Bruce Petrick

A 40nm 16-core 128-thread CMT SPARC SoC processor

This next generation of Chip Multithreaded (CMT) SPARC SoC processor, code named Rainbow Falls, doubles on-chip thread count over its predecessor the UltraSparc T2+. The chip offers high levels of integration and scalability with twice the number of cores, a larger L2 cache, and higher maximum I/O bandwidth, within the same power envelope. Sixteen 8-threaded enhanced SPARC cores (SPC) provide 128 threads in a single die, delivering the highest thread count for a general-purpose microprocessor. The new cache coherency further allows up to 4-way glueless systems with a total of 512 threads. Each core communicates with the unified 6MB L2 cache through a crossbar (CCX) delivering 461GB/s (Fig. 5.2.1). A gasket (CXG) is also introduced to manage the congestion and synchronization of the massive interconnect between the 16 cores and the crossbar. This facilitates a synchronized delay control between any core and any L2 bank for partial core product binning and testing.

A dual-core 64b UltraSPARC microprocessor for dense server applications

A dual-core 64-bit microprocessor optimized for compute-dense systems such as rack-mount and blade servers for network computing was developed. The chip consists of two UltraSPARC II cores, each with its own 512 kB L2 cache, a DDR-1 memory controller, and symmetric multiprocessor bus (JBus) controllers. The 206-mm/sup 2/ die is fabricated in 0.13-/spl mu/m CMOS technology with seven layers of Cu and a low-k dielectric. The chip offers a highly efficient performance-per-watt ratio with a typical power dissipation of 23 W at 1.3 V and 1.2 GHz. A short design cycle was achieved by leveraging existing designs wherever possible and developing effective design methodologies and flows. Significant design challenges faced by this project are described. These include deep-submicron design issues, such as negative bias temperature instability (NBTI), leakage, coupling noise, intra-die process variation, and electromigration (EM). A second important design challenge was implementing a high-performance L2 cache subsystem with a short four-cycle core-to-L2 latency including ECC.

Design and implementation of an embedded 512KB level 2 cache subsystem

Dual on-chip 512 kB unified second level (L2) caches for an UltraSparc processor are implemented using 0.13 /spl mu/m technology. Each 512 kB unit is implemented using 34 million transistors to achieve 1.4 GHz and 2.6 W at 13 V and 85 C. This fully integrated subsystem is composed of data and tag SRAMs along with datapaths, controller and test engines. The unit achieves one of the shortest on-chip L2 cache latencies reported for 64b microprocessors, with a data latency of only 4 cycles including ECC correction for 128-bit data. The design solutions to build this integrated short latency L2 cache are discussed.