Pages:
Author

Topic: ALLMINE INC - FPGA Cryptominer - page 54. (Read 51510 times)

hero member
Activity: 1118
Merit: 541
May 09, 2018, 05:07:07 PM
#16

Sounds like you should sell the gpus and pick up better general purpose miners. The FPGA can even have opencl code compiled for it.


And that's why I am here;)

What will be time period to deliver after ordering and how much upfront?


Need to have some additional talks with Xilinx and check BOM before i say anything concrete; but i'd estimate 30->90days. It really just depends on the state of the market and if things are or are not available. A single item of the BOM not being available could delay things for weeks. That's just me being completely honest with you instead of saying "YA YOU'LL HAVE IT NEXT WEEK! GIVE ME MONEY!".

member
Activity: 531
Merit: 29
May 09, 2018, 05:03:35 PM
#15

Sounds like you should sell the gpus and pick up better general purpose miners. The FPGA can even have opencl code compiled for it.


And that's why I am here;)

What will be time period to deliver after ordering and how much upfront?
hero member
Activity: 1118
Merit: 541
May 09, 2018, 04:59:12 PM
#14
This new FPGA trend sounds interesting and earnings do sound good to everybody but don't you think making so many devices available could cripple coins?
Every coin developer is striving to be ASIC resistant but when using FPGAs they won't be able to counteract them since you can change the algo as fast as they fork.

I mean difficulty would increase but you still would have a lot of coins flooding the market and driving them on a bear trend to very low values thus ruining your ROI too.

Lyra2z alone can handle 1-2K devices before diminishing returns start to kick in.
Lyra2rev2 performance will be about the same as the asic that was announced with double the power usage. Lyra2rev2 could support 50K devices before diminishing returns start.
Cryptonight v7 could handle another 50K-100K devices before diminishing returns kick in.

Not sure why whitefire hasn't targeted any of those yet, or cryptonight/lightv7. The lyra2z performance will make you want to kick your 1080 TIs to the trash bin (over a 50x increase in performance for 1/2 the power).

This all without greatly affecting GPU profit margins (maybe a 20-40% decrease). There's room for high multiple 100k of units, possibly millions.




Have'nt looked at Lyra2z, but the Phi network would be destroyed for GPU's even if 5000 FPGAs came online.


Sounds like you should sell the gpus and pick up better general purpose miners. The FPGA can even have opencl code compiled for it.

member
Activity: 531
Merit: 29
May 09, 2018, 04:56:53 PM
#13
This new FPGA trend sounds interesting and earnings do sound good to everybody but don't you think making so many devices available could cripple coins?
Every coin developer is striving to be ASIC resistant but when using FPGAs they won't be able to counteract them since you can change the algo as fast as they fork.

I mean difficulty would increase but you still would have a lot of coins flooding the market and driving them on a bear trend to very low values thus ruining your ROI too.

Lyra2z alone can handle 1-2K devices before diminishing returns start to kick in.
Lyra2rev2 performance will be about the same as the asic that was announced with double the power usage. Lyra2rev2 could support 50K devices before diminishing returns start.
Cryptonight v7 could handle another 50K-100K devices before diminishing returns kick in.

Not sure why whitefire hasn't targeted any of those yet, or cryptonight/lightv7. The lyra2z performance will make you want to kick your 1080 TIs to the trash bin (over a 50x increase in performance for 1/2 the power).

This all without greatly affecting GPU profit margins (maybe a 20-40% decrease). There's room for high multiple 100k of units, possibly millions.




Have'nt looked at Lyra2z, but the Phi network would be destroyed for GPU's even if 5000 FPGAs came online.
hero member
Activity: 1118
Merit: 541
May 09, 2018, 04:54:53 PM
#12
You already have some software for this? Why would we need whitefire's software?

