James Warnock, Brian Curran, John Badar, Gregory Fredeman, Donald Plass, Yuen Chan, Sean Carey, Gerard Salem, Friedrich Schroeder, Frank Malgioglio, Guenter Mayer, Christopher Berry, Michael Wood, Yiu-Hing Chan, Mark Mayo, John Isakson, Charudhattan Nagarajan, Tobias Werner, Leon Sigal, Ricardo Nigaglioni, Mark Cichanowski, Jeffrey Zitz, Matthew Ziegler, Tim Bronson, Gerald Strevig, Daniel Dreps, Ruchir Puri, Douglas Malone, Dieter Wendel, Pak-Kin Mak, Michael Blake

This work describes the circuit and physical design implementation of the processor chip (CP), level-4 cache chip (SC), and the multi-chip module at the heart of the EC12 system. The chips were implemented in IBM's high-performance 32nm high-k/metal-gate SOI technology. The CP chip contains 6 super-scalar, out-of-order processor cores, running at 5.5 GHz, while the SC chip contains 192 MB of eDRAM cache. Six CP chips and two SC chips are mounted on a high-performance glass-ceramic substrate, which provides high-bandwidth, low-latency interconnections. Various aspects of the design are explored in detail, with most of the focus on the CP chip, including the circuit design implementation, clocking, thermal modeling, reliability, frequency tuning, and comparison to the previous design in 45nm technology.

The new System z microprocessor chip (“CP chip”) features a high-frequency processor core running at 5.5GHz in a 32nm high-κ CMOS technology [1], using 15 levels of metal. This chip is a successor to the 45nm product [2], with significant improvements made to the core and nest (i.e. the logic external to the cores) in order to increase the performance and throughput of the design. Also, special considerations were necessary to ensure robust circuit operation in the high-κ technology used for implementation. As seen in the die photo, the chip contains 6 processor cores (compared to 4 cores in the 45nm version), and a large shared 48MB DRAM L3 cache. Each core includes a pair of data and instruction L2 SRAM caches of 1MB each. In addition, the chip contains a memory control unit (MCU), an I/O bus controller (GX), and two sets of interfaces to the L4 cache chips (also in 32nm technology). The CP chip occupies 598 mm2, contains about 2.75B transistors, and has 1071 signal IOs.