SparkPool (OP)
Newbie
 Offline
Activity: 18
Merit: 1
|
 |
April 11, 2018, 09:07:55 AM |
|
Introduction
Many people asked us why the hashrate displayed on rig (Reported Hashrate) is higher than that shown by the mining pool (Effective Hashrate). Though many are told that less than 5% difference is a normal range, miners still found it confusing. The information asymmetry problem confuses everyone except those vested interest groups. In some circumstances, mining pools are accused of stealing computing power from miners. However, things are not that simple. To fully understand how this system works, you are suggested to take a 10 min reading this post. Let’s open the black box and find the truth therein today.
Before we step into the main part, here is some background information:
1. The work of the mining pools is nothing more than grouping computing power (hashrate) in order to get a stable block generation. They play an important role in distribution, stability protection (anti-DDoS) and efficiency (improve block utilization, lower uncle rate, etc).
2. The stability protection and efficiency performance depend on the technical development of each pool. Many pools are spending tens of thousands of dollars every month in anti-DDoS and technical optimization.
3. As for distribution, there’s much to say. In fact, there is no essential relationship between your payout with the calculated hashrate and the rejection rate. In other words, the payout distribution of any mining pool is a black box. As known to all, the payout distribution is based on the effective “shares” submitted, but your share proportion can be adjusted by the mining pool easily. It all depends on a pool’s conscience and ethical performance. So don't be fooled by a single rate shown by the pool or mining software, but believe the crypto you are paid every round.
4.The way mining pools change fee is not the same as the mining softwares (a.k.a. clients). Each time you mine, the mining softwares occupy part of your time to mine for themselves (for example, on the Claymore, 1-1.5% of the time, you are mining for the author’s ETH address). In the mining pool, your fee is automatically charged after every new block generation.
5. Most newbies spend all day only looking at the hashrate gap and staring at the currency value fluctuation. They have been trapped in many sections by the hardware machine sales, together with hosted mining farms, the electricity, mining softwares, cloud mining and greedy pools. All is not within the scope that we mainly talk about today.
6. SparkPool does not sell mining rigs, run hosted farms or sell software to miners. However, we want to tell you the truth and explain how the system works.
Enough for warming up, here we go:
You start your rigs, however, the reported hashrate and the effective hashrate show a gap. Many veterans say that a difference of 5% is normal. But how is the gap composed?
You must distinguish between the reported hashrate and the effective hashrate or known as the calculated hashrate.
Reported Hashrate
Reported Hashrate is used by mining clients (e.g. Claymore, Phoenix, Ethminer, and other third party mining softwares). The mining client submits the computing power (a.k.a. hashrate) of your actual hardware (what you see on your console) to the pool. For example, the RX570 could be overclocked to 29MH/s. The reported hashrate is a convenience feature used to allow you to compare it to your calculated hashrate shown by the mining pools.
Most importantly, Reported Hashrate does not directly related to your payout from the mining pools. It is a value only displayed on the rigs and known by the users.
Effective Hashrate or Calculated Hashrate
Effective Hashrate takes the number of valid shares your miner has submitted over a period of time and uses a formula to convert the number of shares into a readable hashrate. This number can fluctuate, sometimes being lower or higher than your actual miner's output hashrate on the console.
If we go deep into the problem, to ease the pressure of data communication between mining client and pool’s server, mining pools set a difficulty threshold with the Ethereum Proof-of-Work algorithm. Every time when those amount of work is completed, the miner will get a share. And the effective hashrate is calculated based on the shares.
Effective Hashrate(24h-avg) = your submitted shares in 24h * difficulty set by the pool
Mining pools calculate your hashrate based on the time interval you submit. For example, if we assumes that the time of submitting a share for 100M should be 10 seconds, then if your interval for each submission is exactly 10 seconds, the pool shows you a calculated hashrate of 100M. But if you submit it in 5 seconds, from the mining pool side, you have 200M.
The effective hashrate shown by each pool are generated differently. Some pools shown lower effective hashrate, while some mining pools shown effective hashrate even higher than your reported ones. This has absolutely no effect on revenue, as explained later. Because mining in general is a probability game.
