Today i show you what i know so far about FPGA's and their possible mining capabilities.
As shown in the video, their hashrates on Ethereum for example could be enormous!
ASIC's are nothing compared to these possible hashrates.
Due to high pricings, and also the expensive software that will probably go along these cards,
The return of investment would only be slightly higher than regular GPU's.
They could possibly be better than the use of ASIC's, due to their flexible reprogammable properties.
Let me know in the comments below what you know so far, or what you think about these.
As i talked with Bits Be Trippin about these, he shares the same opinion as me about them.
Until now, this is all we have so far and i wanted to keep you guys updated!
I will hopefully see you guys in a next video where we find out more details about this topic.
Make sure to also check out Bits Be Trippin's YouTube Channel:
Buy Complete NVidia/AMD rigs at:
You could buy your parts from amazon using the following links:
Buy anything using the following links:
Buy your graphics cards with Hashrate Benchmarks NOW!
Send me your Mining Rig photos on Instagram!: https://www.instagram.com/buriedone_blockchain/
Join our Discord at the BuriedONE Blockchain Community if you seek help! Discord: https://discord.gg/6PwuaTq
Just my assumption, buriedone..
*The ethminer says its "on difficulty 20, where as my running pheonix miner says its mining difficulty is 4300 mh..
*Eventhough several shares has been "found", the accept (A) counter number is not increasing from 0.. except for (F) counter, which is i have no idea what that is.. unless accepted shares is represented by (F), which likely means "found" or share accepted.
Again, just my assumption & from observation.. im just an amateur beginner-ish, never use ethminer except claymore qtminer & currently using phoenix miner, just what i read from ethminer bitcointalk & github..
Its running in benchmark mode, is the only assumption i can think of.
we'll see in the future whether its real..
Want to tip or donate,
Any donation/contribution for building/upgrading my rig is much appreciated. =)
p/s; if this ad is wrong, let me know.
These numbers are so far from realistic it is laughable. First, the UltraScale V9 cards show above are $7000 ea. if you can buy them. You might get 1 or 2 if you have a good Xilinx rep. Then, the only FPGA that can even come close to any kind of performance are the chips with integrated 8GB HBM, and as of August 2018 those aren't even released, and Xilinx has no date for when they are going to be available. Finally 1.9 Gh/sec for Ethash is a pipedream. Simple math will tell you that it is impossible even if you wait until the integrated HBM comes out.
Oh, and the cost of a single FPGA at Digikey? $32,000 USD. That for probably around 400 Mh/sec - 500 Mh/sec at best, on chips that don't even exist. There is no actual footage of anything in this video. The statements made here are impossible. Do not be fooled.
On June 19th, I purchased the XILINX 1525 with mods and DIMMs from the 1st bulk order on this a bitcointalk thread.
I have the receipt to prove it and I'm sure FPGA.land/Squirrels Research will back me up. It will arrive sometime
in August and then I can ship it out. I really wanted to run this card, as projections are $20-$50 per day.
Unfortunately, I need to sell it to pay off some debt. I'm asking $6,000 to cover taxes, shipping, fees, etc.
I know that may seem high but as far as I have researched, Batch 2 is not even close to ready, and people
could be waiting until next year sometime. I will accept ZEN, BTC, ETH and possibly PayPal.
PM me if interested. I'm open to reasonable offers. Highest offer gets it.
A CPU, GPU, ASIC or FPGA all consist of logic building blocks (Logic gates AND, NOT, OR gates, Memory blocks, Flip-Flops, clock drivers etc). The thing how differ is that a FPGA are reconfigurable how this building blocks are connected "wire" together. In a ASIC this logic connection is fix you can not change this configuration you can ONLY change the program sequence on a CPU, ASIC or GPU. But the FPGA you can change everything, both program sequence and configuration of how the logic are connected. So a FPGA is ultimate flexible compared to ASIC but a ASIC with same logic configuration consume less power then the FPGA.
The smallest of Xilinx's Virtex Ultrascale+ chip is 12000$, https://www.digikey.fr/products/fr/integrated-circuits-ics/embedded-fpgas-field-programmable-gate-array/696?k=fpga&k=&pkeyword=fpga&pv457=i1313763&FV=fffc007a%2Cffe002b8%2Cffecdad9&quantity=0&ColumnSort=0&page=1&pageSize=25
Just a heads up for anyone looking for some more technical details - Matt over at 'The Technicals' had an interview with Olivier Tremois from Xilinx, the company/people who make the 'more desirable' unit (the one that's like ~$4k)
Correct me if I am wrong but I don't believe that these FPGAs would be running ethminer. They would require not only a custom bit stream but also custom mining software not publicly available and not made for commodity GPUs
Finally electroneum mines
So far it was only possible for a small group of people to mine Electroneum with the smartphone in a beta version. Finally, the mobile miner for Android can be downloaded in the Google Playstore for Everyone.
