senseless
|
|
May 07, 2018, 04:38:30 AM |
|
Regarding the question about bitstream compatibility, the answer is no, a bitstream built for VCU1525 will not work on the XUPP3R. The clock & USB UART are on different pins, so I must 'build' two versions of the bitstream, one for each card. Although the change in the code is tiny, it still takes the tools 5+ hours to re-run place & route, with multiple runs needed to get a success.
Can you inplement a double set of interfaces to support both boards by the single firmware? BUFGMUX_CTRL can be used to select the clock signal, if clocking assignment is different. It is probably possible to build one firmware for both board. I get the general idea but I still have a hard time seeing how to do that with the USB-UART pins. As far as 'resale' of these FPGA cards, there is also a tremendous future for AI mining. In fact, if tens of thousands of people were mining crypto with these boards, and an AI mining server became available, they are so powerful that you might end up with a 'Skynet' type superintelligence, or the Kurzweil 'singularity' event. My point is that if, for some reason, mining crypto became unprofitable, someone could re-sale access to the FPGA hardware for AI, weather prediction or any other compute-heavy tasks. Suppose you have 2 different sets of pins. RXD_BITTWARE and TXD_BITTWARE is one set, RXD_XILINX and TXD_XILINX is another set. The uart module inside FPGA has RXD input and TXD output also. Then make connections as follows: TXD_BITTWARE = TXD TXD_XILINX = TXD RXD = RXD_BITTWARE & RXD_XILINX The merging function varies between "or, xor, and" and it is depending on internal or external pulling unconnected pins up or down. I'm curious why he's not just using PCI-E for comms. You need the board in that PCI-E slot for the +75W. Get rid of that mess of USB cables and use an AXI4-Lite interface. The PCI-E block would be identical on both boards.
|
|
|
|
2112
Legendary
Offline
Activity: 2128
Merit: 1068
|
|
May 07, 2018, 04:42:42 AM |
|
do people who know c language and assembly language, who have experience in programming micro-controllers like zilog, atmel and pic ....can easily adapt to using this? like doing the stuff that you are doing? create firmware for this FPGAs
It doesn't help much and in my opinion it is actually detrimental. Those devices you mentioned have the similar conceptual limitation: they encourage thinking of solving tasks in terms of sequences of steps. Very much like many early "home computers" programmable with BASIC. the lack of developers in this arena and the lesser utility/other uses of FPGA (in general population) makes me think twice in investing more in this kind of stuff...unless i knew a guy in person that i can work with, not just some random guy in the internet..
The thing is that a random guy can conceivably be much better at FPGA programming, even without previous exposure to computers. I believe it starts much earlier and is conditioned with the type of toys one played as a kid. Here's my theory: If as a kid somebody played (1) with mechanical toys that allow to build working things out of building blocks or (2) with electrical toys that allow building working circuits out of components connected with pluggable wires then as an adult such person will have natural aptitude for (a) FPGA programming or (b) parallel programming or (c) electronic engineering To such a person the FPGA development kit is very much like their dream toy from the childhood upgraded to a true professional tool with nearly unlimited quantity of building blocks and connecting wires. So I disagree with your premise that the aptitude for FPGA programming is rare, it just isn't discovered and developed in many people. In the past I was frequently surprised that some people with advanced computer science degrees from reputable schools have extremely hard time understanding that some device can concurrently work doing many things simultaneously. No need to explicitly state the order in which things need to be done. You may now know it, but FPGAs don't need to be programmed using some language like Verilog or VHDL. They can be designed using schematic capture of assembling blocks and connecting them with wires. And this comes naturally to some grade school kids with no computer science exposure whatsoever. So the TL;DR of the above is: it really depends on what kind of toys you played with when you were a kid.
|
|
|
|
netmonk
|
|
May 07, 2018, 04:51:53 AM |
|
... the lack of developers in this arena and the lesser utility/other uses of FPGA (in general population) makes me think twice in investing more in this kind of stuff...unless i knew a guy in person that i can work with, not just some random guy in the internet..
