The NVIDIA Deep Learning Accelerator (NVDLA) is a free and open architecture that aims at promoting a standard way of designing deep neural network (DNN) inference engines. The analogy between open-source software and hardware points to FPGAS as ideal implementation platforms for open hardware accelerators. However, the instantiation flexibility enabled by reconfigurable logic should be correlated to the capacity of cost-effective devices. This paper explores the resource utilization-performance trade-offs spanned by the main precompiled NVDLA accelerator configurations on top of the mainstream Zynq UltraScale+ MPSoC. For the sake of comprehensive end-to-end performance characterization, the inference rate of the software stack is matched to that of the accelerator hardware, thus identifying current bottlenecks and promising optimization directions.
Cross-Layer Hardware/Software Assessment of the Open-Source NVDLA Configurable Deep Learning Accelerator
Bertozzi D.
Ultimo
Supervision
2020
Abstract
The NVIDIA Deep Learning Accelerator (NVDLA) is a free and open architecture that aims at promoting a standard way of designing deep neural network (DNN) inference engines. The analogy between open-source software and hardware points to FPGAS as ideal implementation platforms for open hardware accelerators. However, the instantiation flexibility enabled by reconfigurable logic should be correlated to the capacity of cost-effective devices. This paper explores the resource utilization-performance trade-offs spanned by the main precompiled NVDLA accelerator configurations on top of the mainstream Zynq UltraScale+ MPSoC. For the sake of comprehensive end-to-end performance characterization, the inference rate of the software stack is matched to that of the accelerator hardware, thus identifying current bottlenecks and promising optimization directions.File | Dimensione | Formato | |
---|---|---|---|
Cross-Layer_Hardware_Software_Assessment.pdf
solo gestori archivio
Tipologia:
Full text (versione editoriale)
Licenza:
NON PUBBLICO - Accesso privato/ristretto
Dimensione
246.92 kB
Formato
Adobe PDF
|
246.92 kB | Adobe PDF | Visualizza/Apri Richiedi una copia |
I documenti in SFERA sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.