Mining Electroneum (ETN) Cryptocurrency works 100% free!
You get about 4 to 5 ETN per day.
At the moment, value on the crypto exchanges is undervalued. ETN is a promising coin that stands before the boom. Experts even predict that ETN will reach $ 5 by the end of 2018.
You can always run the app in the background and even get your coins while you sleep. The battery is only charged like a normal app and is not warm by that.
Now only for Android, later also for iPhone.
Sensation: 100% free!
1.) download free app from the Play Store (electroneum),
3.) Enter code F796EF in the APP below to earn free coins
for more speed, you get more ETH
4.) Start mining. Better you can not earn money by the way.
It can come because of the large rush to delays, bookings could be made a little later, nothing is lost.
Here we go
Oh my god, do you even understand how ethereum mining works, and that its bound to IO not processing power, did you even look up what the Acorn M.2 does? Youre a fucking idiot if you think we are gona see asics or fpga's for ethereum that does more than 2x the speed per watt compared to a gpu, this is typical in this noob world, vitalek long time ago created the most asic resistant algorithm, and then idiots come up later with shit like equihash saying its better more asic resistant than ethereum, let me give you the numbers, and these numbers come from a guy that understands this shit have put it on paper and calculated how much performance gain the most efficient asic could gain, for ethash its x2 for equihash its x100 for sha256 (bitcoin) and x11 (dash) its 1000X. What you dont get is the fpga is just the step before an asic, if you have programmed a fpga to mine some algorithm you now also have the recipe to make an asic that does the same, its the same logic youre going to implent its just pure hardware gates and not flip flops. And therefor more efficient, so no the fpga lost in 2011 to asic in bitcoin and it will loose again, HOWEVER, with fpga controllers comming up with fast pci express interfaces we can start offloading some of the more intensive work on a fpga leaving gpu's to do what they do best like float and memory intensive shit. But for that you would need a real x8 speed to each gpu not just the x1 most have =) And again if it becomes a big thing I can take the fpga logic, and make an asic from it. And have a device that does the same at half the power use.
GET FREE ETHEREUM FOR INVITING PEOPLE
ETH Invites is a new and unique project that gives users Ethereum just for inviting other people to join our project! Earn up to 0.01 ETH for 1 invited user which equals to about 7 USD
this is straight up bs! XILINX marketing MOVE to sell more fgpa's CHIPS ) NO MEMORY ON THOSE CARDS TO SUPPORT THAT HASHRATE! ASK YOURSELF A QUESTIONS WHY THE HELL I DON'T GET TO SEE ANY ACCEPTED SHARES ON THAT VIDEO ? IF THAT WOULD BE POSSIBLE BITMAIN WOULD COME UP WITH ASIC FOR ETHEREUM LIKE A YEAR AGO
BuriedONE Cryptomining I know scammers exist just look on Alibaba, Amazon or EBay... Sigh, best source is the original source that is bitcoin forums, the guy who posted is more knowledgeable, and I don't think he's faking, though to run a fake CLI script with delays is not hard. True proof is display wallet, pool data, and cli running in real time, that is harder to fake but not impossible (just copy GUI style). But live stream and share public address can be easily verified. Also there was the software native copy of a miner ASIC optimized running on 1050 that was faster than bitmain, yet no one talking about it... Hmm... Its all on bitcoin forums, now if the recent ASIC software can be ripped and ran on FPGA imagine the performance...
Brian M No brian, potential scammers are selling them online and people bombarding me with questions if they need to buy these new cards... Like thes potential scammers: https://fpga.land/products/vcu1525-withmods
Hi i watch your video and i like it, i am very interested to this card.can you tell me how mutch did you get per day with it.does it have software also and how mutch is it ROI for it.thanks man,please give me some more information.
1000% BS. Ethereum IS NOT COMPUTE BOUND AT ALL. Compute efficiency has 0.000000% impact on speed. FPGA is about 10x more efficient than GPU at COMPUTE. FPGA BITCOIN MINERS were made, because BTC requires 0 bytes of RAM, and can be done entirely in logic circuits. This board has DDR4 slots, best case assuming quad channel, performance would be abysmal. Maybe they are getting 1.9 Megahashes? LOL. Sad you have this channel but have absolutely 0 knowledge of mining. I remember you straight up believing rumors of Vega getting 100MH, ignoring the math completely.
There is an EIP for further strengthening Ethereums ASIC resistance. It suggests perhaps a 2x efficiency improvement with a proper Ethereum ASIC vs GPUs. The EIP would further strengthen ASIC resistance. Google "EIP 1057" if you care to look. With Proof of Stake getting closer I would be weary of buying and Ethereum ASIC. GPUs always have the benefit of resale, if you don't buy at high prices... like now. If prices recover before long, and there isn't another rush to buy all the GPUs, it may still be profitable next year, and years to come, so long as there are enough ASIC resistant coins. RX 570 still best ROI (with bios mod.....). I track profit on most cards for common coins/algorithms, message me if you want to see my chart.
