Could M.2 CPU mining accelerators also be engineered?
Acorn CAN do this. must we do intel cpus? I am looking for a new thread ripper with your m2 boosters and a 1080ti. or 2 I would get a board with 2x m2 slots. I would then boost the cpu with one m2 from you I would then boost the 2 1080tis with the other m2 from you. Acorn doesn’t particularly care which CPU you use, although of course both worth a bit differently. |
|
|
|
Could M.2 CPU mining accelerators also be engineered?
Acorn CAN do this. |
|
|
|
Under Euro 3 bank regulation, online swift by retail customer (us) can't be faster than 3 days from Europe to US. SO if you pay today in euro Allmine will get them on Monday in their account.
To further speed up, as normal and not business customer, you must walk to your bank.
Well, at least that's what they told me this morning
Makes sense. |
|
|
|
Hi, I want to make order tomorrow cause I will have funds on my acccount but I'm located in EU so it probably won't reach your account till monday. Is it ok if I order tomorrow and also send funds by wire and send screenshot from my bank account or funds have to be in your account by friday evening ?
Thanks
If you’re making wires for for thousands of dollars in FPGAs, do us a favor and pay the tiny fee to expedite. I’m happy to deduct it from your order, rather than manage the Will they/won’t they game for a week. @GPUHoarder, I haven't heard anything about expedited wire transfers from EU to USA. What do you mean exactly? I am commenting here cause I will be doing wire transfer from the EU as well and I know that it schould take from 3 to 5 business days or sometimes even longer. Usually you choose class of service, and you can pay ~100€ to expedite. Varies by bank. Regarding stock - this entire batch is being produced for us, no stock is on hand (essentially operating like a group buy) We’re getting down to the likely end of stock here - if you need expedited service to purchase your FPGAs, email sales@squirrelsresearch.com directly. Jessica will be happy to help quickly |
|
|
|
Well - something is clearly broken. Checking with team. EDIT: It appears the Squirrels Research domain has been hijacked. Please do not email or interact with this domain until we can clear up what happened. EDIT2: The domain is now firmly back under our control and has been migrated to our normal, secure configuration. The MX records were only blanked, and no phishing attempt was made. Please be careful and diligent out there. |
|
|
|
You speak of 30 MH of increase using it of hybrid form, that increase with how many GPU's would be?, You have approximate date of commercialization? A greeting from Spain.
Maybe I missed it, but some clarification would be needed.. Is the bump in hash per gpu or combined bump per gpu group? I have the exact same question re: the ballpark 30 MH figure. I was also disappointed that the GTX 970 and RX 550 aren't on the list. I'm a baller on a budget! Finally, (another repeated question) does the software work in combination with existing mining software to bridge the gap between the GPU and FGPA or does it do all the mining itself? Incidentally could researchers benefit from it to do whatever research it is they do with GPUs? They aren’t excluded, just not listed in the priority for testing survey. |
|
|
|
DIY FPGA Mining rig for any algorithm with fast ROI DIY not any not fast not ~LOL~ Too much hype sometimes lead to lies and misrepresentations. I’m not sure any of those statements are false. Lots of DIY, there’s no algorithm you can’t implement (though some not fast too), and there are definitely those with fast ROI. |
|
|
|
At these current BTC and crypto prices...and the FPGA card prices... looks like mine (and perhaps) others view on all this is gonna get pushed out in the future ...for more than a bit... a couple questions 1) at $3.2k a card (damn, this looked more do'able a month or two ago) is there any 'resale' market on these cards? What I mean is, the folk buying these for industry ...likely need to buy 'new' for the warranty...or they'd lose their job, kinda viewpoint 2) so as to the 'what comes first' do you mine with these FPGA cards on the hope can do 'current' coins or do you do a combo of both current coins that this FPGA card would make sense to mine and/or a specific FPGA coin? 3) What is the best guess of someone getting a working prototype up mining ...whatever coins (above) as a public example? thanks, just trying to catch up...but damn $3.2k a card is 'rough' if there is no real 'resale' value to their industrial used nature (now). (I'm a simple miner (or was looking iffy now) this makes my head hurt.  Searing 1) There is a resale market but it's much smaller. In addition, with the modifications to the device it may be difficult to sell at all. I'm not sure how to say this... There are other things I'd like to do and having the card may be worthwhile to you for other purposes in the future outside of crypto. 2) There's no algo that exists which would be faster on a gpu than a FPGA. The GPU is instruction set based and does have some features in silicon beyond those instruction sets. The problem coin devs have is there's really no way to make an algo that is faster on gpu than fpga. The FPGA are specially suited for this purpose and while there are some limitations with the current generation of cards; in the future, it's not going to be a problem. 3) Ya i've posted me mining on the VCU118 development board. https://imgur.com/a/tmebe6WHi, I want to make order tomorrow cause I will have funds on my acccount but I'm located in EU so it probably won't reach your account till monday. Is it ok if I order tomorrow and also send funds by wire and send screenshot from my bank account or funds have to be in your account by friday evening ? Thanks If you’re making wires for for thousands of dollars in FPGAs, do us a favor and pay the tiny fee to expedite. I’m happy to deduct it from your order, rather than manage the Will they/won’t they game for a week. |
|
|
|
Hey side topic. I did see a couple lines or two in one of the threads about water cooling. Another user had mentioned that they could help design.
Any headway here?
It is going to be available as a “factory” upgrade offered before shipping |
|
|
|
@senseless
According to whitefire, X16R, X16S , X17 "requires 2-3 FPGA cards connected together at 320GBps with 2 x QSFP28 cables, since the algorithms do not fit in a single FPGA effectively."
So would the 16GB memory circumvent that? Seeing if I would need an even amount of FPGA or not or would 1 FPGA would be able to achieve the same thing and swap out the bitstreams within the 16GB memory.?
There are theoretical designs where x16r could be enabled by storing state and swapping bitstreams with RAM, I don’t know that they are advantageous in any way. |
|
|
|
Kaspersky detects Javascript trojan dropper on Zetheron's website. Bad press for them.
Lux Team in full swing right? |
|
|
|
The focus here should be on the opportunity and what reasonably priced, high end FPGAs combined with the current crypto ecosystem can do for this community. Instead we have arguments over false statements and attacks on us personally because we do business differently (and perhaps more transparently?) than others here.
Yes, as long as you are completely transparent about it, especially when large orders of these placed by farms can have and have had the same effect as ASICs on the real GPU mining community. I dont see anyone mentioning anything about the fact that these will most certainly not come anywhere near the profitability you have been advertising once they start shipping. Thats my one pet peeve of ASIC manufactures and projects like these...why are you telling newbs that have no idea what is going on what the profitability is on them on April 30th?? Or Bitmain releasing a new miner without at the very least saying how many units are in the first batch. People dont understand difficulty, and if you want to help out the community, that should be the main disclaimer at the top of the page for any new project that will add a very substantial hash-rate to current networks. I don’t have much to add to Senseless’ statement, other than to reiterate that we have never made any claims regarding profitability of these cards. To that end all I can say is I wouldn’t have done this if I wasn’t convinced that they would be, and that we have significant resources devotated to ensuring they are. |
|
|
|
btw, you obviously have intimate knowledge of xilinx, fpga, and have been on the forum as long as I have.... Where did you say you were from again? I think I can taste a little salt in the air.
How is poking at your business dealings/model him being salty? Few of the higher profile members having this discussion here have been here for long enough to know how most of these pre-orders end up. Its more concerning to me that more of these discussions are not happening. I would not touch this with a 10 foot pole if it was not for that fact that is seems like your using xilinx's supply chain/board partner and design to product these, so it at least takes a good portion of the failure risk out of the equation. Still your collecting a huge sum of money...since they are so un-trusting of the market as you put it, what happens if they get spooked next month and decide not to go through with the deal? Deal is binding, but I think the frustration is coming from the fact that we’ve worked (and are working) extremely hard to make all this happen. We have big financial risk here, as well as tons of time hours and logistics. I personally have barely slept over the past week (anyone on the discord can attest to that) to make sure all of this goes smoothly. That tends to make people a little cranky when someone makes a bunch of unsubstantiated statements that you personally know aren’t true attacking the legitimacy of your work. The focus here should be on the opportunity and what reasonably priced, high end FPGAs combined with the current crypto ecosystem can do for this community. Instead we have arguments over false statements and attacks on us personally because we do business differently (and perhaps more transparently?) than others here. A good number of people on here have gone to great lengths to verify, and vouch for, this. People have contacted Xilinx and verified that we are reputable and are authorized to make this sale. I’m sure if their legal team would clear it any of the Xilinx team watching this thread would confirm that as well. With all said, we stand by what we are doing here and the long term of everyone who chooses to get involved. I am personally really excited about what is happening here, and I think it is going to be game changing. |
|
|
|
For some people that dont want to spend 3500 or just want to get our feet wet before we spent that amount you mention this board The Avnet KU040 which is 9 and change how would we use this would your mining software work with other fgpas I understand certain boards wont be able to mine the same algos as others
Right now the work is going into VCU1525s because we’ve put 5000 of them in market. Once that’s all up and running it shouldn’t be an issue for devs to build some less powerful bitstreams for the KU040, but that’s not necessarily going to be the best board. |
|
|
|
Read through this thread and i think it's great what you guys are trying to achieve, however you have to realize some/most of us will remain skeptical until we at least see these units in action, with actual real world numbers posted for everyone to see, hopefully those results/proof will come sooner than later.
The exact same FPGA cards are available in AWS, so it would be trivial to verify it all for one's self, but the end-to-end software stack to do so has not yet been made available. And expecting people to buy rather expensive hardware without making available the software stack required for public peer review is... bold. This is what i meant, the hardware is there and capable but at what numbers/hashrates, for that we need the software, hopefully it will soon be out for everyone to see cause as it stands, without the software to mine, these will just be expensive paperweights until the software is released, i do know there's other uses for the specific fpga in this thread but the paperweight comment is for this specific use, as in mining. Okay. So remain skeptical and wait for everyone to report back. Is that so hard to do? It is worth pointing out that many of us developing this mining software/bitstreams do not even have VCU1525 cards in our possession yet. That means our numbers are based on the same/nearly same chip on VCU118 dev boards. I'm sure many will be happy to provide more details as they get their development VCU1525s to test with. |
|
|
|
Read through this thread and i think it's great what you guys are trying to achieve, however you have to realize some/most of us will remain skeptical until we at least see these units in action, with actual real world numbers posted for everyone to see, hopefully those results/proof will come sooner than later.
The exact same FPGA cards are available in AWS, so it would be trivial to verify it all for one's self, but the end-to-end software stack to do so has not yet been made available. And expecting people to buy rather expensive hardware without making available the software stack required for public peer review is... bold. 1. The AWS FPGA cards are under several restrictions that make them drastically different in performance. 2. They are the same "chip", but drastically different setup required and a bunch of work no-one has time to do. Again, you're operating as if I'm trying to sell you something. I am not. I have no expectation that you will buy a VCU1525, or anyone else. There is a whole discord full of 1300 people that wanted to be able to get these cards for prices that were reasonable. I helped facilitate that. If you don't want to buy it you don't have to, no one is forcing you to. That option didn't exist before. Too many people on here are selling their wares and trying to scam people into buying their hardware. This isn't my hardware, I'm not "profiting" from selling you it, in fact it is taking a tremendous amount of my time, energy, and sanity to do so. There is a lot of work to do to take what were private miners operated by technical people that know them in very specific hardware setups, and adapt them to be secure as well as easily usable by non-FPGA dev members of this community. That work is underway, but it won't be fully ready for release until these cards are shipping. Similarly if we waited till all the software was 100% ready for the typical miner to plug in and use, we wouldn't be getting cards till 2 months after that. You are welcome to choose to wait and see, and also you may find out there are no cards available at that point. |
|
|
|
|
I think I should be clear that the group buy / sale is for the VCU1525 hardware, below market prices, and the hardware only. It includes access to the shell platform that allows us and others to securely publish bitstreams, including those with the performance numbers on the website, but we are not "selling" a packaged miner or some specific performance promises on some specific algorithms. This is Xilinx's hardware - we are not nor do we want to be Bitmain.
We've simply made the VCU1525 that the first page of this thread references available in a more cost effective group buy, and promised that we will support bitstream and miner development for it alongside the rest of the community.
|
|
|
|
GPU,
Can you explain the difference between your Accelerator and an EVO 970? You guys are evolving the mining world to the next phase.
thanks
The difference is one is a hard drive and the other is an FPGA accelerator that happens to fit into the same slot. |
|
|
|
When do you expect to start shipping ?
I don’t want to overquote - August. |
|
|
|
|
Show Posts
