Thinking about becoming a crypto miner and need power? Really Cheap Power?
Well nothing is cheaper than make it yourself sans(without) the overhead of transmission, distribution, power generation and profit taking charges foisted on you by the power utility, IF you know what you are doing. But I will save that for another post.
Today's post is about "highly distributed" #P2P crypto storage as a service with a payout in crypto automatically made by the network to the storage and/or distributed compute job node operator, running cost effectively off-grid.
OFF-GRID CRYPTO PROTECTED DATA STORAGE needs OFF-GRID POWER - And it's No Place for "Power Sucking Alien" storage hardware. :)
Of course Electrical Engineers still plying their trade have a "one up" on everyone, to create an energy generation, storage and delivery system "Behind the fence", optionally hooking up the grid as a backup "trickle battery charge" system to create their own #crypto-mining, storage and/or distributed compute job power house node.
A little more 3rd party energy storage information can be found here
CRYPTO MINING REALITIES: ONLY THE BIG CAN PLAY (in the big cryptos)
Now the barriers to crypto mining the big PoW "Proof of Work" coins, #BTC, #ETH and other mine-able cryptos found in the Top 20 on https://coinmarketcap.com/ make crypto mining largely out of reach for most.
The big mining pools dominate, all wired up to power in countries hosting their rigs where the price of power is super cheap and plentiful, always on, (Iceland anyone?)
There is financial risk jumping into mining crypto some of the other #PoW Proof of Work blockchains, so selecting the right mix of coins to mix requires a lot of research, not covered in this post.
That said, there are other way to make money in the world of crypto outside of mining which will also benefit from low cost smaller scale OFF-Grid power generation, energy storage and delivery to keep data storage and computing functioning.
The good news is these crypto environments are not PoW energy pigs solving crazy hard puzzles. Better for the environment No? Environment concerns aside, it's really more about who you pay, yourself, or some centralised entity concentrating wealth?
ANOTHER RARE OPPORTUNITY TO EARN CRYPTO IS EMERGING: NEXT GEN CRYPTO PROTECTED STORAGE SERVICES
It's an any one can play (with a small budget) opportunity, and it's just around the corner, no mining required. If you can get the equipment configured, AND setup right on affordable SSD "Solid State Drive" hardware utilizing a bit of special #SSD optimization software (only on #Linux #Ubuntu), there is opportunity to actually make some decent coin in the crypto Storage as A service market.
No its not #Sia, or #Storj and definitely not #Filecoin, (although it could work them as well) and yes IPFS might be in the mix, but not the way you think.. so read on...
Maidsafe: The SAFE Network: Earn crypto storing encrypted data.
Some of us may remember MaidSafe's 2015 early ICO (Ayr, Scotland), and their early (now conquered) struggles with Mastercoin liquidity (Ripple also uses Mastercoin, coloured coin tech) needed to finance the growth of their development efforts. What has seems to be a never ending project by some to re-invent the Internet, is actually about to come to fruition in the form of the SAFE Network, and nextgen version of the Internet, as David Irvine and team test the last piece of their puzzle, the Vault. The SafeNetwork Vault is akin to #IPFS "Inter Planetary File System" https://en.wikipedia.org/wiki/InterPlanetary_File_System , except, when you join the SafeNetwork, it pays you automatically for allotting some of your storage to the SafeNetwork, and you earn more the more trusted (ie reliable) your data storage becomes. Well, in my book that means keeping the lights turned on, and the power turned ALL THE TIME to keep the SafeNetwork Storage Vault operating.
Before I dive into this, yes I could have looked at more uni-dimensional crypto storage as a service plays like Siacoin, Storj and the infamous Filecoin to name a few, they all came well after MaidSafe. In fact, MaidSafe's David Irvine started on the project in 2006, before crypto really existed. That said, Maidsafe has much more experience and is more multi-dimensional, in terms of value offered, which in the end, will give you the reader more opportunities to earn #SAFE crypto for offering resources beyond just storage and bandwidth, and get more value operating a Safe node running Vault than what the others can currently offer, or plan to offer.
