal to one plot? Or do a series of your plots constitute one block?
I'm imagining that you have 629 blocks, right now doing nothing but sitting there waiting to be picked by the lotto, right? So you win a block and now one (or more) of your plots are being utilized by the blockchain, is that an accurate understanding?
Plots are essentially lottery tickets that are used to validate transactions contained in the blocks. The more lottery tickets you have, the better chance at winning. Every 10 seconds a new challenge is sent out looking for a specific combination contained in the plots. Each plot only has a 1 in 512 chance of passing a filter which is used to minimize the number of disk reads required. If one or more of the plots passed filter it moves on for a chance at generating a proof and winning the block. Other than generating proofs for blocks, plots are not utilized by the network. https://thechiafarmer.com/2021/04/26/how-chia-farming-works/ |
|
|
|
To put this in perspective, with 62TB I currently have 0.0012% of the total network plotted space, which when calculated with how many opportunities there are to win a block in a day works out to an average expected time of 18 days to win a block at the current network state as is shown in the Chia farmer.  Deviations from the average are normal. Just because someone won a block much faster than expected doesn't mean that is an outcome you should expect. In fact the opposite is true, statistically speaking the more people that win faster than average, the less probability that will happen to you. Especially when the network keeps growing at an exponential rate as it currently is. You also hear about the people that were lucky disproportionately from the people that were not as lucky. This video has a good explanation of how probability works in relation to Chia farming and has a spreadsheet you can use to estimate your expected outcomes based for network growth. https://www.youtube.com/watch?v=_yEBS-q0sK4 |
|
|
|
On cards with GDDR5X memory you need to increase the memory timings. Current ETH miners have a configuration parameter to do this automatically. With Pascal GPU's in PhoenixMiner you use the -straps 2 command line option or in the miner config file. -straps <n>
Memory strap level (Nvidia cards 10x0 and P10x series only). The
possible values are 0 to 6. 0 is the default value and uses the
default timings from the VBIOS. Each strap level corresponds to a
predefined combination of memory timings ("-vmt1", "-vmt2",
"-vmt3", "-vmr"). Strap level 3 is the fastest predefined level and
may not work on most cards, 1 is the slowest (but still faster than
the default timings). Strap levels 4 to 6 are the same as 1 to 3
but with less aggressive refresh rates (i.e. lower "-vmr" values). |
|
|
|
|
Yes, the time to win given is an estimate on how often you should win a block based on your percentage of the total plotted network space. Luck is also a factor which is why some people with only a couple TB of plots have managed to win. If you hover over the estimated time it tells you your calculated percentage of the total network plotted space and that it can take 3 to 4 times longer than the estimate given. The problem is the network is growing so fast that unless you are plotting 3-4 TB of new plots a day, you are falling behind the growth rate of the network, so you're estimated time keeps growing. The shortage of hardware also makes it hard to be able to keep up. The network is now at a point where they really need pooling to have more predictable earnings and wider distribution.
|
|
|
|
Iv'e been plotting for about a month. Started about a week after launch and currently at 60TB with no XCH farmed, but it's all good. Started slowly at first and ramped it up about two weeks ago with my new Desktop 3960X Threadripper build. I'm already at capacity after scavenging all the unused storage I had and even my NAS. New high capacity HDD are OOS or too expensive, so I'm ready to go and looking forward to pooling which should be released soon. The fact is Chia has been in development since 2018 and the development team has always been very transparent with the project. If you go on YouTube you will see numerous interviews going back years that were given to various tech outlets discussing the project status. The testnet lasted several months prior to launch and anyone that had been following could participate. It's also true that the mainet launch was announced well in advance, so that anyone that was following the project and wanted to participate could have done so. Many early adopters did and now they have been heavily rewarded for it. Regarding the Chia foundation fund allocation, the fact is development teams to launch a project on the scale of Chia cost money and the foundation is governed by strict rules as to how it is spent, which no different than many other successful projects in the Crypto space. https://www.coindesk.com/5-takeaways-from-chia-networks-new-white-paper |
|
|
|
|
No problems running 5.6c with Polaris RX580's in Linux 5.12 and AMD 21.10 drivers. It runs flawless and I also have full overclocking and power settings through the miner config or terminal with rocM. However with a Nvidia Pascal rig it doesn't show the temperature with Phoenix 5.6 and it has some issues with slowing down the primary GPU when mining and using the system. The problem is Nvidia Linux drivers are garbage.
|
|
|
|
|
Not all motherboards support 6 or more cards, even if you have the PCI-E slots on the board. A few settings you can try in the BIOS:
- Make sure you are using the current BIOS for your motherboard
- Change the PCI-E link speeds to GEN 2 or GEN 1
- Enable '4G Decoding' if you have that option
- Disable the integrated GPU and set the Primary display to PCI-E (Usually PCI-E slot 1)
- Disable the integrated audio controller
|
|
|
|
|
Just getting started now farming Chia with the expectation to solo mine is not realistic at this point. The net space growth rate has exceeded even the most optimistic expectations of the development team. For this reason, they have made officially implementing pooling their #1 priority and we should hear some updates soon. They have suggested that in order start farming on a pool, you would need to re-plot, so that the plots can be portable and can be used at different pools if you choose. This would mean in the beginning everyone looking to join a pool would effectively be starting from scratch, so there is an advantage in having the hardware ready to go if you plan on doing pool farming.
I don't think comparing Chia as a Burst coin 2.0 is a fair comparison. For one, the development and marketing teams at Chia are first rate and have wide recognition in the tech industry. Then there is also the programming language developed for Chia which would enable things like smart contracts, comparable to ETH. If the 'green crypto mining' that is being promoted catches on in a new market sector, I could definitely see XCH being a leader in that sector.
|
|
|
|
|
The max speed of your GPU is irrelevant. Given what you are trying to do has been obsolete since 2015 when ASIC's took over Scrypt mining, there are no Scrypt optimized miners for any cards since then I'm surprised you managed to get it working at all. If you really want to GPU mine DOGE, simply mine what is most profitable for your cards to an exchange address and trade it for DOGE. That way you are utilizing your GPU's to their fullest potential and not wasting time trying to mine Scrypt with GPU's.
|
|
|
|
|
Absolutely your ISP can see you are mining and so can anyone else that can do packet analysis on your network. If you use SSL your traffic is encrypted but they can still see you are connecting to a mining pool, unless you route through a VPN, then they will know you are connecting to a VPN.
|
|
|
|
When you have <10 GPU's in an average room A/C is a reasonable option, albeit expensive to run. In a hot climate for an average 250 sq/ft room you need around 6000 BTU's in air conditioning just to cool a standard indoor room with no additional heat sources. With 10 or more GPU's in the room it's going to heat up quickly unless the A/C runs pretty much 24/7, which would be very expensive and your A/C would breakdown quickly. Best thing I did was add an 8" inline fan and vent the air through a crawlspace in the attic. It lowered the room and GPU temperatures dramatically. https://bitcointalk.org/index.php?topic=4465680.msg40192306#msg40192306 |
|
|
|
It's in the 'sensors' tab, the second one. 'GPU Load'  |
|
|
|
|
Show Posts
