Topic: ZTEX USB-FPGA Modules 1.15x and 1.15y: 215 and 860 MH/s FPGA Boards - page 49. (Read 182443 times)

ah, thanks for the explanation

It's a link

Virtex FPGA's are to expensive for bitcoin mining. They contain a lot of features which are not required for bitcoin mining, e.g. multiplicators and high speed serial I/O's.

The fact that GSM can be hacked is known since several years. AFAIR it was published by group of scientists from Israel.

Publishing ready-to-use software would just reduce the amount of effort required to abuse this technology.

Whats this?

http://semiaccurate.com/2011/10/25/xilinx-sets-new-fpga-record/

You are aware that keeping things secret only HELP governments.  If the government wants to crack GSM it will crack GSM.  They won't be buying your product to do so.  They will be employing their own cryptographers.

Hiding things in secrecy only help governments.  Information in the open levels the playing field.  If GSM is cracked IN SECRET then nobody knows and the govt can spy without public knowledge.  If GSM is cracked IN PUBLIC then the public is at least aware of the danger and hopefully public pressure pushes for superior algorithm which can't be cracked.  This may in fact defeat a govt program which has already cracked GSM by bringing public attention to the vulnerability.

An example would be WEP.  WEP was horribly flawed.  It would have provided no resistance to any govt program.  However the wifi industry claimed it was secure.  Hell it is the Wired Equivelency Protocol.  It is as secure as a wireline.   Had it not been broken publicly and so spectacularly to make it the laughing stock of the industry and public WPA may never have been developed.

If you want to protect citizens from governments then make information free (free as in free speech not free beer).

GSM hacking and computation of rainbow tables are other cryptographic applications. But I will not implement ready-to-use solutions for this due to moral reasons (could help governments to tyrannize their people)


At http://www.ztex.de you can also find boards for these applications.

A few word about the possibility of reselling:

This FPGA Board become inefficient (higher power costs than returns) when bitcoin mining is dominated by future generation FPGA's (even upcoming GPU's will not be as efficient as current FPGA's). If that ever happens it will take a very long time due to the high investment costs.

In other words, it will take a very long time until these FPGA board become inefficient (if that ever happens). Then the boards are to outdated in order to sell them.

(sorry for the OT tangent)

5830s overclocked (and carefully clock managed, but not overvoltaged, 324MH/s) bought in bulk, 6x to a board, slowest cpu, network booting, a lot of careful tweaking. All other parts cut to a minimum. GPUs screwed to a wooden frame many linear feet long spaced as wide as possible. Ribbons ordered from hong-kong and shipped on the slow boat. I'm also fairly power efficient for a gpu setup— you can gain a fair bit by careful psu selection and running at 240v.  (and good power quality is important for stability at high clock rates)

Total cost per node are 109*6 (gpus) + 210 (mb) + 4*6 (ribbons) + 60 (cpu/ram)  + 100 (power) + 8 (marginal price of a ethernet switch port) = $1056 plus some modest shipping costs and a negligible amount of wood and screws.  yields 1944 MH/s.  Er. That a bit better than I said, though its a bit closer once you throw in some for shipping.


Its hard to say. There are a lot of gpu miners... and once the cards are a couple generations old I expect an influx of used cards (especially a lot of lemon used cards that have been driven hard and not well cooled as many miners are guilty of) will undermine the used market price. It's hard to guess by how much.  These ATI GPUs are only so vastly superior to the alternatives for a few apps.. a lot of gamers would rather have nvidia.

If you want to build a DES cracker however, then that FPGA will be a pretty smoking solution.  I expect the FPGA vendors could probably improve their sales by offering tools for other applications, just to increase customer confidence that the FPGAs are useful for things other than mining.  DES reversing and WPA cracking are obvious things that come to mind. ... lack of fast IO limits SDR applications, lack of memory and fast IO limits imaging applications, these are best for crypto.

25% less on your build?  That's impressive.  Care to share your secrets?

Very true:  Do you own math/homework. Whether it's hype, anecdotes, or my calculations, it might not be right for you.

I'm ambivalent about the resale value of the FPGAs. If BTC was no longer viable or you just wanted to get out of the game, I think you would take a much bigger hit on the residual value of the FPGAs than a GPU.

Certainly they are being an option worth serious consideration.  Multi-FPGA and better hash rate at a lower cost is what could be very interesting.  Keen to hear about the 2nd gen....

People should run their own numbers in any case— your $/MH on the GPU farm is higher than my actuals by about 25% but people build with different degrees of efficiency.

You're certainly making the point that FPGAs are now in the realm of competitive for new deployments, even for buildouts with fairly short term planning horizons. Ztex's modules are also attractive because they've clearly been designed in a manner which would facilitate re-purposing them for things other than mining, which is a reason they should retain some value even if future SASIC (or even ASIC) devices are more efficient for mining.