Because I'm not releasing my software. My goal is to not let bittware bilk the community. Whoever wants to create firmwares, provide them, and charge a fee (like whitefire990) -- or not -- is more than free to do so.





member
Activity: 531
Merit: 29
May 09, 2018, 04:54:07 PM
#11
You already have some software for this? Why would we need whitefire's software?
hero member
Activity: 1118
Merit: 541
May 09, 2018, 04:45:47 PM
#10
This new FPGA trend sounds interesting and earnings do sound good to everybody but don't you think making so many devices available could cripple coins?
Every coin developer is striving to be ASIC resistant but when using FPGAs they won't be able to counteract them since you can change the algo as fast as they fork.

I mean difficulty would increase but you still would have a lot of coins flooding the market and driving them on a bear trend to very low values thus ruining your ROI too.

Lyra2z alone can handle 1-2K devices before diminishing returns start to kick in.
Lyra2rev2 performance will be about the same as the asic that was announced with double the power usage. Lyra2rev2 could support 50K devices before diminishing returns start.
Cryptonight v7 could handle another 50K-100K devices before diminishing returns kick in.

Not sure why whitefire hasn't targeted any of those yet, or cryptonight/lightv7. The lyra2z performance will make you want to kick your 1080 TIs to the trash bin (over a 50x increase in performance for 1/2 the power).

This all without greatly affecting GPU profit margins (maybe a 20-40% decrease). There's room for high multiple 100k of units, possibly millions.


sr. member
Activity: 616
Merit: 279
May 09, 2018, 04:35:00 PM
#9
This new FPGA trend sounds interesting and earnings do sound good to everybody but don't you think making so many devices available could cripple coins?
Every coin developer is striving to be ASIC resistant but when using FPGAs they won't be able to counteract them since you can change the algo as fast as they fork.

I mean difficulty would increase but you still would have a lot of coins flooding the market and driving them on a bear trend to very low values thus ruining your ROI too.
hero member
Activity: 1118
Merit: 541
May 09, 2018, 04:23:39 PM
#8
On the other thread, whitefire990 mentioned "The intro price (at Avnet) on the VCU1525 is $3995 USD". That was the quote I got when emailing Jason directly per unit. How come your quote is higher even after negotiations?

That price is intro, digikey is selling for $5200 and avnet will be right around the same. Let's assume they (digikey and avnet) got a better deal than me from Xilinx because of their 30 year billion $ business relationship. And, as I mentioned, at 10K MOQ the board would be $3,000-$3,500 shipped. That's a little more than half the price of what the units will cost in June/July from digikey/avnet.


sr. member
Activity: 544
Merit: 250
May 09, 2018, 04:21:16 PM
#7
On the other thread, whitefire990 mentioned "The intro price (at Avnet) on the VCU1525 is $3995 USD". That was the quote I got when emailing Jason directly per unit. How come your quote is higher even after negotiations?
hero member
Activity: 1118
Merit: 541
May 09, 2018, 03:46:54 PM
#6
Did you read this:

Sorry, I've been meaning to make a post related to this.  We have been working with the OP to determine the optimal configuration and board for this application, as we have a few to choose from.  The XUPP3R may not be the optimal one, we have other boards, including one with larger core power supplies and even the VU13P instead of the VU9P. The OP is working to determine the combination of FPGA size, clocks, logic, memory and power use that maximizes ROI.  We do not want to sell you boards that may not be the best fit, that is in no one's best interest.

So you have not missed out, you perhaps have saved yourself from buying a non-optimal board (we actually have not yet taken any orders or shipped any boards to miners so no one is in that camp).  You can still contact us and we will get back to you as soon as we have enough details to let you place an order.

So VU13P might be better? I think you are going too fast. I am interested in 2 but things need to settle down.

Edit: And OP(whitefire99) is still working specs.

I've already been at this for over a year. I've been mining on these boards since June 2017 and working on them since Jan 2017. The XCVU9P is where you want to be at. The XCVU13P isn't produce in the same quantity and it will not be as cost effective. The fact they don't know which device to get and are just figuring this out now shows how late to the game they are.

