Yes, every of the 8 DES bytes contains 7 bit (and 1 parity bit).
The 7 bits in the last byte are always 0 (for all 8 ECM keys I have seen).
So it looks like they use only a 2^49 key space for the ECM keys.
May be someone has a recorded Tandberg stream (CAID 1010 and EMM tab ID 82 *AND* 83) from an non olympic event. So we can see if the key space is only 2^49 for olympic events or generally.
Found on the web
"DES works by encrypting groups of 64 message bits, which is the same as 16 hexadecimal numbers. To do the encryption, DES uses "keys" where are also apparently 16 hexadecimal numbers long, or apparently 64 bits long. However, every 8th key bit is ignored in the DES algorithm, so that the effective key size is 56 bits. But, in any case, 64 bits (16 hexadecimal digits) is the round number upon which DES is organized."
I did a quick test on FPGA with a open core taking 56bit key in and 64bit data in / out to use a sample I did this:
Code:
[COLOR="RoyalBlue"]See Text file Attached[/COLOR]
I loaded the above 56bit key into the DES core on my FPGA and it does output the correct result, so far fine. :thum:
Sure I will need to add in many cores to run parallel on the FPGA from what I see on the resource use I think it might be possible to have around 100 cores on the FPGA I use.
The fmax I do see are around 200 Mhz with a result on every clock, so this will give around 20 Ghash/s so 2^56 will be around 1000 hours for a full search.
If 2^49 it will be about 8 hours on FPGA
But we have to be sure the last bit are not used, recall that in RAS you will type in a 7 digit decimal number, maby this will fille out the unused bit's?
No idea how fast it can be done on CUDA card, could nice to test if on the new GTX1080 card
I will not be able to do more test on FPGA until midd next week