As far as im not leaving on american continent, i called my local Bittware reseller several time last week. He is well introduced into this FPGA field, selling and developping on those kind of board and technology since more than 10 years. Therefore, he has knowledge and skills, also has in his professionnal network some independant developpers with more skills and knowledge in this area. He already inform them of the existence of this thread and waiting their feedback. Local reseller are valuable consellor.
|
|
|
|
flankiee
Newbie
Offline
Activity: 2
Merit: 0
|
|
May 07, 2018, 06:32:34 AM |
|
Regarding the question about bitstream compatibility, the answer is no, a bitstream built for VCU1525 will not work on the XUPP3R. The clock & USB UART are on different pins, so I must 'build' two versions of the bitstream, one for each card. Although the change in the code is tiny, it still takes the tools 5+ hours to re-run place & route, with multiple runs needed to get a success.
Can you inplement a double set of interfaces to support both boards by the single firmware? BUFGMUX_CTRL can be used to select the clock signal, if clocking assignment is different. It is probably possible to build one firmware for both board. I get the general idea but I still have a hard time seeing how to do that with the USB-UART pins. As far as 'resale' of these FPGA cards, there is also a tremendous future for AI mining. In fact, if tens of thousands of people were mining crypto with these boards, and an AI mining server became available, they are so powerful that you might end up with a 'Skynet' type superintelligence, or the Kurzweil 'singularity' event. My point is that if, for some reason, mining crypto became unprofitable, someone could re-sale access to the FPGA hardware for AI, weather prediction or any other compute-heavy tasks. I also think so, building a Skynet, making huge energy consumption a meaningful thing.
|
|
|
|
greyday
Member
Offline
Activity: 434
Merit: 52
|
|
May 07, 2018, 06:52:57 AM |
|
Regarding the question about bitstream compatibility, the answer is no, a bitstream built for VCU1525 will not work on the XUPP3R. The clock & USB UART are on different pins, so I must 'build' two versions of the bitstream, one for each card. Although the change in the code is tiny, it still takes the tools 5+ hours to re-run place & route, with multiple runs needed to get a success.
Can you inplement a double set of interfaces to support both boards by the single firmware? BUFGMUX_CTRL can be used to select the clock signal, if clocking assignment is different. It is probably possible to build one firmware for both board. I get the general idea but I still have a hard time seeing how to do that with the USB-UART pins. As far as 'resale' of these FPGA cards, there is also a tremendous future for AI mining. In fact, if tens of thousands of people were mining crypto with these boards, and an AI mining server became available, they are so powerful that you might end up with a 'Skynet' type superintelligence, or the Kurzweil 'singularity' event. My point is that if, for some reason, mining crypto became unprofitable, someone could re-sale access to the FPGA hardware for AI, weather prediction or any other compute-heavy tasks. I also think so, building a Skynet, making huge energy consumption a meaningful thing. Part of the point of FPGAs and AI specific chips is energy efficiency, but yes.
|
|
|
|
arielbit
Legendary
Offline
Activity: 3416
Merit: 1059
|
|
May 07, 2018, 06:54:09 AM |
|
do people who know c language and assembly language, who have experience in programming micro-controllers like zilog, atmel and pic ....can easily adapt to using this? like doing the stuff that you are doing? create firmware for this FPGAs
It doesn't help much and in my opinion it is actually detrimental. Those devices you mentioned have the similar conceptual limitation: they encourage thinking of solving tasks in terms of sequences of steps. Very much like many early "home computers" programmable with BASIC. the lack of developers in this arena and the lesser utility/other uses of FPGA (in general population) makes me think twice in investing more in this kind of stuff...unless i knew a guy in person that i can work with, not just some random guy in the internet..
