Superscalar processors improve performance by exploiting instruction-level parallelism (ILP). ILP in a basic block is, however, not sufficient on non-numerical applications for gaining substantial speedup. Instructions across branches are required to be executed in parallel to dramatically improve performance. That is, speculative execution is strongly required. Boosting is a general solution to achieving speculative execution. Boosting labels an instruction to be speculatively executed, and the hardware handles side-effects. This paper describes the efficient implementation of boosting in terms of cost/performance trade-offs. Our policy in implementation is beneficial in code scheduling heuristics, penalties imposed by code duplication to maintain program semantics, and area cost. This paper also describes a branch scheme which minimizes branch penalty. Branch delay causes crucial penalties on the performance of superscalar processors since multiple delay slots exist even in a single delay cycle. Our scheme is the fetching of both sequential and target instructions, and either of them is selected on a branch. No delay cycle can be imposed. This scheme is realized by a combination of static code movement and hardware support. As a result, we reduce branch penalty with small cost. Simulation results show that our ideas are highly effective in improving the performance of a superscalar processor.

A complete single chip multi-format Phase Shift Keying (PSK) demodulator ULSI for Japanese BS digital broadcasting is reported. The carrier recovery system shows the pull-in range up to +/-5 MHz. The clock recovery system cancels the poor group delay characteristic and the orthogonality degradation caused by the analog front end, and improves the BER performance by 0.2 dB. Thus the requirement to the analog front end is relaxed. A digital PLL ensures minimum program clock reference jitter in the output data stream, which simplifies jitter management in the succeeding MPEG2 system decoder. It integrates two 8-bit 60 MHz ADCs, 58 MHz VCO, 1 Mbit SRAM and the 450 K-gate FEC-demodulator core. Implementation of 1 Mbit de-interleaver RAM facilitates the use of a low cost receiver. The 8.8 milion transistor chip occupies the 72 mm2 in a 0.25 µm triple-metal CMOS technology.