Bitcoin Forum
November 10, 2024, 12:43:43 PM *
News: Latest Bitcoin Core release: 28.0 [Torrent]
 
   Home   Help Search Login Register More  
Pages: [1]
  Print  
Author Topic: Hash vs. Real Hash  (Read 893 times)
sspano01 (OP)
Newbie
*
Offline Offline

Activity: 2
Merit: 0


View Profile
June 09, 2014, 01:27:57 AM
 #1

Hi Folks,

I have a question on an FPGA miner design I am working on. Its running with CGMiner now at about 1GH/s, well, in fact it has 8 cores in it with each core running at 125MHz, so I know that each core is "hashing" at 125MH/s which is 1GH/s. However, sometimes it doesn't get a solution for a requested job from cgminer. That means that my average hash rate on cgminer drops down to around 750MH/s. Of course, sometimes it gets solutions very fast and shows my hash rate up near 4GH/s.

What is the story with the "no solution" issue and is there a better way to report hash rate in cgminer? Ideally, I would like to see always a 1GH/s worst case rate since that is what the hardware is running at.

Thanks
shorena
Copper Member
Legendary
*
Offline Offline

Activity: 1498
Merit: 1540


No I dont escrow anymore.


View Profile
June 09, 2014, 09:33:23 AM
 #2

-snip-
Its running with CGMiner now at about 1GH/s, well, in fact it has 8 cores in it with each core running at 125MHz, so I know that each core is "hashing" at 125MH/s
-snip-

If each core runs at 125 MHz and can caluclate 1 hash per cycle your hash rate is 125MH/s. It seems however that the ICs you are using are not makeing a hash within a single calculation cycle.

Im not sure where your ICs find a shortcut in sha256 which is the only explanaiting I have to them something "solving" this faster.

Im not really here, its just your imagination.
sspano01 (OP)
Newbie
*
Offline Offline

Activity: 2
Merit: 0


View Profile
June 10, 2014, 12:37:03 AM
 #3

Hi

The logic design has two SHA cores in it for each "hash" engine and runs fully pipelined. I can see in the simulation that the first hash takes 2 clock cycles, and then a new hash/nonce attempt is generated each clock cycle after that. It seems like that if no solution is found, that the code in cgminer drops the request and then its not counted as a hash attempt?

I just a single serial interface and no queue running. Maybe the issue is job-to-job latency? I need to implement the queue to make sure the chip has continuous work
ajareselde
Legendary
*
Offline Offline

Activity: 1722
Merit: 1000

Satoshi is rolling in his grave. #bitcoin


View Profile
June 11, 2014, 10:47:35 AM
 #4

Hi Folks,

I have a question on an FPGA miner design I am working on. Its running with CGMiner now at about 1GH/s, well, in fact it has 8 cores in it with each core running at 125MHz, so I know that each core is "hashing" at 125MH/s which is 1GH/s. However, sometimes it doesn't get a solution for a requested job from cgminer. That means that my average hash rate on cgminer drops down to around 750MH/s. Of course, sometimes it gets solutions very fast and shows my hash rate up near 4GH/s.

What is the story with the "no solution" issue and is there a better way to report hash rate in cgminer? Ideally, I would like to see always a 1GH/s worst case rate since that is what the hardware is running at.

Thanks


You are talking about local hashrate, which is in your case 1 gh, but sometimes your luck is higher , sometimes lower, and that is what is making your miner appear on your pool sometimes as 750mh and sometimes above 1gh.
Pages: [1]
  Print  
 
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.19 | SMF © 2006-2009, Simple Machines Valid XHTML 1.0! Valid CSS!