The thing is that a random guy can conceivably be much better at FPGA programming, even without previous exposure to computers. I believe it starts much earlier and is conditioned with the type of toys one played as a kid. Here's my theory: If as a kid somebody played (1) with mechanical toys that allow to build working things out of building blocks or (2) with electrical toys that allow building working circuits out of components connected with pluggable wires then as an adult such person will have natural aptitude for (a) FPGA programming or (b) parallel programming or (c) electronic engineering To such a person the FPGA development kit is very much like their dream toy from the childhood upgraded to a true professional tool with nearly unlimited quantity of building blocks and connecting wires. So I disagree with your premise that the aptitude for FPGA programming is rare, it just isn't discovered and developed in many people. In the past I was frequently surprised that some people with advanced computer science degrees from reputable schools have extremely hard time understanding that some device can concurrently work doing many things simultaneously. No need to explicitly state the order in which things need to be done. You may now know it, but FPGAs don't need to be programmed using some language like Verilog or VHDL. They can be designed using schematic capture of assembling blocks and connecting them with wires. And this comes naturally to some grade school kids with no computer science exposure whatsoever. So the TL;DR of the above is: it really depends on what kind of toys you played with when you were a kid. thanks for the input....the way you explain things to me makes me feel that I can do it already haha. the person that I have in mind is a geek/nerd..kind of jack of all trades...when I ask him about programming last year when we met, like can you still program in the programming languages we encountered in the past? his answer: yes, and all of them are the same even the ones that he didn't use yet. things that person knows..making websites, programming in different languages, electronic engineering, micro controllers, computer networks.....and baking cakes! my heart bled when i learned he is baking cakes! but they are making good money with his sister and if i pulled him away from that venture.. i don't know, he is a mainstay in that business..maybe i'll pay him a visit sometime..i'll wrap my head around this FPGA thing first.. I remembered when I'm still in high school, in our programming class, where there is a guy who knows how to program...and i know a bit in programming that i can change variables and colors and i'm the one who knows windows command line that i'm pirating copies to my class mates so at the end of the class everybody have their own unique programs (in their floppy disk) haha..
|
|
|
|
stash2coin
Jr. Member
Offline
Activity: 108
Merit: 1
|
|
May 07, 2018, 06:54:42 AM |
|
VCU1525 card is the Development Kit! This is not mass produced. And it can not ramp up production.
I had a conference call with Xilinx before announcing this project and they agreed to transition the VCU1525 into a production product for a slightly higher price; the production version will be released in June/July, in much higher quantities. Interesting but i dont see the logic here Xillinx is selling the chip used in these cards on its own for utterly ridiculous price, now they will provide much cheaper source of this chip, if your not interested in crypto mining but you have a product to sell using this chip, buy this card desolder the chip reball it and use it in your own product. I dont see this happening in the long run. Its a great marketing strategy but wont bring big revenue for Xillinx in a long run.
|
|
|
|
oskarwu
Newbie
Offline
Activity: 30
Merit: 0
|
|
May 07, 2018, 06:56:44 AM |
|
Regarding the question about bitstream compatibility, the answer is no, a bitstream built for VCU1525 will not work on the XUPP3R. The clock & USB UART are on different pins, so I must 'build' two versions of the bitstream, one for each card. Although the change in the code is tiny, it still takes the tools 5+ hours to re-run place & route, with multiple runs needed to get a success.
Can you inplement a double set of interfaces to support both boards by the single firmware? BUFGMUX_CTRL can be used to select the clock signal, if clocking assignment is different. It is probably possible to build one firmware for both board. I get the general idea but I still have a hard time seeing how to do that with the USB-UART pins. As far as 'resale' of these FPGA cards, there is also a tremendous future for AI mining. In fact, if tens of thousands of people were mining crypto with these boards, and an AI mining server became available, they are so powerful that you might end up with a 'Skynet' type superintelligence, or the Kurzweil 'singularity' event. My point is that if, for some reason, mining crypto became unprofitable, someone could re-sale access to the FPGA hardware for AI, weather prediction or any other compute-heavy tasks. The same idea as mine , but this kind of business opportunity will belong to mining FARM owners with enough resource to transform their hardware computing service from crypto mining resouce to AI mining resource ...
|
|
|
|
oskarwu
Newbie
Offline
Activity: 30
Merit: 0
|
|
May 07, 2018, 06:58:44 AM |
|
Regarding the question about bitstream compatibility, the answer is no, a bitstream built for VCU1525 will not work on the XUPP3R. The clock & USB UART are on different pins, so I must 'build' two versions of the bitstream, one for each card. Although the change in the code is tiny, it still takes the tools 5+ hours to re-run place & route, with multiple runs needed to get a success.
