Archive for the ‘Tests and Reviews’ Category

More Information About the Jasminer X4 1U Ethash-Etchash ASIC Miner

28, Oct 2022

Yesterday we have published our First Hands-On Experience with the Jasminer X4 1U Ethash-Etchash ASIC Miner and today we are continuing with some useful information about the device for anyone that might own it already or plans to get a Jasminer X4 to play with. As we’ve already mentioned this ASIC miner for Ethash and ETChash is quite easy to work with both in terms of hardware and in terms of software as it pretty much modular in terms of hardware and with open root SSH access to the software information on the device. So, today it will be more about the software and what you can do in order to tweak or play around with the miner to optimize it more, or maybe overcome some lacking feature etc.

The Jasminer X4 1U Ethash-Etchash ASIC miner comes with two options for configuring and operating the miner – a web-based interface and a software tool for Windows only. Both feature pretty basic settings only – mining pool settings (up to 3), what algorithm/coin to mine (ETH/ETC) as well as the operating frequency of the miner (200 or 225 MHz). The rest is handled automatically or set to a predefined value, thus kind of limiting what is user accessible and making you think of some workarounds if you need something specific. We actually had a few to work with and have found a solution, though it would have been nicer if there was a bit more user control available… an advanced settings page in the web interface maybe. Oh, wait, there actually is one, but it is hidden and not fully functioning (advanced.html) in the latest firmware.

In the advanced control page besides the 200 and 225 MHz operating frequency there is also an option to enable 250 MHz as well, then there is also an option for fan control available (Auto and Manual modes) and there is also an “Operation Mode” that can be either High-efficiency, Balance or High-performance (not sure what these actually do). But even if you work around these settings trying to set them does not work apparently as there are other measures set to kind of limit your access or the features might not be properly implemented at all… not sure which one it is. We’ll get back to these in a moment with some more details.

The sl_miner control software that is only available for Windows operating systems has pretty much the same settings for the miner, plus an easy option to scan an IP range to see what is the IP address of your miner generated by the DHCP server, so that you can access its web interface. This software makes it easy to configure and monitor your miner from a single window and what it does better than the Web-based control panel is making it easier to control multiple Jasminer X4 miners that you might have. With the sl_miner software you don’t have to open different web-based control interfaces for each of the miners if you have more than one that is.

Now, have in mind that once you boot up the Jasminer X4 miner it may seem to be idle and not working for a while or at all. You will need to have some patience as the device needs about 20 minutes of time in order to generate the DAG file before it can actually start mining, so be easy on the reboots or config changes that will require you to restart the miner. The good news is that changing pool settings and operating frequency of the miner does not need a reboot of the system, changing the algorithm you use however does.

If you are feeling adventurous you can login onto the Jasminer X4 control board over SSH with putty or another such client and explore the device on the software side. The default username is root and the password is also root and you are ready to start looking around. The miner software is called Jasminer and is version 2.05 in the latest firmware and to make things a bit easier for you we have shown the list of commands supported by the miner. Most of the things here are not much of an interest, apart from the option to set the fan speed and to change the operating frequency of the device. Notice that on the screenshot it says JasMiner-Brick 2.05 and there is a reason for that although we are using the Jasminer X4 1U here, but more on that in a moment.

Do note that the jasminer miner software is executed and running as soon as the device starts and there is a monitoring script that will restart it if it senses that the miner has stopped for some reason. So, just issuing a “killall -9 jasminer || true” command (without the quotes) will stop the miner for a moment and then it will start back again, you need to also stop the monitoring script in order to manually run jasminer with your own settings. For stopping the monitoring script as well you will need the following command “killall -9 monitor || true“, do note that it would be wise to start it afterwards or reboot the miner to automatically revert things besides the config file.

By default, the miner runs with a config file that is being created and changed from the web interface or the sl_miner software, the exact command line that is used to start it is as follows:

jasminer --version-file /etc/machine --config /media/configs/jasminer.json --syslog

Inside the config file you only have the pool information, the algorithm and operating frequency along with some API data. There is no option to manually control the fan speed from the config file, you need to run the miner with the “--fan_speed” set to the percentage value you want and that can be a problem if you want to do it permanently. The configuration file is saved on the memory of the device, but the rest of the software is loaded into memory from a bootable image, so no permanent changes will be saved and available when you reboot the device unfortunately.