So in most cases, if there is a large gap between reported hashrate and effective hashrate, you have to check on Claymore whether there is a GPU card dropped. If all run well, we can move to the problems presented below.
Inappropriate overclocking abuse.
How much a rig can be overclocked depends on miner’s experience. Different brands’ GPU’s performances are also different. You can find more information on Google. If overclocked too much, the client will give you an alert. Your submitted shares will be invalid and rejected. If you have found too many alarms, please downclock a little and re-check.
The impact of this overclocking problem will be 0%-10%. Some people will reach 50%, but it is completely avoidable.
Network Delay
Network delay is a problem that can only be optimized but cannot be avoided. It takes time to transmit data between local rigs and the node server. Lag within 50ms is perfect, and within 150ms is normal. If it goes higher, you have to find a closer server node. To solve this, each mining pool has its own server nodes in various locations of the world.
Why this will impact effective hashrate? Mostly because delayed submissions will be considered as invalid shares, or discounted when the valid calculation was missed while the others are mining the next block. Even though your calculation is correct, the submission was too late for the pool to utilize. Note that, mining is a continuous work, and it is inevitable for everyone to submit few invalid shares sometimes.
The network delay problem would roughly contribute the hashrate gap around 1%. If the difference is more than 2% means there is a space to be optimized.
The service fee of a mining software (client)
The creator of mining client does not extract the crypto you mined, but instead somehow extract 1% of your rigs' working time to mine for himself.
Since the creator has contributed to make the mining software, it is reasonable to pay for the technical intelligence. Most miners of the world use the Claymore. However, many copycats are build based on the core of the claymore. These copycats could miss report hashrate to generate profit for themselves.
The impact of the mining software like Claymore to the effective hashrate gap was 1% single-mode and 1.5% dual-mode. There are also Ethminer, Phoenix, and Genoil in the market. It is said that there is a slight difference between them. You should try it yourself.
As calculated above, those reasons add up to 2-3.5% difference of mining. That is network delay and software extraction. There are some hardware problems, such as the heat of your rigs, wire rod, network connection, power supply stability should not add more than 1%. Advanced mining farms rarely have these problems, so I think 3% is a theoretical long-term hashrate gap. So a gap of 5% is also taking into account the situation where you use copycat mining software that stealing your hashrate.
The mining pool
As just said, a hashrate gap of 3-5% is the cost for the mining pool after taking your hashrate. The rest of your real income is related to the submitted shares, luck, and payout patterns.
With PPLNS, it is normal for you to see your calculated hashrate fluctuates every day. So be better to see your return in the long-term test.
With PPS, your income is stable. Service fees are relatively higher to compensate the risks of pools.
With PPS+ (PPS Plus), PPS+ combines the advantages of PPS and PPLNS. Compared to the PPS model, which only awards miners block rewards and does not allocate tx fees, PPS+ assigns bonuses to miners and allocates tx fees. At the same time, the deficiencies of fluctuations in the PPLNS are avoided.
Let's talk about REWARD DISTRIBUTION in ETH mining. ATTENTION!!!
a. Main Block Reward
b. Uncle Bonus
d. Referring Uncle Reward
c. Transaction Fee Reward
Generally speaking, when pools distribute the payout, many of them only give you the part of Main Block Reward. They keep the rest themselves. However, many miners even don’t know there are other rewards. How much will it be worse?
For example, the PPS model does not allocate Tx fee. When the transaction volume increases and the network is busy, there will be many Tx fees.
So, the actual service fee of two mining pools could differ 7% or even more than 10% if all the rewards are taking into consideration.
To sum up:
1. Reasonably optimize your Hashrate.
2. Choose a transparent, unbiased mining pool to avoid cheating. Test and you will know.
3. The original intention of the blockchain is not speculation. Blockchain technology is not only about cryptocurrency but to create trust and consensus at the lowest cost. Please support those who have been continuously dedicated to technology and community in the blockchain ecosystem. Pay for knowledge and professionalism. Stay away from institutions or individuals who only resort to making money.
4. When you don't learn yourself and always listen to the recommendations of the mining industry middlemen, you are actually slaughtered.
|