Can you inplement a double set of interfaces to support both boards by the single firmware? BUFGMUX_CTRL can be used to select the clock signal, if clocking assignment is different. It is probably possible to build one firmware for both board. I get the general idea but I still have a hard time seeing how to do that with the USB-UART pins. As far as 'resale' of these FPGA cards, there is also a tremendous future for AI mining. In fact, if tens of thousands of people were mining crypto with these boards, and an AI mining server became available, they are so powerful that you might end up with a 'Skynet' type superintelligence, or the Kurzweil 'singularity' event. My point is that if, for some reason, mining crypto became unprofitable, someone could re-sale access to the FPGA hardware for AI, weather prediction or any other compute-heavy tasks. BTW , do you plan to support high-end Intel/Altera device/board in the coming future ? Thanks ~
|
|
|
|
m5
Newbie
Offline
Activity: 82
Merit: 0
|
|
May 07, 2018, 08:01:48 AM |
|
I got the same lead time in EU. Can you recommend any rack server for these? I guess they won't fit in normal 2U servers like HPE Proliant as most spec only one full width full height card as the board is double height if I understand it correctly.
|
|
|
|
Ginzink
|
|
May 07, 2018, 08:07:49 AM |
|
If i were to buy just one of theese cards, could i not just put it in my computer then as it fits in the PCI slot? But communicates through usb and only use PCI for power was that it?
|
|
|
|
EggdraSyl
Member
Offline
Activity: 140
Merit: 30
Bismuth core dev & EggPool.net Operator.
|
|
May 07, 2018, 08:19:34 AM |
|
If i were to buy just one of theese cards, could i not just put it in my computer then as it fits in the PCI slot? But communicates through usb and only use PCI for power was that it?
Sure. Risers are only needed if it does not fit in the regular pci-e slots.
|
|
|
|
senseless
|
|
May 07, 2018, 08:28:46 AM Last edit: May 07, 2018, 08:41:46 AM by senseless |
|
I got the same lead time in EU. Can you recommend any rack server for these? I guess they won't fit in normal 2U servers like HPE Proliant as most spec only one full width full height card as the board is double height if I understand it correctly. I'm using the 3u8g-c612. It's about $3K. If i were to buy just one of theese cards, could i not just put it in my computer then as it fits in the PCI slot? But communicates through usb and only use PCI for power was that it?
It requires the PCI-E power, a full 75W. But you can provide that through a powered riser. I would expect him to implement a PCI-E interface at some point though. VCU1525 card is the Development Kit! This is not mass produced. And it can not ramp up production.
I had a conference call with Xilinx before announcing this project and they agreed to transition the VCU1525 into a production product for a slightly higher price; the production version will be released in June/July, in much higher quantities. Interesting but i dont see the logic here Xillinx is selling the chip used in these cards on its own for utterly ridiculous price, now they will provide much cheaper source of this chip, if your not interested in crypto mining but you have a product to sell using this chip, buy this card desolder the chip reball it and use it in your own product. I dont see this happening in the long run. Its a great marketing strategy but wont bring big revenue for Xillinx in a long run. Some (Most) of your assumptions here are false. Can't really say more as it's all covered under NDA. You're also looking at a 50% +/- recovery rate removing and reballing a FPGA.
|
|
|
|
eugenetritante
Newbie
Offline
Activity: 9
Merit: 0
|
|
May 07, 2018, 08:40:49 AM |
|
wow 100$ per VC I will save and buy too so I can start mining too. thanks for the information.
|
|
|
|
melpheos
Jr. Member
Offline
Activity: 557
Merit: 5
|
|
May 07, 2018, 09:27:24 AM |
|
@OP Whitefire990 : you write that you will distribute the bitstream file but what about the miner itself ?
|
|
|
|
Reggie0
Member
Offline
Activity: 107
Merit: 13
|
|
May 07, 2018, 09:47:57 AM |
|
Regarding the question about bitstream compatibility, the answer is no, a bitstream built for VCU1525 will not work on the XUPP3R. The clock & USB UART are on different pins, so I must 'build' two versions of the bitstream, one for each card. Although the change in the code is tiny, it still takes the tools 5+ hours to re-run place & route, with multiple runs needed to get a success.