So, manual fan control can be tough issue to deal with the Jasminer X4. The automatic mode active by default apparently follows the temperature and adjusts, but you might just want to have the fans running at 100 percent all the time for instance. Or you might want to modify the miner with more powerful fans and run them at lower RPM to make them silent while still providing more than adequate cooling for the ASIC chips. There are some possible workarounds here that might help you do things like that, but they will most likely not be permanent and will need to be reapplied after a reboot due to the way the device operates.

One thing that you can permanently do in order to have the fans running at maximum RPM is to leave only the power pins of the cooling fans connected to the miner’s control board. That however will present you with another issue – no reporting of FAN RPM to the software and this will trigger a safety measure that will prevent the device from mining. It will still run and calculate a DAG, but as soon as it needs to start hashing it will just not start as it will think that the fans are not operating and it will overheat if it starts mining. You will end up in the same situation if you are thinking about some alternative cooling methods as well such as immersion cooling where you might not need to connect any cooling fans at all. Getting fans reported at zero RPM will trigger a failsafe mechanism that will prevent the miner from starting to hash… there is an easy solution for that as well.

We have remembered that there is a smaller Jasminer X4 Brick miner that has a lower hashrate and is completely passively cooled, so why not flash the firmware of that miner (yes, it will work on the X4 1U as well) and get rid of the annoying fan safety measure. Remember that this miner does not have fans, so no an alarm firing when fans are reporting zero RPM… they do it anyway as no fans are connected there in the first place. The miner will however run any fans connected to the control board just fine, it will just not prevent you from mining as there is no fan failure detecting mechanism available in this version of the miner.

There is a small catch however, by default the web-based interface for firmware upgrades will refuse you to flash the firmware for the X4 Brick, the only thing you need to do is just rename the original file you download from the Jasminer’s website from JasMiner-X4-Brick-20220726-183045.tar.gz to JasMiner-X4-20220726-183045.tar.gz and then you will be able to flash the X4 Brick firmware onto your X4 1U just fine. Performance and feature wise we see no difference with the X4 Brick firmware compared to the original X4 1U one apart from the lack of fan failure mechanism in place and the fact that just a single operating temperature is being reported by the device and not a range.

So far, so good. But what about running the miner at a higher frequency in order to increase the mining performance? By default, you have only 200 MHz and 225 MHz available, but you can also make the miner run permanently at 250 MHz with a little playing around. You just need to SSH to the miner, go to the /media/configs folder and edit the jasminer.json configuration file changing the freq setting to 250. Note that this is a semi-permanent setting change, meaning that it will remain in place after rebooting the miner, but will be overwritten as soon as you change the mining pool for instance. You will simply need to redo the change again in order to have the miner boot at 250 MHz operating frequency.

It seems that this setting might be removed as an available option due to the fact that it can be a problem for a lot of even all of the Jasminer X4 to operate properly at that setting with their default setup. In our case the device does produce way too many errors and as a result only about 50 MH/s hashrate is being reported instead of getting more than 10 times higher performance. You are welcome to try and report if it works for you any better or not and if it does what hashrate are you getting. It might be a problem of proper cooling, or the 300W power supply not being enough or simply because the chips are not Ok with the higher frequency.

Anyway, below you can find a quick list of the three operating frequencies of the miner with the power usage measured at the wall and a hashrate that is expected at that setting. Do note that the power usage reported is with our silent modification to the miner applied and that lowers the actual power used as compared to the original cooling setup:

200 MHz – 237W – 450 MH/s+
225 MHz – 265W – 520 MH/s+
250 MHz – 294W – ~50 MH/s

The Jasminer X4 1U does manage to work just fine at 225 MHz producing over 520 MH/s of hashrate stable and problem free, so that is the recommended operating frequency of the device in order to maximize performance. If the 250 MHz setting works for you things might be better, but not all or any might do. If not working properly at the higher operating frequency you will be getting much lower hashrate than expected with higher power usage as the miner is actually operating, but producing a lot of errors and not properly hashing most of the time.

