Table 3 Overview of secret key rate and GPU (NVIDIA GeForce GTX 1080) throughput at the maximum reconciliation distance with Rcode = 0.02 multi-edge codes and Nprivacy = 1012 bits

From: Quasi-cyclic multi-edge LDPC codes for long-distance quantum cryptography

Reconciliation dimension

Maximum reconciliation efficiency

LDPC code

Maximum distance (km)

Finite secret key rate \(K_{{\mathrm{finite}}}^\prime\) at max distance with frep = 1 MHz (bit/s)

Upper bound on key rate \(K_{{\mathrm{lim}}}^\prime\) at max distance with frep = 1 MHz (Kbit/s)

GPU raw throughput \(K_{{\mathrm{GPU}}}^{{\mathrm{raw}}}\) (Mbit/s)

GPU info. throughput \(K_{{\mathrm{GPU}}}^\prime\) (Kbit/s)

\(K_{{\mathrm{GPU}}}^\prime\) speedup over \(K_{{\mathrm{lim}}}^\prime\) (\(K_{{\mathrm{GPU}}}^\prime\)/\(K_{{\mathrm{lim}}}^\prime\))

d = 1

β = 0.960

Random

122.40

0.021

5.152

0.613

0.111

0.020×

  

QC, q = 21

122.40

0.042

5.152

1.887

0.686

0.130×

  

QC, q = 50

122.40

0.023

5.128

1.966

1.426

0.280×

d = 2

β = 0.970

Random

127.20

0.005

4.129

0.612

0.222

0.050×

  

QC, q = 21

127.20

0.020

4.129

1.856

2.70

0.650×

  

QC, q = 50

127.20

0.009

4.129

1.983

0.360

0.090×

d = 4

β = 0.970

Random

127.20

0.010

4.129

0.604

0.439

0.110×

  

QC, q = 21

127.20

0.030

4.129

1.818

3.938

0.950×

  

QC, q = 50

127.20

0.069

4.129

1.855

2.692

0.650×

d = 8

β = 0.990

Random

142.40

0.037

2.049

0.604

1.409

0.690×

  

QC, q = 21

142.40

0.066

2.049

1.724

7.160

3.50×

  

QC, q = 50

142.40

0.064

2.049

1.808

4.207

2.05×

  1. At the maximum distance points of \(\ell\) = 122 km, \(\ell\) = 127 km, and \(\ell\) = 142 km, the excess noise \(\epsilon\) is, respectively, set to 0.032, 0.037, and 0.052, and Alice’s modulation variance VA is, respectively, set to 14.0, 17.5, and 34.1