OTOH, the density of one S6-LX150 per board is kinda poor, as I assume these are losing a fair amount to commons and PCB costs— (the bare FPGAs are about $190 each IIRC).  Sort of a tossup, the single chip boards are more reusable but it's still kinda pricy even with the quantity discounts.

You're right.     Sorry about that.  I pulled the wrong numbers so the output was too high for both FPGA and GPU.  I have updated the calculations, which now shows that FPGA and GPUs are roughly equivalent at 8 cents/kWh, everything else being equal.

50% would be high even for pool hopping.  Real world (especially today w/ pools putting up countermeasures and less pools being hoppable at all) hopping "only" nets you 25%-35%.

  You're right, its .65 BTC per day on the FPGA setup and 1.83 BTC per day on the GPU setup in the example at current difficulty. Maybe his calc uses a pool hopping formula to get the numbers?

Your BTC/month is almost 50% higher than the real earnings at these difficulty levels. They'll fall at the next change, but not enough to earn 30/month at 950 Mh/s. It seems to me that it won't fall this much unless the price falls even more.

  I don't know about that. ;p  At 50GHs that amounts to 1164 Watts less in electricty costs on this Ztex board. If electricty costs are at all an issue, this will add up pretty quick. ~28KW per day. Or $1,226 per year @ .12 KWH.  Yea, maybe not a big deal unless you're approaching it from a business perspective.

  Awesome board, Ztex!  Thanks for making it available to the community.

  

Polymer capacitors are electrolytic capacitors.


... and to THT bypass capacitors


Yes,  the USB-FPGA Module 1.15x has to be firmware and bitstream compatible to USB-FPGA Module 1.15d.

Nice job!

Glad to see you added the northbridge sink mounting holes.  But you might want to consider putting in two sets of holes at different distances like I did -- there isn't an official standard for how far apart they're supposed to be, and the distance can vary by as much as 6mm -- enough that any single pair of holes will rule out a significant fraction of the heatsinks out there due to being either too close or too far.

Why the electrolytic capacitor?  I see you've got at least one polymer (i.e. does-not-explode-and-leak) cap on there... curious why you didn't use that for both of them.  I've switched to all polymer+ceramic.

Also curious: is there a reason why you aren't using the FGG484 package?  It's slightly cheaper than all the others (not sure which one yours is, but it's not the FGG484).

Stacking these FPGA boards is not feasible due to several reasons

• Stackable pin headers are expensive and either have to be soldered by hand or have to be pressed in (both is expensive)
• At high currents you get problems with the contact resistance
• The space between the board is usually to small for a proper cooling
• Last but not least: At least all clock signals would have to be terminated (in order to avoid that reflected signals are misinterpreted as clocks). But termination is impossible since such a stack is everything but impedance controlled.

Thanks for your calculations. But don't forget, for larger amounts (group order) you get higher volume discounts. See the initial message for bulk prices.

Edited: Original calculations overstated BTC/day for both FPGA and GPU.  Revised calculations show that FPGA makes sense at a much lower power cost.

It seems that the viability of this product rests completely on electricity pricing.  If you compare FPGA to GPU with the same amount of money invested it appears (using round numbers) that somewhere in the range of 12 8 cents per kWh is where either solution is roughly equal in profit.  If your power costs more than that, you're better off with an FPGA.  If you can get cheaper power, go for GPUs.

As an example, let's say you invest $2000 into mining.  With that, let's assume you could either buy 5x ZTEX FPGA modules or two GPU mining rigs with 8x GPUs.

FPGA
Investment: $2000
FPGA's: 5
Total Mh/s: 950
Monthly power cost at $0.08 for 45W: ~$3
BTC Payout: 0.6508 BTC/Day or ~20 btc/month
USD Payout at $2.70: $54
Monthly net revenue from $2000 investment (BTC minus power): $51

GPU
Investment: $2000
4-GPU rigs: 2 (each having 4 x Radeon 5850 GPUs producing 333 Mh/s each)
Total Mh/s: 2664
Monthly power cost at $0.08 1700W: ~$99
BTC Payout: 1.825 BTC/Day or ~55.5 btc/month
USD Payout at $2.70: $150
Monthly net revenue from $2000 investment: $51

Of course this comes with the standard caveats that BTC prices and difficulty will change and that you might be able to overclock GPUs and get even more Mh/s, etc. etc..  Also, if AMD can deliver on huge cuts in power consumption with the 7xxx series GPUs, the power indifference line will increase.

But, if you're considering investing in mining equipment and are looking at this FPGA or GPUs, your electricity rates may have already made the decision for you.