Few more things that we already talked about in the previous article, but that need to be mentioned here as well for people that don’t see both posts. The whole lineup of Jasminer X4 ASIC miners comes equipped with 5 Gigavbytes of memory, so besides the device not usable for Ethereum mining anymore (due to the recent switch from PoW to PoS for that one), you will not be able to mine ETHF or ETHW or any other fork that split off Ethereum at the Merge that occurred last month. These already have passed a DAG size that can fit inside the memory (you could mine them with a lower hashrate for a while), so think about mining younger Ethash coins with smaller DAG size than 5GB or ETChash ones such as Ethereum Classic.

Also, currently the Jasminer X4 product lineup does not support dual-mining of ETC/ETC coin with ZIL, something that could increase the profitability of the mining hardware easily with 30-40%. Unfortunately, not supported for the moment and the same goes for using the Jasminer X4 on the Nicehash mining platform where you could also aim for higher profit than just mining ETC. Nicehash is not currently supported and the miner is not connecting to any of the pools on the platform at the moment, though a future compatibility update adding support for these two things would be a nice addition.

We do hope that this extra information and tips can be of use and help you get more out of your Jasminer X4 1U ASI miner and if you do have some questions feel free to ask. We are going to continue playing with this miner for a while for sure and will report more about it later on, hopefully along with good news regarding improved or added features. Meanwhile next we are going to target our attention to a competitor of Jasminer X4 – the iPollo range of ASIC miners that have been on our radar for a while now too.

First Hands-On Experience with the Jasminer X4 1U Ethash-Etchash ASIC Miner

27, Oct 2022

The Jasminer series of ASIC miners for Ethash and ETChash by the Chinese company SUNLUNE have been quite big (and expensive) when they were announced near the end of last year. They were the most efficient Ethereum ASIC miners out there at the time and with the high mining profitability of ETH at the time there was interest and demand even being quite expensive miners. Now, with the end of Ethereum’s mining phase and the low profitability of mining for pretty much everything the Jasminer ASIC miners are easier to get a hold of and you may get a decent deal for the price of one. The big question however is should you invest in a Jasminer X4 ASIC miner now and that is precisely what we are going to talk about here.

We’ve managed to score a decent deal for about $2000 USD on a Jasminer X4 1U-C ASIC miner as we were trying to get our hands on one of these for a while now to play around with it and see what the hype is all about and share our first-hand experience from the device. So, below you will find some of the things we have noticed playing for a few days with the device as well as some useful tips and tricks that might help any of you that might already have an X4 miner or are maybe planning to buy one. One of the first thing we love to do is to open up the hardware and explore what we have inside and how we can improve it further or optimize it, then play around with the software as well to see what else may be lingering in there too. But we will get to that in a bit, for now let us just share our first hands on experience with the device…

Jasminer X4 1U is rated at 520 MH/s with a power usage of 240 Watts (with 10% variation according to the specs) with a power efficiency of 0.46 J per Megahash or hashrate. The device is equipped with 8x Jasminer X4 ASIC chips in a 45×45 mm size (678 square millimetres die size) with 1TB memory bandwidth, 5GB of memory and a hashrate per chip of 65 MH/s. Each of the chips has 384 computing cores and 384 on-chip dies according to the manufacturer with a 512-bit bus width and an operating frequency of 800 to 1 GHz and a power consumption of 23 Watts. The 8 ASIC chips are placed on a single board with a controller in the middle and there is also a separate controller board, a power supply and a number of fans inside the case.

The Jasminer X4 1U-C that we got is pretty much the same as the Jasminer X4 1U model, however it is rated to 450 MH/s with 240 Watts of power usage instead of 520 MH/s with the same power consumption. We are not entirely sure if there is actually a hardware difference between the two units or it is more a matter of software and more specifically the clock rate of the chips 200 MHz clock for the X4 1U-C and 225 MHz clock for the X4 1U and this is something you can control from the miner configuration (the 800 MHz to 1GHz specs of the chip). Shouldn’t this however mean that these chips should be able to run at 250 MHz as well then, well, it does, but you are not allowed to do it by default, though you can read how to do it yourself below.