Bittware is also not going to give you a good deal. For each board they sell they're going to gross about 2/3.

The VCU1525 schematics and bom are available for anyone to produce. I have already sat down with Xilinx (in Shenzhen) and negotiated chip pricing. If there is enough interest, I could facilitate a chip purchase, put the chips on boards and deliver them at a regular price of $4,000-$4,500. To hit a $4,000-$4,500 and stay at $4,000-$4,500 there would need to be at least 1,000 boards sold. If order volume went up to or over 10,000 units pricing could be dropped to $3,000-$3,500 per unit.

You would be able to use the device with the firmwares produced by whitefire990

Any interest?

Payment methods would be Crypto (obviously) and Credit Card / Paypal (identity verification would be required and shipping to your registered / verified address)


So this would require his software? what if I want to develop my own? would I have a paper weight?

If you want to develop your own firmwares you'd need a copy of vivado. A copy of vivado would not be included with this board. But you could purchase a single unit from avnet or digikey and obtain a copy of device-locked Vivado with that (the software would only allow bitstream generation against the VCU1525, if you wanted to be able to generate bitstreams for any device you need to buy vivado outright at a cost of around $4K).
newbie
Activity: 37
Merit: 0
May 09, 2018, 03:41:17 PM
#5
The VCU1525 schematics and bom are available for anyone to produce. I have already sat down with Xilinx (in Shenzhen) and negotiated chip pricing. If there is enough interest, I could facilitate a chip purchase, put the chips on boards and deliver them at a regular price of $4,000-$4,500. To hit a $4,000-$4,500 and stay at $4,000-$4,500 there would need to be at least 1,000 boards sold. If order volume went up to or over 10,000 units pricing could be dropped to $3,000-$3,500 per unit.

You would be able to use the device with the firmwares produced by whitefire990

Any interest?

Payment methods would be Crypto (obviously) and Credit Card / Paypal (identity verification would be required and shipping to your registered / verified address)


So this would require his software? what if I want to develop my own? would I have a paper weight?
member
Activity: 531
Merit: 29
May 09, 2018, 03:41:08 PM
#4
Did you read this:

Sorry, I've been meaning to make a post related to this.  We have been working with the OP to determine the optimal configuration and board for this application, as we have a few to choose from.  The XUPP3R may not be the optimal one, we have other boards, including one with larger core power supplies and even the VU13P instead of the VU9P. The OP is working to determine the combination of FPGA size, clocks, logic, memory and power use that maximizes ROI.  We do not want to sell you boards that may not be the best fit, that is in no one's best interest.

So you have not missed out, you perhaps have saved yourself from buying a non-optimal board (we actually have not yet taken any orders or shipped any boards to miners so no one is in that camp).  You can still contact us and we will get back to you as soon as we have enough details to let you place an order.

So VU13P might be better? I think you are going too fast. I am interested in 2 but things need to settle down.

Edit: And OP(whitefire99) is still working specs.
hero member
Activity: 1118
Merit: 541
May 09, 2018, 03:12:46 PM
#3

The VCU1525 schematics and bom are available for anyone to produce. I have already sat down with Xilinx (in Shenzhen) and negotiated chip pricing. If there is enough interest, I could facilitate a chip purchase, put the chips on boards and deliver them at a regular price of $4,000-$4,500. To hit a $4,000-$4,500 and stay at $4,000-$4,500 there would need to be at least 1,000 boards sold. If order volume went up to or over 10,000 units pricing could be dropped to $3,000-$3,500 per unit.

Any interest?



I'm interested in getting between 2-4 units, Which payment method is accepted? paypal, creditcard, or only crypto?

