This paper describes ID R&D team submission to the text- independent task of the Short-duration Speaker Verification (SdSV) Challenge 2021. The top performed system is a fu- sion of 9 Convolutional Neural Networks based on the ResNet architecture. Experiments’ results of optimal NN architecture search are shown. We also present and investigate the subnet- work approach to solve the auxiliary tasks such as gender or language detection. Verification scores refinement step using quality measurements of a trial pair allowed to further mini- mize the target metrics. A comparative analysis of all systems used in the fusion has been provided on the VoxCeleb-1 test set, SdSV-2021 development and evaluation sets. The final submis- sion achieves 0.69% EER and 0.0319 minDCF on the challenge evaluation set.