Ok I got you so what would you suggest then? Just stick to 1080ti rigs? I think that FPGAs might hold great potential and seem like the next logical step & you shouldn't be so quick to write it off until we have official numbers.
You are stuck on ASICs being magnitudes more efficient at **********COMPUTE**********. COMPUTE bound algorithms benefit greatly from ASICs. ASICs are worthless at MEMORY BOUND OPERATIONS. Ethereum algorithm is 100% memory bound, compute performance is completely irrelevant. If ASICS are so great at every algorithm WHY is Bitmains"ETH ASIC" LESS efficient than GPUs????? It is because a massive amount of memory is used... and memory takes power. Not just amount of memory, but wide, high speed memory busses.
If it were possible to magically make some ETH ASIC like you assume, they would have been made 2 years ago. Ethereum mining is like 85% of GPU mining market, ASICs have been made for like 5 other algorithms in last 6 months, but mysteriously no one makes ETH ASIC because gains are minimal.
wtf are you talking about? asics are magnitudes more efficient(hash to watt ratio) than GPUs on the specific algo they were designed for!?! are you talking about the versatility of gpus? Of course they're more versatile, but the hashrates - watt ratios are waaay too low compared to Asics. Btw there is a Asic for Ethereum mining: the Antminer E3. And the Technicals who commented here is an absolutely serious channel. Your accusations are completely baseless.
Technically, it is impossible, I dont' care if some shill channel "confirmed" it. It is impossible. Ever wonder why no one makes ASICs for Ethereum, even though mining network size is 10-100x that of other coins with ASICs? It is because ASICs are near pointless given the crazy memory bandwidth bottleneck. You can't just ignore the memory part and make an ASIC, or program an FPGA. There is an EIP for further ASIC resistance. Very smart people have laid out all the facts, and there is at most a 2x efficiency improvement over GPUs. Bitmain made an ASIC, it is slightly worse efficiency than GPU.
its fake check dag size 0.9999 gb and eth dag = 2.5gb or 2.6gb this fake its not easy to hit 1000mb hash with 1 card this will never happen cuz u need big big big big gpu processor and no one can do it even bitmain cant
Sorry, I posted my comment to wrong video.
Zetheron Technology - FPGA mining software including for Xilinx VCU1525 for FREE.
Prior to this website he had a thread on BitcoinTalk outlining his FPGA vision.
The company is pretty much impossible to get a hold of. The telephone number directs you to a webpage that directs you to call the exact same phone number. There is a operator, but it seems his job is literally to help you as little as possible, and to just transfer your call to the number that directs you to the webpage that directs you back to the same phone number again.. They don't seem interested in selling anything. It seems like a front company for drug smugglers or something, that wouldn't surprise me in the least.
If you are even considering those mining numbers, then you know next to nothing about Ethash, ASIC or FPGA's .
Please read up before making a video about something you haven't the faintest clue of.
If we consider mining as an FPGA use-case this is where FPGA's would land on different parameters :
Ease of Use : ASIC >> GPU >>>> FPGA
Cost : ASIC > GPU >>>> FPGA
Efficiency : ASIC > FPGA >> GPU
Switching Algo : GPU >>> FPGA >>>>>>>>> ...... ASIC
I’ve been an early adopter since 2011, I even remember the old days of soldering our own power PCI n PCIe risers for LTC during the shortage. Now, THIS HERE frightens me. People think profit n ROI is everything or the best for the community. It’s not. Fast monsters like these, like ASICS, did create n open the door for “get in-cash out fast” people with deep pockets making the price and market share of coins rise in a parabolic line which’s always creates a bubble like 2013, 2017 n other years. The community needs gradually responsable upgrades. This is gonna be a cycle hard to break.
100% fake. I've heard from mining software devs that produce custom software for GPU and some FPGA. FPGA is good on ETH but not amazing, they can get 1.5 to 2x the hashrates at Best but still memory limited. FPGA and asic alike is amazing for compute heavy tasks but the more memory bound it becomes the less viable they are. There is currently a public FPGA dev releasing software on a monthly basis for different coins. So FPGA could be an option still to explore with some risk taking and research.
Fpga buying info https://bitcointalk.org/index.php?topic=3688965.0
Public dev Whitefire990 https://bitcointalk.org/index.php?topic=3459858.0
There is not much results wise you can see at the moment but you can preorder a unit from https://fpga.land and they should provide the bitstreams to you as they are developed from devs. Bitstreams is the bios if you will that tells the FPGA what to do in term of mining an algo. I personally would not advise anyone to take this risk at the moment at least until real profitable results can be seen. Especially with crypto in this bear market you could very likely profit magnitudes higher by buying some high potential coins.