How much energy, you as an operator of a SafeNetwork Vault, are willing to pay for, is up to you, and is partly determined by your bandwidth speed and your designed system response to any request to store (write) data from other users looking to safely store their data on the SafeNetwork to your Vault equipped SAFE node.
For example, If your running a power hungry #HP Z70 Rack mount system with 6/8/12 #AMD #Opteron 64bit cores with 8GB ram or more, and several "spinning rust" hard drives, of course the power bill will be higher, as those 500GB/1TB drives, as they need to be spinning (not in sleep mode) and responsive to increase your chances of winning a storage request from theSAFE Network serving user request to save data and earning some Safe coin. Your Safe Vault crypto earning might be adequate, but the overall ROI "Return on Investment" will be marginal, if fact you may find yourself losing money pretty quickly with such a power sucking alien rig, as you slip down the performance rating in the SAFE Network, with new hardware being added with better performance.
TK Note- My old HP Z70 (I sold it a few years back to some #CAD/#CFD engineer, something I also do for a living), had a 430 Watt draw just to turn it on sucking 2.5 Amps, add more drives and you end up looking like a 1000 Watt microwave turned on 24 by 7. So depending on your local rate that could be US$ 6 to 10 cents per hour being sucked out of your wallet, At 10 cents, that is $2.40/day or $876.00 per year to run maybe 2 TeraBytes of Spinning Disk or about 333, 6 Gigbyte HD movies which will be likely what many people will want to backup, and do it cheaply. US $73.00 for 2 TB of storage service is way over market for basic ISCSI Storage as a Service, so the Z80 would never cut it and be a loser right from the start ).
What the Safe Network will pay you to have 2 TB available at all time is up for discussion, as this is hard to decipher from their Forum and Whitepaper.
Safe Network is storage centric for now with Vault in Alpha 2 Test Mode (Maidsafe promise to deliver a functioning network with Vault support sometime in 2020, and IMO appear to be on track). Maidsafe also have other Dapps running today including the Jam music player and the ability to host Website pages in a distributed and reliable manner through their very secure SAFE Browser. all very cool, running the Parsec Consensus algorithm to protect your data and transactions.
There is also a promise of being paid for Distributed Compute Jobs running on other systems you might sign up to run on the Safe Network. You can try out their Browser, and some other apps here .. https://safenetwork.org/
I have been following the Safe development effort since late 2017, track their activity on Github. IMO they are a gritty determined bunch of clever developers who really are "All in" to make this "reboiling" and transformation of the Internet into a new secure, "people first" Internet operating on a lean PoS protected distributed anonymous ledger which pays node operator in crypto to help the SAFE Network grow and flourish. So there is some staking involved to play on the Safe Network. Maidsafe's dogged determination and focus, looks to be bearing fruit as they work their way through the Alpha 2 test network milestones. IMO Maidsafe is worth signing up to the forum https://safenetforum.org/ (you need to learn on their forum to advance, gain access to more info, so no tyre kickers) to get a better feel for how the SAFE Network works. A quick video explaining Maidsafe is here
SafeNetwork is a "#DAO" Distributed Autonomous Network of sorts, and completely anonymous, totally focused on privacy like #Monero or #Cardano, and other privacy centric coins. Once they release the SAFE Network into the wild, not much can stop it, IF they get the Node adoption, and that will mean can the Node operator make money, which means to they have their system build act together coupled with cheap low cost off-grid power generation, storage and delivery. Part of Maidsafe's "not for profit" biz model being adapted in their affiliate and partner companies is to also create, then service and support separate private instances of the SAFE Network.
Time will tell if the SAFE Network, once released into the wild, running on its own "a la today's #BTC Bitcoin network" , will fly with the general populous.
IMO it looks like the SAFE Network will be realised and get initial decent user adoption, likely by end of 2020, provided Maidsafe keep working through their development road map milestone's successfully and more or less on time. It's a big "boil the ocean" project in a new form, which is now IMO about 75% complete per the road map.
The real long term trick will be can MaidSafe attract and grow community developers, growth largely driven by having a good documentation platform and training engine behind the Safe Network as well. The SAFE Network forum is a good start for education, Documentation as always in any project needs more work beyond the scantness found on #Github.