The mining device comes in a standard 1U server case, so it is easy to deploy in a rack and be used in a data-center environment or server hosting place. If you want to keep it at home it may not be the best idea as by default the miner is pretty noisy (up to about 80 Decibels according to specs), so not suitable for home mining at all. You can of course modify it in order to make it run silent without too much of extra cost or effort as it is just with 200-300W of power usage spread over a number of chips over a wide area with a huge aluminium heatsink on top of them.

Inside the 1U miner case there are 9 small server AVC fans that are typically very noisy due to the high RPM that they operate at. The particular model is AVC DBTB0428B2U, with a size of 40x40x28mm, dual ball-bearing and rated at 12V 0.48A (5.76W at maximum speed of rotation). They are PWM controllable 4 wire fans and although they start with low RPM and not that much of a noise, as soon as hashing starts they up their rotation a lot and become very noisy and power hungry… at max RPM these could get up to almost 52W of power usage just for the fans. Modifying the miner with fewer larger fans can make it not only more power efficient, but almost silent in terms of operation.

Even if you stop and replace the 9 cooling fans for the ASIC chips there is one more thing you need to take action on. The power supply is also a server type one and even though it is wated at just 300 Watts it does come with yet another small and noisy fan that although not as loud as the others still makes the miner noisy when operating. You have the option to either replace the standard PSU with a silent one (you only need 3x 12V 6-pin PCI-E power connectors to power everything) or disconnect the small PSU fan and replace it with a larger silent one.

One of the first things we did with the device was to make it silent, meaning that we disconnected all of the 10 noisy server-type fans and have replaced them with 5x Arctic Cooling F12 120mm fans (rated at 12V 0.21A), making the Jasminer X4 1U-C run pretty much silent. The main Jasminer control board has 2x 4-pin fan sockets that you can utilize (each of them has a splitter board connected for the small server fans). So, you can have 4 fans on the main heatsink where the ASIC chips are connected via a 4-fan splitter cable to the first socket and a single fifth fan connected to the second socket. You need to either have fans connected or a fan simulator as if no RPM is being reported an alarm will trigger. You’ll need to open the power supply and disconnect its fan from the PSU main board (there might be some glue on top of the connector to hold it in place), or just use a different silent PSU altogether.

It is interesting to note that the main control board of the miner has three sockets for ASIC miner boards to be connected and just one is being used in the X4 1U miner, but the other two slots are fully functional, so if you add a second and a third ASIC board they should be active as well making the miner 1040 MH/s (Jasminer X4-Q anyone?) or 1560 MH/s respectively. The most powerful model with 2500 MH/s Ethash ASIC hashrate – Jasminer X4 High throughput server that the company sells is probably equipped with a different control board with 5x 8-chip ASIC boards. So, if you have multiple X4 1U units you may play around and turn them into fewer miners with up to 3 computing boards per control board and more powerful 800 or 1000W power supply.

So, let us get to mining performance and power usage that we have measured. First is stock settings where we are using the default hardware configuration of the device (just as it came) running at a clock frequency of 225 MHz and then the same settings with our silent modification described above.

Default
– 225 MHz clock
– 297 Watts of power usage
– 55-61 C ASIC temperature
– 530-560 MH/s hashrate average
– 76 Decibels noise level (dBA)

Silent Mod
– 225 MHz clock
– 265 Watts of power usage
– 53-56 C ASIC temperature
– 530-560 MH/s hashrate average
– 44 Decibels noise level (dBA)

The official mining profitability calculator on Jasminer’s website reports around $1.25 USD daily profit for the Jasminer X4 1U device mining ETC at the moment, but that is with power cost of just $0.05 USD per kWh and most of you are probably paying at least multiple times that price. With an average price of around 25 cents per kilowatt hour of electricity you are just about to be able to pay for the electricity used by the miner at the moment. The above calculation is made using just Ethereum Classic (ETC) mining and no dual-mining or Nicehash with ETChash mining that are likely to be more profitable at the moment with 30-40% higher earnings on average.