Regardless it is interesting to read about.
DDR4 memory could not possibly sustain such a hashrate. You do understand that bitmain ASICs are identical in performance to gpu rigs (but use less space and energy)? there is no way ddr4 fpgas can do that much hashrate, sorry
There's something wrong with the shown hashrate based on the available memory bandwidth for this Xilinx FPGA card.
We know each hash of ETH requires 8192 bytes of memory access (ie. DAG access per hash is done by a 64 iteration loop with each iteration accessing 128 bytes of the DAG, thus 128 bytes x 64 iterations = 8192 bytes per hash). Now a hashrate of 1.9 GH/s means it would require around 15564.8 gigabytes per second of memory bandwidth. In terms of bits this means that hashrate requires around 124518.4 gigabits per second.
This Xilinx card comes with 4x DDR4 DIMMS, with each DIMM providing a 64 bit data bus, thus the 4x DIMMS provide a total data bus width of 256 bits. The required memory bandwidth of 124518.4 gigabits per second (to get the 1.9 GH/s hashrate) means the DD4 DIMM module must have a speed per data pin of around 486.4 GHz (equivalently 486400 MHz). But there doesn't exist any DDR4 DIMM with data pin speeds which reach the needed 486.4 GHz.
So the DDR4 DIMMs used by this Xilinx card can't even provide the required memory bandwith to give the hashrate shown.
I can't care less with your try hard behaviour. Don't mine with them if you don't want to. You'll soon regret it.You are comparing FPGAs with normal hardware.FPGAs are super specialised and efficient hardware, equal to or better than ASICs.
+Sergei Nikitin - I took another look at the specs of the VCU1525's FPGA chip, the VU9P, and it seems it's processing performance is rated at around 21.3 INT8 TOP/s (https://www.xilinx.com/support/documentation/selection-guides/ultrascale-plus-fpga-product-selection-guide.pdf), this is 8-bit performance so for 32-bit performance (as used by the ethash algorithm) this means around 5.3 TFLOPs.
Now since generation of a single node of the full DAG (from the light DAG) requires a 256-iteration outer loop with a 4-iteration inner loop performing a quad FNV 32-bit operation, then each DAG node generation needs 4K FNV operations. Since the ethash DAG sampling loop itself uses 64 iterations for each hash, then this means 256K FNV operations are needed for each hash. Now an FNV operation is itself composed of a 32-bit multiplication and a 32-bit xor operation, so this means 512K 32-bit operations are needed for each hash. So the 1.9 GH/s hashrate shown in the video means that an on-the-fly generation of the DAG nodes (from the light DAG) would require around 996 TFLOPs, which is much higher than the rated spec for the VU9P. So it doesn't look like the VU9P FPGA with it's 5.3 TFLOPs can brute force the DAG node generation to get the hashrate shown in the video.
There are 3 major pitfalls with reaching consensus around the outcome of an event.
These fully specified, manipulation resistant, and publicly verifiable events form a necessary foundation for sound prediction markets. Without this foundation, prediction markets are subject to confusion, manipulation, and abuse. Though sometimes tricky, many prediction events exist that satisfy all of the criteria listed above. As the world moves forward into the realm of decentralized prediction markets, it will be important to keep in mind the pitfalls associated with many naive prediction events.
Sia , by Nebulous Inc., is a blockchain-based decentralized cloud storage platform.
Capital Markets Blockchains Are Finally Getting Go-Live Dates.
Assembled in New York this week, a handful were even confident enough to give firm timetables for production. For those tired of blue-sky talk, it was refreshing to hear large-scale financial infrastructure projects discussed openly and frankly, in clear terms of where they are and when we can expect to see things going live.
Underscoring the seriousness of the undertaking, ASX recently produced an 87-page progress report. Roll-out is targeted for late 2020 or early 2021.
In the weeds.
The enormity of such a project may not be obvious to those unfamiliar with the creaky plumbing of the capital markets.
At the completion of phase one, DTCC will have nodes set up internally for every firm that it knows will run one, plus some general nodes that will take care of supporting the transactions and processing for the firms that do not wish to support a node of their own.
For this project, DTCC has taken a multi-vendor approach. Ethereum-inspired startup Axoni is providing the technology, with IBM helping to manage the project, and R3 providing best practice guidance on areas like selecting the right data models.
Luxembourg is the largest fund management hub outside of the U.S. The jurisdiction holds many trillions of dollars worth of assets under management.
The KPMG-led project includes banks like BNP Paribas, Credit Agricole and others, as well as over 400 asset managers. The technology used is ethereum-based Quorum, the popular open-source project run by JP Morgan.