Thread: GMP-ECM vs mprime for F12? View Single Post 2019-03-12, 19:57   #10
mathwiz

Mar 2019

127 Posts Quote:
 Originally Posted by ATH Here is the sigma from the F14 factor: https://mersenneforum.org/showpost.p...9&postcount=18 So you can test with: ecm.exe -sigma 8585974330888598 110e6 110e8 < f14.txt where f14.txt is just "2^16384+1"
Yup.

Code:
Input number is 2^16384+1 (4933 digits)
Found number: 1*2^16384 + 1
Using special division for factor of 2^16384+1
Using B1=110000000, B2=11000000000, polynomial x^1, sigma=0:8585974330888598
dF=16384, k=4, d=158340, d2=11, i0=684
Expected number of curves to find a factor of n digits:
35	40	45	50	55	60	65	70	75	80
73	308	1481	7942	46946	302925	2115419	1.6e+07	1.3e+08	1.1e+09
Using gwnum_ecmStage1(1, 2, 16384, 1, 110000000, 1)
Step 1 took 3102146ms
Estimated memory usage: 1.64GB
Initializing tables of differences for F took 13ms
Computing roots of F took 1634ms
Building F from its roots took 3879ms
Computing 1/F took 1515ms
Initializing table of differences for G took 139ms
Computing roots of G took 1282ms
Building G from its roots took 3860ms
Computing roots of G took 1282ms
Building G from its roots took 3860ms
Computing G * H took 844ms
Reducing  G * H mod F took 1269ms
Computing roots of G took 1283ms
Building G from its roots took 3853ms
Computing G * H took 843ms
Reducing  G * H mod F took 1268ms
Computing roots of G took 1282ms
Building G from its roots took 3879ms
Computing G * H took 846ms
Reducing  G * H mod F took 1270ms
Computing polyeval(F,G) took 8930ms
Computing product of all F(g_i) took 622ms
Step 2 took 43778ms
********** Factor found in step 2: 116928085873074369829035993834596371340386703423373313
Found prime factor of 54 digits: 116928085873074369829035993834596371340386703423373313
Composite cofactor (2^16384+1)/116928085873074369829035993834596371340386703423373313 has 4880 digits
Peak memory usage: 994MB  