This article is from the original Iris of the IPFS Force Zone
How did the miners prepare the day before the second phase?
Today I would like to share with you, what do you think of Filscan's so much data and how to quickly understand the strength of a certain miner?
01 32GB miner
The second stage of the testnet only supports 32GB sector size. Judging from the most recent official network reset, only 6 miners have exceeded 1T computing power, namely t01002, t05268, t03774, t03848, t02286, and t06681.
Supporting only 32GB sectors will undoubtedly raise the threshold for participating in the test network, which is not conducive to the healthy development of the community. Both the official and large miners are responsible for advancing technological progress and lowering the threshold for small miners to participate.
The way to view the miner sector size is to directly click on the miner’s name to jump to the miner’s details page to view the miner’s sector size.
02 Original computing power and effective computing power
The original computing power is the computing power stored by the current miner, that is, the total computing power we have been able to see before; the effective computing power refers to the computing power stored after the miner accesses the order, including both effective storage and self Filled storage. According to the official instructions, the hashrate stored in the order will add a weight to the original value (the hashrate increases). Therefore, the effective computing power>original computing power will be caused.
This is the official way to encourage miners to improve real and effective storage-effective computing power basically determines the block generation. It is foreseeable that a strong miner, the more orders he receives, the greater the effective computing power, and the block generation rate will increase accordingly. At present, the original value and effective computing power of miners on the computing power rankings are the same, which means that the miners are just filling in their own data and have not started to perform effective storage.
There are two factors that affect effective computing power, one is data with deal, and the other is to verify the real data of customers. Judging from the current situation of miners, all miners are storing garbage data, and they have not passed their own billing methods.
The point is that, according to the current official economic model code design, self-issued orders will increase the effective computing power by 0.1 leverage, and storing real data will increase the effective computing power by 10 times the leverage.
03 Computing power growth curve
Now let's take the miner t01002 as an example to check the effective computing power growth of a miner. Other miners are similar.
From a trend point of view, the computing power of the t01002 node began to appear last night, and until 4 o'clock, the computing power was slowly increasing; after 4 o'clock, the growth curve suddenly became steep. This shows that from this node, t01002 miners have increased their investment in storage hardware, and the amount has increased a lot.
In terms of details, the growth of all miners (including t01002) is such a staggered growth of "growth-platform-growth-platform". This is actually a good explanation. The platform is actually being sealed by the miners. After the packaging, the computing power can be displayed, and there will be an increase. Then, if you want to see the growth trend in detail, you can also quantify it by the hourly growth rate-growth rate = (later 1 hour computing power-first 1 hour computing power) / first 1 hour computing power.
With specific data, we can also clearly compare the specific strengths of various miners in storage and sealing. The way to view the miner growth cancellation is to click on the miner number (red box) of a miner in the above figure, and you can get a separate miner growth curve.
04 Miner News
Combined with the growth rate of miners' effective computing power, we can also see a series of actions that miners are doing on filscan. Then it is this "message" that can reflect the miner's hardware configuration (such as whether to use GPU). The screenshot above is the first time the miner of t01002 went from "data pre-submission" to "data submission proof" (phase 2), which was calculated to be about 24 minutes.
According to official test results↓↓↓
A 2080ti GPU takes 57 minutes to complete in the second stage. Comparing the two, from 57 minutes to 24 minutes, this means that the t01002 miners used at least 2 to 3 2080ti GPUs in this process.
We know that the seal process of miners is divided into four stages, and the messages of each stage are submitted to the chain. By analyzing the messages of these four stages, we can grasp more information about this miner. For example, the peak investment of miner t01002 is around 300 servers.
05 Block rate and win rate
Regarding the block generation rate and the winning block rate, this is the situation that directly determines the current profit of the miner. The influencing factors of the block generation rate have just been mentioned and are linked to the effective computing power.
But we also see this situation: the t01002 node ranks first in the effective computing power rankings, and has lost nearly half of t03425's computing power; in terms of block generation and win rate, it is far inferior to t03425. What is the reason for this?
At present, we can analyze the reason for the problem of reading the stored data. The hardware of t01002 is currently very strong from the growth curve and the submission time of the second phase. Then the most likely problem is its There are certain problems in the performance of distributed storage, and the level of response is to read.
／End.