Now back to power requirements and what Compute and Storage equipment to use along with what type of Energy Generation , Storage and Delivery equipment to use to power such gear.
TK alert- The next is section is for deep geeks only, or the madly inspired ;)
But First an SSD Storage Primer
At first thought, many will think SSD "Solid State Drives" would be the way to go to become a Safe Network Node Operator offering up crypto Storage as a Service by a running a SAFE Vault, as Drive response + ample Bandwidth is likely to generate more rewards on the Safe Network. Yes this is the right formula, generally speaking.
However, success is in the details. Not all SSD drives and SSD appliances are created equal in terms of speed and durability. SSD speed and durability has to do with block write size and how the controller handles the WRITE to the SSD Media. Getting the mix right to yield a decent ROI with some durability is not easy.
Generally a 2U rack of 24 SSDs set up to deliver 8 TB of storage, will suck power in the neighbourhood of 200 Watts operating with say a low cost but high quality SuperMicro Equipped 4 core 64 bit compute platform in a 2U chassis with a 24 slot SSD bus, running 8GBytes of RAM. Expect best case 100,000 IOPS with 4K block size, the typical block write unit which fits in a standard IP packet, running standard Linux Ubutntu BTFS or some form of ZFS.
SSD Controllers are DMA hardware interrupt driven devices with the WRITE algorithm burned into EEPROM Electrically Erase-able Programmable Read Only Memory which is fast, quickly loaded into the Drive controller local ram at boot time, code designed to take Layer 5 (ISO Model) queued blocks in a FIFO "First In , First Out) manner and write to the SSD media. This is an over simplification of what is really going on, but is enough to describe how things work for now.
TK-Note, there is not much difference in SSD read speeds between vendors, so reads are not a problem, reads are generally 5X to 10X faster than writes, and use much less CPU and RAM resource, as there is minimal queuing going on, and no indexing going on where the latter are a standard part of any write to storage.
SATA https://en.wikipedia.org/wiki/Serial_ATA HDA Controllers are the baseline, you can spend more money on fancy SSD exploiting a faster motherboard bus option, to connect the SSD drive controllers, it won't buy you much, but read on to find out a different , software only way of accelerating writes which amounts to US$ 10 cents per Gibabyte of storage overhead, running on Linux Ubuntu. (Because serious Storage Appliances do not use MS Windows or Mac OSX)
SSD Drive Durability Shocker: 4-5 months use with 50% use rates having 50% WRITES and 4K block size
Now for the interesting part. SSD durability today in the real world of Storage as a service really SUCKS!
Durability of SSDs as we know them today from the cheapest vendor to the most insanely priced vendors like Pure HP Nimble, NetApps, EMC, etc.., all of them, without exception, are "durability challenged" when it comes to small block WRITES of 4k.
For instance, any of these "fancy" SSD vendors using any type of SDD media (on cards) from Micron or Fujitsu, or Samsung, etc.. in 1, 2, 4 or 8 TB formats, where the READ/WRITE ration is 50% Reads and 50% writes, will see the SSD drive produce "media flaw" errors at an alarming rate (over 1%) starting in month 4. which for a bank or insurance company, is a non starter,, these clients simply throw away the drive and install new ones, as IOPS are everything for Transaction per second services, and errors reduce IOPS because of costly data moves chewing up what would have been "IOPS through put CPU cycles", reducing the actual write speeds significantly, by as much as 355 depending on how scattered the data is on the SSD. That means less chance of a higher overall score as a Vault in Safe, meaning less chance of getting selected as a virtual storage device and less crypto earned.
That's right, ALL SSD with 50% writes running 50% of the time with 4K block size are essentially dead by month 4 or 5.
Worse yet when such SSD media flaw errors show up in the fancier systems, like Pure Storage et al, where the master code running on the main cores controlling the slave hardware SSD controller will trigger a SSD drive controller read copy sequence to automatically move the blocks near the flaws to other un-used or lightly use media sectors, usually operating under RAID 5, 6 or even wasteful RAID 10 redundancy configurations. Ok now you have duplicates, now what? Well the fancy assed SSD systems you bought for way too much money will enter a "De-dupe" De-duplication cycle to remove the old files. These set of data protection methods employed by the fancy vendors is what takes time and CPU cycles way from crypto earning WRITES.