Im ready to dump most of my gpus and step up my game and switch over to FPGA cards



Crypto definitely, I'm not sure about paypal and credit card. Stripe would probably close my account with all of the fraud i'd receive. My paypal account is a bit more established (18 years and MM transacted). I doubt it would have any issues from paypal with the amount of fraud. But, I'm sure you can imagine what a PITA it would be to deal with that. I would probably do paypal / credit card manually on a case by case basis after asking for proof of identity documents, if at all.


full member
Activity: 846
Merit: 115
May 09, 2018, 03:12:02 PM
#2

The VCU1525 schematics and bom are available for anyone to produce. I have already sat down with Xilinx (in Shenzhen) and negotiated chip pricing. If there is enough interest, I could facilitate a chip purchase, put the chips on boards and deliver them at a regular price of $4,000-$4,500. To hit a $4,000-$4,500 and stay at $4,000-$4,500 there would need to be at least 1,000 boards sold. If order volume went up to or over 10,000 units pricing could be dropped to $3,000-$3,500 per unit.

Any interest?



I'm interested in getting between 2-4 units, Which payment method is accepted? paypal, creditcard, or only crypto?

Im ready to dump most of my gpus and step up my game and switch over to FPGA cards

hero member
Activity: 1118
Merit: 541
May 09, 2018, 03:01:16 PM
#1
News / Updates:
05/29/18 - We are going to delay pre-sales at this time. We have calls scheduled with both Intel and Xilinx for this week. News to follow.
06/01/18 - Store opened at https://fpga.land/ for user registration. We are still not taking orders at this time but you can register and receive updates when hashrates, pricing and order information becomes available.
06/04/18 - Any developers in the community wishing to develop bitstreams please send me a private message. If you can't make a crypto core in RTL, this doesn't apply to you.
06/04/18 - We will NOT be able to accept credit cards for batch 1. Credit card processors are unwilling to allow us to process credit cards to sell you FPGA that any other site who doesn't market specifically to the crypto community would be able to sell and accept credit cards for. If we sold them as "FPGA Accelerators", no one would care. I should add, this is one of the reasons crypto was created in the first place.
06/05/18 - Pricing information has been released on our website https://fpga.land/
06/07/18 - Orders will open at 4PM EST on 06/07/2018
06/07/18 - Paypal disabled due to fraud and charge backs
06/08/18 - Sales have closed at this time. We may reopen sales if devices free up from cancelled or unpaid orders.
06/08/18 - Additional stock as opened up. Payments via crypto and bank wire only. Paypal has suspended our account.
06/12/18 - Paypal requested additional information to re-open our account for pre-orders. We won't be able to provide them the information by the time the sales close. They have agreed though to continue working with us in the future. We'll be allowed to sell via paypal for future batches.
06/14/18 - Several RTL source codes have been released by ourselves as well as Sprocket to bootstrap community RTL development. You can find our github repo here and sprocket's github repo here. Between the 2 of us we're covering around 90% of the algos in used today.
06/16/18 - Sales have closed. Register on the website to receive notice of openings for future sales.
06/21/18 - For those of you who purchased your fpga through us but wish to host with mineority will be given the opportunity to do so. We'll send out an email and update with details when we have them.
09/13/18 - Our pre-production BCU-1525 units shipped from Xilinx today. Volume production to follow; Current expectations are to start shipping volume units in early Oct. Additional shipping updates expected to follow once we have more information on exact shipping dates. We're hopeful we'll be able to ship all orders before the end of Oct.
10/20/18 - We've been busy shipping BCU-1525 units. To date a minimum of 800 units have been shipped out to customers of all sizes who did not opt to wait for additional upgrades (waterblocks, passive block upgrades, etc). We still have more units in the warehouse to prep for shipping and more units on the way scheduled to arrive next week.
11/2/18 - New build of minerator, bmc firmware and 60Mh/s lyra2z coming soon! We're expecting some community dev releases this week as well.