Can you inplement a double set of interfaces to support both boards by the single firmware? BUFGMUX_CTRL can be used to select the clock signal, if clocking assignment is different. It is probably possible to build one firmware for both board. I get the general idea but I still have a hard time seeing how to do that with the USB-UART pins. If the alternative serial pins are mutually unused on the other board then: - you can transmit on both TX pins. - enable the pull up for RX pins, because in idle state the serial lines are in high state. Then you can make a common RX signal: RX_COMMON=1'b1^RX_A^RX_B.
|
|
|
|
heavyarms1912
|
|
May 07, 2018, 03:10:08 PM |
|
Things to ponder.
FPGAs are expensive. Dev time is significantly higher than other code. Compile times are even higher (days or even weeks) If you don't get modified bitstream files to keep up with any algo related changes, you'll be stuck with an expensive piece of HW.
With all that said this is a good initiative.
|
|
|
|
BittWareFPGATech
Newbie
Offline
Activity: 10
Merit: 6
|
|
May 07, 2018, 03:10:46 PM |
|
Looks like Bittware reduced their standard warranty for the "crypto" models. I believe they typically have 1 yr mfg warranty on their products. Kind of disappointing the warranty was reduced that much for a $6k product.
Yes, a 90-day warranty raises a LOT of concerns. Has Bittware done stress testing and has some reason to believe the failure rate ramps up after that timeframe? I could understand a 90-day warranty on support issues with a 1-year hardware warranty. I was in the process of arranging my finances to purchase a few cards, but if the manufacturer has that little confidence in their product, they're sending me a message I can't ignore. Even the Chinese manufacturers (with their notorious quality) give 180-day warranties. Just for some background, we have been shipping this board for 1.5 years now, and have had almost no field failures. It has gone through full validation, including thermals. We manufacture in the US using a Tier 1 contract manufacturer. Our standard warranty is 1 year, we have some products with 3 year warranties, and some customers who pay for extended warranties. We also generally provide unlimited support, as our normal customer is one who buys 10s or 100s (or even 1000s) of boards. We have been building high reliability boards for decades, with one of our boards being used as part of the traction control system on the TGV (and all Alstom train and subway cars) for well over 10 years. When this mining opportunity came up, we looked for ways to provide a deep discount as we know this community is very cost sensitive. We also have concerns with the way that these boards will be used for mining, mostly from an ESD point of view. Our normal customers install these boards in servers, and are in properly controlled environments with ESD protection taken very seriously. So we lowered the warranty to reduce cost, and for fear of the boards being abused. Perhaps we went too far lowering it to 30 days. We also limited support, however, our normal support includes helping people develop their FPGA code on our board. In this case you are using code already developed. Note that we do have onboard monitoring for temperature, voltage, and current, with thresholds set to shut down the board if it appears to be entering a situation that would cause damage. We provide a utility (console and GUI versions) that you can use to talk to this onboard controller via the USB and see what is going on with all the sensors. Unlike GPUs and CPUs, FPGAs have no inherent thermal throttling, unless the FPGA developer builds that into their logic, which is very rarely done. FPGAs also burn more power as they get hotter, so from a power efficiency point of view, it is better to keep them cool. As for cost we presented, while I know it seems steep for this community, this is deeply discounted from our standard price, taking into account the reduced warranty and support. From what I can tell, Xilinx has a 90 day warranty on their devkits, which is what the VCU1525 is. I'm not sure how long they will sustain that low cost, we know of smaller board manufacturers that pay more than that for just the VU9P chip itself. Sorry for the typo "... went too far lowering it to 30 days". We've lowered it to 90 days, not 30.
|
|
|
|
heavyarms1912
|
|
May 07, 2018, 03:15:24 PM |
|
My rig is already paid off so i might be tempted to purchase two in order to mine at least raven (which apparently requires two, i'm not sure exactly why - lack of memory to contain all the instruction ? )
Raven is using 16 different algos. The dev time complexity of implementing 16 different kernels into an FPGA will be significant. @whitefire990 are you coding on pure VHDL/Verilog or are you using HLS/HLX?
|
|
|
|
s1gs3gv
Legendary
Offline
Activity: 1316
Merit: 1014
ex uno plures
|
|
May 07, 2018, 03:39:06 PM |
|
do people who know c language and assembly language, who have experience in programming micro-controllers like zilog, atmel and pic ....can easily adapt to using this? like doing the stuff that you are doing? create firmware for this FPGAs
easily ? no.
|
|
|
|
|