Now for the fancy vendor sneaky stuff, to get write speed, these Fancy systems use WRITE AMPLIFICATION, to write massively in parallel 3, 4, 5 up to 10 times more copies of the same file, so they hit their quoted "IOPS" speeds of 150,000 to 200,000 Input Output Packets per Second (measured in 4000 byte or 4K packets). Need to run more efficiently, sneaky vendor x, y or z will see compression add-on software to deal with the problem they caused in the first place, that is eating up all your storage 3-5X per write to go fast. (For the record running a CVault on Amazon or MS Azure, would make absolutely no economic sense at their current over inflated prices for Storage and IOPS).
SSD Storage Myth Buster: WRITE Amplification is used by all fancy SSD system vendors to get WRITE speed and the cost of storage capacity
The Storage Density of these fancy vendor systems is typically configured with maximum 8TB per card, with 24 Cards per 2U (about 4" or 110mm tall) of rack space to create 192 TBs of RAW Storage (re member the amplification and counter-acting compression vendor surcharges) before RAID 5, 6 or RAID 10 is turned to create some basic form of storage redundancy. As a potential crypto Storage as Service Node, looking to earn crypto , one will need to create a serious "Storage as a Service" machine, with dual power supplies in most cases, with each 2U "Storage appliance" sucking 250 Watts on each of its dual power supplies able to provide 500 Watts peak (during 100% busy write to SDD use case)... operating a somewhat less power levels, say on average 400W total, depending how busy the SSDs are. Much better than the Z80 discussed earlier for sure, but how cost effective, given the price tag to get there (the old HP Z80 can be has for US $400 these days with 6 cores and 8GB Ram coupled to 2 TB of storage).
Lots of crypto Storage as a Service "Build Cost" and ROI questions for sure, but not a lot of answers out there in this regard, as IMO the Crypto Storage as a Service market is nascent and has really yet to emerge. So more research required to really validate the SAFE Network Vault business proposition to the prospective Vault Node Operator looking to earn some crypto offering crypto Storage as a Service.
One thing is for sure, Expect to spend at least US$ 30K retail building such a muscular Storage rig, Given it will be useful with the right durability enhancement Software, which is out there, for SSDs, you can expect a rig like this to provide decent service and earn crypto for about 3 years, maybe longer, before you have to upgrade your Storage Attached Bandwidth from 20mbit/second speeds at a minimum to be more than 100mbits/s to stay current and be on a par with a minimalist local 100mbit second rate over Ethernet Attachment to the Storage Appliance, which means you are partly reliant on the Infrastructure Operator to keep up.
Be sure, the most profitable business out there is "Storage as a Racket", these "fancy" vendors make insane margins, and are hardly transparent about performance and durability in the process, its a fancy vendor "storage cartel" for sure, so buyer beware. How to DIY and beat the odds to become a profitable crypto Storage as a Service provider,, regardless of the crypto play you select, is found later in this post.
IOTA QUBIC- Distributed Computing as a Service: Crypto style
The Qubic platform has been in development for 21 months well, staffed and generating lots of progress, that said, early 2020 deployment of a production network is likely the earliest we will see Qubic which is now operating some of its functions on the test network managed by the IOTA Foundation. IOTA is quick to stress they are still in Research & Development mode when they seep of Qubic.
IOTA's Qubic promises more than just distributed computing as a service, operated via smart contracts using IOTA as the crypto-currency to facilitate such transactions. Qubic is being integrated into the main IOTA nodes, now under development as BEE Nodes. It is IOTA's intent all existing production nodes of IOTA operating in snapshot mode or permanode mode will be upgraded at some point ot the emerging Bee Node which includes QUBIC capabilities.