Allmine Developer / Miner website: https://all-mine.co/
BCU-1525 Instructions: https://miner.all-mine.co/setup-instructions
Discord Link: https://discord.gg/QMavGwv


FAQ

What are the hardware requirements for FPGA mining?

Currently there are no known specific hardware requirements for FPGA mining beyond basic requirements for GPUs. That is, a basic CPU, minimal ram, minimal disk, and x1 PCI-E 3.0 connectivity. There may be other hardware needs in the future for specific bitstreams. Some developers may opt to use the PCI-E bandwidth to communicate between cards or offload some processing from the FPGA to the CPU. In these instances, those developers will make known these hardware requirements. You may opt to upgrade or change your infrastructure then to support those bitstreams. The FPGA present a new way of doing things and there are countless possibilities for how various resources could be used in mining. Because this is an open but new development environment it will take some time for things to mature and the most efficient way of doing things to become evident.

What operating systems are supported for FPGA operations?

At launch we are planning only Linux support. When you're mining, even in windows, all that's really necessary is for you to be able to edit a file and run a command line interface application. We do not see using linux as a blocking point for the operation for miners. Detailed setup and installation instructions will be provided. We're also planning integration into some mining-specific linux distributions which will allow easy web interface management. However, even without the web management it should not be difficult for someone who's operated a GPU miner or ASIC miner to also operate a FPGA miner.

Can the FPGA multi-mine?

Yes! The FPGA can use resources however is needed for the developer to achieve their goal. It would be possible to mine several different algorithms at the same time on these devices. There are situations where some resources are more optimally used for one thing than another. Using the FPGA it's possible to mix and match algorithms and designs to achieve the best mix of resource utilization for profitability. An example of this, CN7 has a better hashrate using blockram. It would be possible to use CN7 on blockram, Lyra2z on the ultraram and using the rest of the logic space for a small tribus or other logic algo mix. Going forward with future generations of devices as they become larger I'd expect even greater variation of mining. It's becoming clear that the most profitable bitstreams will be multi-algo bitstreams.

What is "the shell" and why do I need it?

"The Shell" fundamentally is a wrapper around any mining bitstream that will run on your FPGA. The wrapper provides a common set of communication and programming functionality to allow anyone's bitstream or design to operate on your FPGA without re-programming the FPGA using the USB cable. We're able to provide this because we've burned an encryption key onto every FPGA shipped. This encryption key is what enables the creation of our secure shell environment. Some examples of functionality that the shell provides are:

    Hardware Management (Temperature, Power (Voltage/Amperage), Clock sources, PCI-E, Programming / Reprogramming)
    Bitstream Management (Bitstream distribution, developer fee collection)
    Mining Software (Bitstream management, Algo management, common communications platform for all bitstreams)

For Developers:

It allows developers an easier time to get started without needing to build their own software, communications, or worry about secure their development fee. They can focus solely on producing the best possible bitstream designs for maximum profitability. Due to the secure nature of the bitstreams it will enable developers who previously may have not released their designs, to release their designs, as they won't fear people attempting to cut out the developer fee and diminish their work.

For Miners:

Miners will get a wider selection of bitstreams, an easier and faster way to switch between them (over pci-e instead of USB), and easier management of their miners with our online config builder and management system. As of now, if you want to switch from one bitstream to another it would take possibly hours of reprogramming depending on the number of FPGA you have. Using our shell it will take seconds to switch bitstreams and starting mining a new one. This creates a better environment for profit switching and increased gains.

In addition, the shell and software provide user side control of clocking and overclocking. The user can define their own parameters for what they find to be desirable for voltage, clock, temperature, etc. Or, if desired, tell the software to overclock to the maximum of the safety limits.





Hashrates:
Lyra2z (8,8,8) - 40Mh/s 225W ~ Still under development
Cryptonight variant 1 / monero version 7 - 14Kh/s 150W



Notes: Only posting hashrate numbers we're 100% sure about. Cores still under development are expected to increase in hashrate.
Pages:
Jump to: