Categories
- 17.3K All Categories
- 9.5K Mining
- 574 Pool Discussion
- 378 Promotional
- 1.4K General Project Discussion (non-technical)
- 485 Education
- 811 Protocol and Client discussion
- 170 web3-js
- 29 Whisper
- 16 Swarm
- 3 RLP
- 303 IoT & Hardware
- 1.2K Smart Contracts and Dapps
- 28 Serpent
- 359 Solidity
- 664 Projects
- 1.2K Reference clients code and builds
- 249 Eth & AlethZero- Cpp Implementation
- 471 Geth - Go Implementation
- 242 Mist
- 15 Node.js Implementation
- 36 Python Implementation
- 49 Mix
- 36 Other Implementations
- 170 Meetups
- 40 Other Events
- 226 Jobs & Skills
- 281 Press and Articles
- 75 Audio/Video
- 296 Ether Sale
- 1.2K Other Languages
- 96 Chinese
- 255 German
- 33 Italian
- 111 French
- 3 Hebrew
- 42 Japanese
- 75 Portugese
- 46 Romanian
- 185 Russian
- 231 Spanish
- 47 Turkish
- 125 Watercooler
[Interest] 200MH Plug and Play Mining Rig
Putting this out there for interest, but I have access to an interesting deal where I can assembled a couple hundred 200MH Ethereum Mining rigs. Preliminary specs are below
-200 MH ETHash rate
~1300 watts at the wall
-Roughly 4U enclosure
-Target under 2 month ROI (i.e. 2-3k USD)
-Simple ethernet connection, and web interface for configuring
This would be just a plug and play system, plug into power and ethernet, go to local ip address to enter pool (or solo) information and fire it up. No fumbling with GPUs, OSes, setup or any of the like.
These are for a massive farm deployment (so they are designated to be modular and easily scalable), but figured id open some sales to public if enough were interested
-200 MH ETHash rate
~1300 watts at the wall
-Roughly 4U enclosure
-Target under 2 month ROI (i.e. 2-3k USD)
-Simple ethernet connection, and web interface for configuring
This would be just a plug and play system, plug into power and ethernet, go to local ip address to enter pool (or solo) information and fire it up. No fumbling with GPUs, OSes, setup or any of the like.
These are for a massive farm deployment (so they are designated to be modular and easily scalable), but figured id open some sales to public if enough were interested
Comments
Nice Performance there ! I'd like to open a small Ethereum mining farm. Is it possible to ship it to Europe ?
Regards,
P.
Here is something looks more realistic in terms of hardware: http://www.vmxlabs.com/
Agreed. I'm pretty sure there are no GPUs out there hashing ETH at >30 Mh/s at the moment.
Incorrect.
The HD7990 does far more hashes (>40Mhps). It does draw far more than 200W.
So, as I said...nothing out there hashing ETH at >30 Mhs right now.
And btw, there is no rig capable of doing 200MH with under 1300watts.
If anyone wants to questions my reliability, just check bitcointalk/litecoin talk forums...same user name. Im a hardware/software engineer and have contributed lots of code base to ASIC miner drivers as well as built the only USB Stick based scrypt miner (and successfully delivered hundreds of them after a successful crowd development/sale).
Figured it was about time I get into some fun with the ethereum scene
btw I love the ethash algo and taking a pass over the OpenCL implementation...im pretty sure I can squeeze a nice parentage margin more of optimizations of the kernel
I have tested a LOT of the better quality GPU's on the market as I get them for free to 'play with', and none of the 280x/380's got even near 100watts that someone claimed @ 20MH.
Also, if it's in an enclosure, you WILL be using 100watts just to cool the system with the high speed fans required..
An no I'm not killing off GPU mining...thats the whole beauty of ethereum...there is a limited supply of GPUs that everyone has access too.. at most I would add around 100GH of hash to the network with this project. Once all the new people are onboard due to the hype and price increase of ether it will level off. Trust me this is round 3 of the same cycle for me...I was there when bitcoin was mined on GPUs, and Litecoin. The only difference is that there wont be an ASIC developed before POS since it makes no sense to do so, ethereum is killing off GPU mining itself
Good luck with that. Do you actually think that no one else before you has ever looked at improving the kernel performance? Your arrogant tone tells me you don't have a clue, yet, but you will. I'll give you a hint: nothing matters but the DAG. So, if you can figure out how to speed up 128 byte random reads of global memory in OpenCL, please do let us know.
if unless you have something in hand and you can buy via credit card and cancel and /revoke payments if/when they don't deliver stay faaaar away.
and remember cloud mining is ALWAYS a scam, if they can build such great machines they would mine for themselves remember they are not out to help YOU just help themselves grow big with YOUR Capital like KNC etc did.
Now we DEFINITELY KNOW you are full of sht..
Unless you can drive over and actually see what you are buying it's too risky
I personally lost 30k in cloud mining scheemes and I am pretty much done with anything that does not involve any form of refundable payments. google LTC and Paycoin scams people lost millions.
Known: So I've updated my sheet with kH/W of which the best cards showing are Nano's (26Mh @ 175W) and R7 370's (15 @ 110W)
200MH/s equivalents So far the dollar budget is there for R7 370 GPU's but the energy budget is not. Noone is going to waste time undervolting and overclocking 2600 GPU's (unless some custom factory deal) so I'm calling 14x R7 370's / unit.
Physical layout No problem there.
Motherboard CPU RAM and connectivity Totals per Unit So except for the energy budget, the rig itself looks doable within the price for those with enough technical experience in configuring high GPU count nodes (which I was researching last night).
The energy efficiency might not be any better than a home built rig, so the 2000W blow out is no big deal.
Low maintenance low configuration drop-in box is pretty attractive. I suggest you have a good read of Genoil's CUDA miner thread for an experience of someone thinking they could do a better kernel...and getting smashed on the wall of ethash. It's certainly better than the OpenCL kernals for NVIDA cards but no better than OpenCL on AMD.
I would rather buy a signle 390 which pulls 280ish and hashes for 31.5 mhs 28.5mhs min
If you want to bundle your miner with ethOS, we can work something out. Would likely save you a ton of work.
Link for convenience: http://www.gnu.org/licenses/gpl-3.0.en.html
For example, we only need to provide source (and we do) to actual ethOS customers (i.e. the users), not to people voicing their misinformed opinions on public forums.
Anyway, I would rather not hijack OP's thread with discussions about ethOS. If you have some concerns about it, you can post about them in the ethOS thread located here:
https://forum.ethereum.org/discussion/comment/20827
linux kernel: https://en.wikipedia.org/wiki/Linux_kernel#Licensing_terms
glibc: http://www.palamida.com/posts/view/19/What_is_the_license_of_the_GNU_C_Library
bash: https://www.gnu.org/software/bash/
With regards to reporting ethOS project, nothing will happen because there are no violations. I am sure that the Free Software Foundation cares equally about small potatoes and large potatoes.
Again, if you have some concerns, voice them in the ethOS thread.