While IOTA itself has focused on building a distributed public ledger platform capable of handling billions of transactions in the Machine to Machine economy, Qubic's feature set promise is not lost on those looking to create crypto Storage and/or Distibuted Compute as a service, where the latter will be supercharged with custom ASICs and FPGAs running compute job specific code running in the much more power efficient ternary or "Trinary" form of computing as opposed to the binary computing we use today. (Think Russian Semenov ternary tech from the 50s re-birthed.)
First looking at crypto Storage as a Service for a direct compare versus Safe Network Vault, QUBIC is very much a development platform and it will be up to the developer to create their Storage as a Service with the tools provided by the Iota Foundation sponsored development effort found on Github. Qubic as a project, is 100% open source, supporting 3rd party system connectivity today to their distributed public ledge #Tangle through IXI gateway code which support Restful APIs. Tangle is a DAG based hugely scalable ledger technology which asks pending transaction subimtters to execute 2 PoW puzzles before their transaction are committed to the Tangle ledger branch. For more on IOTA gohttps://www.iota.org/, for more Qubic specific try this link https://qubic.iota.org/
The main difference between current IOTA offering as it operates on the test network with the road map promising upgrades to the Bee nodes including Qubic functionality "Phase 1", is simple, IOTA is a PUBLIC distributed ledger which is pseudo-nonymous, while Maidsafe's SAFE Network is anonymous and very private focused. Both offer the ability to hard fork and create different public or private versions of the current networks, with only IOTA being in current production, while SAFE Network and Vault remain in test mode.
IOTA's promise is big, they are backed by some of the biggest companies in the world making use of the IOTA tech today, Bosch being one of the biggest users of the tech creating an edge industrial hub demonstrate and prototype early IoT 'Internet of things' connected to the IOTA Tangle. That said Qubic like SafeNetwork Vault is not in production.
COMPARISONS: APPLE vs ORANGE (with Storage overlap)
IOTA's new development of the #QUPLA scripting language (Looking like JS , sort of) to allow any developer to create applications which can compile easily into their #ABRA flow programming language which itself can be written directly to existing #FPGA and #ASIC chips is certainly a big differentiation over SAFE Network. That said Maidsafe is focused on the C2C & P2P market use of crypto, targeting key social networking apps, and distributed website apps and general purpose distributed computing using existing computers, desktops, laptops, tablets and even beefed up mobile phones with a real OS.
Maidsafe is likely to find it's SAFE Vault storage service with crypto payout hugely popular with the average consumer provided they make it easy to install and update. The Vault will certainly take market share from existing cryptos like #Sia, #Storj and non-crypto plays like #IPFS which is supposedly the core of #Filecoin.
IOTA is arguably the EU's largest Open Source Crypto project, with the biggest community and the IOTA Foundation do generate decent documentation and development examples and use case studies today which help the interested IoT and Smart City focused developer to get up to speed quickly.
Both have followings on Discord and are worth looking up there.
Brave BAT: a wild card in the emerging crypto storage as a service market?
That said, and its purely guesswork on my part, you could see #Brave also move into the crypto storage as a service space pretty easily with their #BAT token , which recently added convertibility and gained liquidity, and they have IPFS running in the Brave browser, so rally only a short step away to tie it all together? I am a Brave and BAT fan, I acutally have earned some Bat on the Brave Browser and via #Publishox, I know Eich's work well and also that of Marshall Rose (being a developer and IP networking guru myself), Rose being one of the early IETF genius wonder boys writing RFCs (and actually figuring out BEEPcore with dynamic security applied per session). the Brave crew is a seriously smart group of successful veteran developers with a big plan, they have nailed the nextgen advertising model on the Internet with Brave and BAT putting the user back in control, so I expect more to come, and it smells like #Freemium crypto storage as a service based on #IPFS. Just my hunch.
IOTA: going places where no crypto has gone before - Sia, Storj and similar, watch out
IOTA is a big project and the mostly lightly traded, most resistant crypto to BTC movement in the Top 20. IOTA just hangs in there with a huge market cap, because the IOTA hodler's truly believe what they see, and that is some excellent progress.
