| Autor: |
Achar, Sumana, D., Jayadevappa |
| Předmět: |
|
| Zdroj: |
IAENG International Journal of Computer Science; Dec2023, Vol. 50 Issue 4, p1410-1419, 10p |
| Abstrakt: |
The chip application has become the trending element in all digital applications because of its easy and flexible use. However, chip-based communication’s chief demerits are power consumption and delay. The huge data broadcasting often raises these two problems in Network-onChip (NoC) architecture. So, the present research work has aimed to create a novel Strawberry-based Recurrent neural (SbRN) framework for minimizing the NoC buffer length and reducing the latency and power consumption. Here, the fitness of Strawberry was utilized to sense the large data size; once the large data size was identified, the compression process was started to compress the data. Moreover, the NoC architecture was designed with an optimized buffer with strawberry fitness. Once the compressed data was present in the network medium, it allowed broadcasting to the other end. If the data is not compressed or not in the minimum compression range, the compression process was again to optimize the data buffer. Finally, the proposed architecture was validated with other conventional schemes and has gained the best outcome by reducing 5% of device utilization and 0.5% of power consumption than other methods. [ABSTRACT FROM AUTHOR] |
| Databáze: |
Supplemental Index |
| Externí odkaz: |
|
