Page 108 - Kaleidoscope Academic Conference Proceedings 2024
P. 108
2024 ITU Kaleidoscope Academic Conference
Table 2 – Performance of CPU, GPU and Alveo with
increasing depth (from 1 to 21) for the given qubit size
Qubits(n) Performance Crossover depth of
Sequence two systems
1 GPU ≈ CPU, Alveo GPU ≈ CPU
2 GPU ≈ CPU, Alveo GPU ≈ CPU
3 GPU ≈ CPU, Alveo GPU ≈ CPU
4 GPU ≈ CPU, Alveo GPU ≈ CPU
5 GPU ≈ CPU, Alveo GPU ≈ CPU
6 GPU ≈ CPU, Alveo GPU ≈ CPU
7 GPU ≈ CPU, Alveo GPU ≈ CPU
8 Alveo, CPU, GPU No crossover
observed
9 Alveo, GPU, CPU Crossover observed
between CPU and
GPU after depth 7
10 GPU, Alveo, CPU Crossover observed
between GPU and
Alveo after depth 5
Figure 12 – Runtime comparison of different qubit sizes with
depths (from 1 to 21)
Figure 13 – GPU ≈ CPU (it is taking almost the same time )
for 7 qubits
Figure 11 – Performance of CPU, GPU, and Alveo with Figure 14 – Crossover observed between CPU and GPU after
increasing qubit at depth 1 depth 7 and for 9 qubits
– 64 –