16 October 2013 in IEEE Transactions on Circuits and Systems I: Regular Papers
A 16-core processor with both message-passing and shared-memory inter-core communication mechanisms is implemented in 65 nm CMOS. Message-passing communication is enabled in a 3 × 6 Mesh packet-switched network-on-chip, and shared-memory communication is supported using the shared memory within each cluster. The processor occupies 9.1 mm 2 and operates fully functional at a clock rate of 750 MHz at 1.2 V and maximum 800 MHz at 1.3 V. Each core dissipates 34 mW under typical conditions at 750 MHz and 1.2 V while executing embedded applications such as an LDPC decoder, a 3780-point FFT module, an H.264 decoder and an LTE channel estimator.
6 April 2015 in IEEE Transactions on Circuits and Systems I: Regular Papers
A domain specific multicore processor for public-key cryptography is proposed in this paper. This processor provides flexible and efficient computation for various forms of RSA and ECC algorithms, fulfilling low-latency or high-throughput requirements of different application scenarios. By using a heterogeneous multicore architecture, the proposed processor enables high speed parallel implementations of kernel arithmetics of public-key algorithms. A long-word-length modular multiplication can be partitioned into parallel tasks executed by the high performance multipliers distributed in multiple cores. Some dedicated communication mechanisms minimize inter-core data transferring latencies of the processor. The proposed processor is implemented under TSMC 65 nm LP CMOS technology. Experimental results show that our design outperforms previous works based on varied platforms in performance, for instance, it can complete a 1024-bit RSA encryption in 0.087 ms at 960 MHz. Moreover, we also study the area reduction techniques for proposed multicore processor from the perspectives of algorithm, architecture, and circuit.
1 March 2012 in Journal of Semiconductors
A register file(RF) with 32×32 capacity and 4-read 2-write(4R2W) ports is presented and analyzed in detail.A new output structure using a MUX and a latch is proposed.It eliminates any dynamic or analog circuit in the read path,and thus it can improve robustness and reduce power at the same time.We also simplify the timing sequence due to the output scheme.The simplified timing circuit not only cuts down the power but also improves the robustness.In addition,less power is achieved when successive read of"0"or"1"is performed.The RF has been fabricated in TSMC 65 nm technology,and the chip test demonstrates that it can operate at 0.8 GHz,consuming 7.2 mW at 1.2 V.
1 January 2015 in IEICE Electronics Express