A 995Mpixels/s 0.2nJ/pixel fractional motion estimation architecture in HEVC for Ultra-HD

This paper presents a fractional motion estimation (FME) design in high efficiency video coding (HEVC) for ultrahigh definition video (Ultra-HD). To reduce complexity and achieve high throughput, the design is co-optimized in algorithm and hardware architecture. Bilinear quarter pixel approximation, together with a 5T12S search pattern is proposed to reduce the complexity of the interpolation and search process. Furthermore, we introduce an exhaustive size-hadamard transform (ES-HAD), to improve coding quality, and determine the best transform size rather than using complex transform coding. Besides, a data reuse method of ES-HAD is applied to reduce the hardware overhead. This design is implemented in 65nm CMOS chip and verified by FPGA based evaluation system. It achieves 995Mpixels/s for 7680×4320 30fps encoding, at least 4.7 times faster than previous designs. Its power dissipation is 198.6mW at 188MHz, with 0.2nJ/pixel power efficiency. Despite high complexity in HEVC, the chip achieves 56% improvement on power efficiency than previous works in H.264.