Topic: Vertcoin - 1 | ASIC - 0 | Lyra2RE | Decentralised | GPU Mineable | Open Source - page 133. (Read 415451 times)

Lira is the miners friendly algo... More miners... More dump...

What's going on with price :/

Size and frequency increases are mainly what you get for more expensive FPGAs. When I refer to size, I mean the on-chip memory size. DDR or flash memory can be connected to FPGAs, of course, but using them increases the latency. For the fastest operations you would want to use the block RAM on the chip only. Use this table to get an idea.

http://www.xilinx.com/products/silicon-devices/fpga.html

Xilinx Zynq chips have a system on chip and an FPGA all on the same package, but have significantly less block RAM. You can run linux on the SoC and have direct access to an FPGA.
http://www.xilinx.com/products/silicon-devices/soc/zynq-7000/silicon-devices.html

As for cost - they are expensive, as you probably expected. The low-mid range Kintex 7 FPGA development board costs $1700 from Xilinx. Development boards have tons of features and hardware that would not be necessary for this application, so cost may come down, but spinning a circuit board for an FPGA is not an easy task.

Actually, it takes dev time to port an algo like X11. Lord knows SIMD or some other obscure hashes probably don't have much for public FPGA code. But I agree, once you get the dev work done, it's about how cheaply you can build them - the dev work is a one-time cost.

I have no more than 700kH/s, rather: 688 - 689 kH/s

statements like this contribute 0 to anything please read my previous post on reasons behind the change, you don't need to be negative coin has reached nice plateau already and many have nice chance to get started in crypto with it in the right way, with few GPU cards and some fun, please think twice before making post like this it as it really does not do anything literarily

it's working..
300khs..

It`s a new beginning.

Changing the algorithm is nice, but the course is rapidly rushed to the bottom, it's the end coin?

Yes.  As stated in my previous post, my miner (SP fork fork of ccminer - http://github.com/sp-hash/ccminer) is displaying higher hash rate than what's reported by the pool.  I'm getting 620-625 kh/s per the pool.  On my EVGAs the gain from overclocking is quite small.

My miners must be submitting shares like crazy because the pool I mine at is reporting 5,000-9700 kh/s all day per 6 card 750ti rig

hi if anyone tell me this works on ATI 5850 ?

i cannot get this works on my ATI 5850 !

this statement I like (from the white paper by VERT) : https://vertcoin.org/downloads/Vertcoin_Lyra2RE_Paper_11292014.pdf

"Unlike Scrypt, Lyra2 (the principal part of the chained algorithm) allows us to change memory usage and time cost independently, giving us more leverage against ASICs."

so VERT acknowledges theoretical and practical possibility and reality of fpga for lyra2 but with caveat that algorithm can adjust and render these systems ineffective (or needed to be reprogrammed on the chip level)

"At this time we have decided not to implement an “N factor” schedule as it is nearly impossible to predict the future. However, Lyra2RE will give us the flexibility to make changes whenever that becomes necessary"

N factor would adjustments could combat fpga/asic with necessity of fpga reprogramming (and invalidate complete asic investment and asic is cheaper to make and handle so risk to consumers of them is rather big)

also this:

"By contrast, Lyra2RE aims to be ASIC-resistant at heart, allowing for less disruption to miners in the future due to our ability to change algorithm parameters rather than change algorithm all together. It will also free up development time to focus on new features without having to worry about constantly implementing new algorithms every time there is an ASIC threat"

so at least and unlike many crypto VERT is very open about the situation

for fpga, a new micro kernel code has to be created and chips flashed for any parameter adjustments, average user will have to ship device to the manufacturer and wait for new boards to be received back, then VERT community or VERT dev itself sees fpga activity again based on the profitability scale and perform adjustment again at some point it becomes impractical to get fpga chips reprogrammed every time

we are looking better and better from the long term investment strategy and I wish this would have been done sooner, but coin has been re-launched and people can rebuild their portfolio from the bottom instead of from the top.

No problem. Neoscrypt still has a little more in it, though. And thanks!

Never mind, hashing at 1,2Mh now 

Anyone has a good setting for Sapphire r9 290 ?
Can't get past ~515 Kh/s 

Okay, good. I was a little worried, there. 7990 is pretty nice, even my 290X can only hit 560 so far, so that card still has a leg up on it.

i believe all  algo can be done on FPGA, it is just about Time and Cost.
i have some computer knowledge and programming ability，checked some new POW algo.
when X11 is released，i think it is bullshit，i haven't touch any X'series coins,no matter what the are.
X11 or even up to X26,all have public FPGA code,it is even easier to build FPGA miner than scrypt.
if you check whattomine site,profit already show them FPGAs are running.
it is a easy way to check if it is mint by FPGAs or ASICs: the calculated profit.
you don't need to find where are the FPGAs or ASICs,just easy calculate.
go to http://www.whattomine.com/ and find out 7 day Profitability,
if it is less than 400%,the algo is rushed  by FPGAs,
if it  is less than 200%,means the coin is dying ,or the algo is rushed by ASICs,
another way,if you have 280x card,calculate the one day profit ,not calculate the electric fee.
if it is less than 0.2$,the coin is dying or the algo is fucked by FPGAs or ASICs.

His miner is probably displaying the wrong numbers, no idea why tho. I'm willing to bet whatever pool he mines on reports closer to 600kh/s per card. With major overoclocking djm said he was able to get 660-670kh/s

How are you getting 720-725 kh/s out of your 750tis ? I have 36 evga 750tis (using your kopiemtu) and the most I can get is 625 kh/s out of each card