Look for the creative forces on the peripheral edge of the IOTA community to jump in right away once Bee Nodes show up in production, as QUBIC is a game changer, not just in crypto Storage as service , modelled the way you like it, but also Smart Contracted the way you want per your business model. I think by late 2020, we will see multiple storage as a service offerings with different business models emerge which will really challenge the current status quo of Sia and Storj as the early storage as a service leaders in the crypto market space
MAIDSAFE: release it and they will come.
This gritty bunch of developers, with some horsepower in India, are making their pennies stretch after their somewhat malformed ICO. They do operate as a not for profit registered foundation in Scotland. That said, a good share of the early Mastercoin coloured Maidsafe tokens ended up with a VC based out of Chicago, so there is an obligation for the Maidsafe development team to perform, as the tokens of MaidSafe will be covered 1:1 in swap for new Safe Coin issues, and one has to wonder how many airdrops the network itself has programmed into to perhaps promote the sign up to SAFE Network and encourage SAFE Network node operation of Vaults.
That said the Maidsafe developed SAFE Browser is an interesting beast, with Maidsafe promising tool up the SAFE Browser to handle the wallet like Brave. This makes the user experience hugely easier, especially with support for existing hardware wallets like #Trezor #Ledger Nano and #Keepkey.
Time will tell.
Having said that, I hold neither, and have never held either one of them, but I am reconsidering my positions and remain very interested, because distributed storage off-grid powered, makes so much sense, especially if you are sovereign being.
Getting cheap hardware to run SSD storage fast: Its all about the software and the smarts behind it.
The Linux Ubuntu Community is one of the largest out there, if not the largest. You can strip Linux down to the bare minimum and run only what you need to run and compile to your hardware target pretty easily these days to get maximum performance out of the cpu core configuration of the motherboard you have selected. The real trick is getting the SSDs to WRITE fast and not wear out. That means getting the dumb ass SSD controller logic out of the way via a software interrupt mask in Ubuntu Linux, and letting special (yes its proprietary, and the facny vendors wished they had it) Layer 5/4 queuing, concatenation per session software manage the SSD writes, increase speed and increase durability.
It turns out, one of the companies I advise, Cloudprox of Hamilton (The Steel City), Ontario Canada, has figured the above out in the large DC space.
Cloudprox is a really tiny startup which is software and data centric. The CEO David Bruce is a bona fide Data architect (without the White suit) having coded all his life for some big names in Canada like rogers, big insurance and big banks. he knows data and Linux, Ceph, etc..., more importantly David knows how to make SSDs go fast and last, on the cheapest of SSD Controllers and Cheap Hardware (SuperMicro is a good place to start), and scale it across merchant/commodity fast storage networking fiber channel 40/56GB cards and switches from Mellanox.
Think paying US $60K for almost 350 TB of optimised storage that lasts at least 5 years more and likely 12 years under normal operation, and its all done in software. Most Importantly depending on block size its 5X to 20X faster, which means there is a higher likelihood you will earn more crypto in any crypto storage as a service setup.
Cloudprox's Turbstorage Software that has been deployed in systems successfully on SuperMicro Hardware the past 3 years using cheap SSD from Micron.
Yes Cloudprox's Turbostorage costs some money, priced about US $0.10 per Gig, but well worth it. Discounts in volume.
Cloudprox Turbostorage ships worldwide today via a SuperMicro Storage Integrator in the Mid USA under full warranty, at 80% less the price of the fancy storage vendors with 5X the performance and durability at a minimum. The same integrator supplies the milspec, govt space and private business and has been in biz for 20+ years.
And the 2U power consumption of the base configuration is a minimal 400 Watts with dual power supplies for 350 TB of storage at about US $60K (two systems with a switch and 2X NICs depending on the fiber-channel vendors selection and configuration. (Mellanox has a lot of options, yes others like Netronome could be used etc..)
Yes you need to front end this storage, lots of ways to do that, and Cloudprox helps you figure that out as cheap as possible, to really get in the game.
Hmm, sound interesting? $300-400K versus $60k for 360 TB of RAID 5 use-able storage?
Makes for an interesting match with any of the crypto Storage as a Service nodes, Sia and Storj included.
If you truly are interested, leave a comment below and I will connect you. Because Cloudprox is stealth, for now. ;)
Live smart and prosper. :)
TK over an Out.