Unfortunately with the Jasminer X4 for now at least there is no option to dual-mine ETC + ZIL on a mining pool like Ezil.me due to the way the miner works (it splits the pool ID at the first dot threating the remaining data as a worker ID, so you cannot have two wallets in it for ETC and ZIL). It is not only a problem with the Ezil.me mining pool, the other bigger ZIL dual mining option K1Pool uses a single ID for pool username, so you can connect the Jasminer X4 to it and it will mine ETC, but when it needs to switch to ZIL mining the miner will simply idle for a while and then continue with ETC mining when it is resumed. So, dual-mining ETC+ZIL is not working for the moment with the latest official firmware for the Jasminer X4.

The alternative higher paying option is ETCHash mining on Nicehash that reports higher profitability probably due to people using it for buying hashrate to double mine ETC+ZIL. Now, we know that miner has trouble with direct dual-mining, but it also has another issue – it is currently not compatible with Nicehash at all. The miner reports any of the Nicehash DaggerHashimoto and ETChash pools as dead and does not connect to them at all and we have confirmed with the manufacturer that Jasminer X4 currently does not have Nicehash support. So, unfortunately not much you can do here to get some extra mining profit for now with dual-mining ETC+ZIL or NiceHash mining, the main competitor iPollo has apparently released an updated firmware for their ASIC devices that supports dual-mining, though we have not personally confirmed.

So, what is the conclusion that we can come up based on our first impressions from playing around with the Jasminer X4 1U ASIC Miner. We can clearly say that the Jasminer X4 1U ASIC miner is a nice piece of hardware that is built well. The hashrate and power efficiency are really good and with something profitable to mine it is really a nice ASIC miner for Ethash and Etchash to have handy to mine. The Jasminer X4 1U device is intended for use in data-center like environments as it is in standard 1U rack-mountable size and is designed with server cooling in mind, so it is noisy. If you decide to pick one of these up for home mining you should be able to pretty easily modify it for silent operation as we have described above. Furthermore, if you wish to play around you can upgrade the device by adding up to two extra hashing boards into the same control board the miner comes with. The miner is relatively easy to play around with both in terms of hardware and in software modifications and while this is just the first hands-on experience that we are sharing there is more interesting stuff about using the Jasminer X4 1U coming right up in separate post(s). Making it run silent is just one of the easiest things to do…

But what about the question is the Jasminer X4 1U ASIC miner worth buying now? With the current level of profit the answer is very simple – NO, the price of the miner at the moment with the profit that you might possibly be making with it is definitely not in favour of buying the device now. You can either barely cover the electricity it uses, even if it is not that much, or make very little daily profit. Furthermore, there is currently no support available for dual-mining Ethash/Etchash coin plus Zilliqa or using the Nicehash platform that can currently give you an extra 30-40% or even more profit. So, you will be more limited in the ways to maximize the potential the device has. If you however just like to play around with it, modifying it, just like we wanted to do and manage to score a good deal on the hardware, then by all means you should give it a try. If you are interested in mining with the Jasminer X4 1U for profit, then this is hardly something that you can currently do and since it is just in the 200s of Watts in terms of power consumption and very noisy by default it is not a good idea to use it as a home heater (unless you make it silent, but still with less than 300W of power usage not much of usable heat generated for the winter).

If you have some further questions about the Jasminer X4 1U ASIC miner feel free to ask and if we can we are going to try and help.

– Visit the official Sunlune Jasminer Ethash/Etchash ASIC miner website…

Trying Out the New Solo Mining Packages on NiceHash

19, Oct 2022

Recently the NiceHash platform has made available a new feature that they call Solo Mining with “Catch the Block” and we have decided to try our luck and see if we can manage to hit a block and get the full block reward. NiceHash have made this new feature very easy to be used, almost like buying lottery tickets and testing your luck hitting a big reward. What you do here is buy a package of hashing power from their platform that will be used for mining a certain crypto coin for some time, and if you validate a block with it then you get the full block reward. All you need to do is choose a mining package (pay for it with Bitcoin) and wait for it to finish mining and see if you were lucky enough or not to hit a block and get a nice reward. Currently mining Bitcoin (BTC), Bitcoin Cash (BCH), ZCash (ZEC) and Dogecoin (DOGE) are supported by thins new mining feature and you have a couple of different mining packages that you can choose from…

