Reconfigurable Computing Applied to AI

Artificial Intelligence (AI) is a field that will offer many opportunities for emerging markets and services and will revolutionize almost every segment of the society. With Artificial Intelligence techniques, it is possible to provide a high precision tool to address classification and prediction problems such as, for instance, speech synthesis or pattern recognition. Artificial Intelligence covers a wide range of activities: medical, multimedia, finance, or advertising, among others. The results are applied to the most varied systems, such as clusters, mobile phones, or even smart sensors. A preliminary market analysis revealed that the Artificial Intelligence market is expected to be worth USD 16.06 Billion by 2022 [1]. However, AI requires intensive processing and high-performance hardware. Thus, for companies such as Intel, IBM, Amazon or Google, energy consumption becomes a serious issue. Their servers and databases are now considering the adoption at the architectural level of low power alternatives, as fast as powerful central computers, multi-core or GPUs (Graphics Processing Unit). Indeed, AI implementations have a high computational cost because most techniques use complex algorithms. This feature makes it difficult to apply AI algorithms to many emerging fields such as Mining Massive Dataset (MMD), 5G Communications or Bioinformatics.

However, novel approaches such as Reconfigurable Computing (RC) can improve the performance of AI algorithms. With the advent of customizable hardware (using field-programmable gate arrays - FPGAs), algorithms can be parallelized and optimized at the gate level to speed up operations, reaching up to 1000x speedup according to what is presented in the literature [2, 3, 4]. Due to the increased use of FPGA-based reconfigurable computing, in addition to consumer, automotive or military electronics, the FPGA market is expected to grow to $ 12.1 billion by 2024. Thus, the combination of these two approaches can make possible the realization of the speed up, low-power and area-efficient AI hardware in the coming years.

[1] “Artificial Intelligence (Chipsets) Market worth 16.06 Billion USD by 2022”, tractica.com. http://www.marketsandmarkets.com/PressReleases/artificial-intelligence.asp (October 04, 2017).

[2] de Souza, A.C.D.; Fernandes, M.A.C. Parallel Fixed Point Implementation of a Radial Basis Function Network in an FPGA. Sensors 2014, 14, 18223-18243.

[3] Kara, K., Alistarh, D., Alonso, G., Mutlu, O., & Zhang, C. (2017, April). FPGA-accelerated Dense Linear Machine Learning: A Precision-Convergence Trade-off. In Field-Programmable Custom Computing Machines (FCCM), 2017 IEEE 25th Annual International Symposium on (pp. 160-167).

[4] Shaikh, F., Kalwar, I.H., Memon, T.D. and Sheikh, S., 2017, June. Design and analysis of linear phase FIR filter in FPGA using PSO algorithm. In Embedded Computing (MECO), 2017 6th Mediterranean Conference on (pp. 1-4).