In order to be able to use the Solo Mining mode you must first have an account at NiceHash and then also have some Bitcoin (BTC) transferred to your platform wallet that you would be using to pay for the mining packages you want to purchase. That may require a certain level of verification if you are registering a new account. Also have in mind that in order to receive your reward you would need to be Tier 2 verified (not required to try your luck) and that requires identity and address verification! There are 3 different mining packages available – Bronze, Silver and Gold with 3 levels of each (S, M and L) available so 9 different packages in total. Bronze is for Zcoin (ZEC), Silver is for DOGE although the L level is for Bitcoin Cash (BCH) and Gold is for Bitcoin mining.

Each of the mining packages available is priced differently with the lowest one starting at sub $2 USD at the moment and the highest priced one is a little shy of $200 USD. The higher the price goes, the more hashing power and time you get to try and validate a block, thus your chances are getting higher. The more you pay however the less you will get if you manage to validate a block, but the chance of doing so is higher (the probability is listed for each package and it varies). With 1:3 or 1:4 probability the chances are high, but you can still buy a mining package and get nothing in return, so maybe not as much like a lottery, but more like gambling.

When you buy a mining packages shortly it starts mining and showing you how much time is left and how are you doing trying to validate a block for the specified crypto coin the mining package you bought for. There is no need to have the page open or anything, it will run for the specified time and you can find the results and your luck afterwards. Of course, there are some pretty graphics that are showing should you decide to spend time watching how things are progressing.

We’ve decided to start with the S or Small mining package for the three options available and then get three of each, so 9 in total for a total of 0.0009 BTC or about $17.37 USD with the current price of Bitcoin. Each of the packages ran for about 1 hour and although we got high percentages of being close to hit the reward, we did not have enough luck and did not manage to pull off validating any blocks with our 9 “lottery tickets”.

If you click on the smaller icon in the ACTIVE order section of any of the purchased mining packages you will get a more detailed graph and stats with information about it, so if you are curious and want to spend more time watching as things progress you can do so. Or you can just leave it to finish and then return to see the results of your luck or chance and if you managed to score any rewards or not and decide to maybe try again or not. Here is also another useful option you can take advantage of – the ability to stop an active mining package from the three small dots on the upper right corner.

The mining duration you get for the S-sized package is about 1 hour, for the M-sized one it is about 2 hours and for the L-sized one it is about 4 hours. There is no exact mention of the amount of hashpower you get with each of the packages, but on the detailed page you see how the spending of the amount paid in BTC is progressing. However when you purchase a mining package you are not charged the full amount, instead it gets depleted over the duration of the mining time you get. So, say you hit a block very early on and you can stop the mining package and not use up the remaining amount. You’ve already beaten the high odds, or would you continue and try hitting another block maybe.

There is a dedicated page showing all of your completed orders as an easy to view list including how close were you to hitting a reward and if any of the mining packages you bought actually managed to hit a block and score you are reward. If you click on any of the packages listed on the COMPLETED orders page you will also get a more detailed full stats for the whole mining time of the package, similar to the detailed data you can watch during the actual mining.

The closest we got a reward with our 9 S-sized mining packages was 89% on a Bronze S package for mining ZCash (ZEC) and that one had 1:118 chance according to the data when purchasing the package, so pretty good, although a little disappointed that we were not lucky enough to hit a block and get a reward. Maybe will give it a couple more tries and you are welcome to try your own luck with a few mining packages and see if you can manage to score a block reward or not.

Update: Meanwhile we have also given it another try with one package of the second M-sized levels for each and unfortunately our luck has not improved much and again we did not get to validate a block on either of the three crypto currencies with the higher mining packages. We are not ready to give up yet, so next come the L-sized ones, hopefully with better results or we are getting completely disappointed from our luck today, but maybe you would have better luck than us.

– Visit NiceHash for more details on their new Solo Mining feature…


top