A block diagram of the Tukwila processor is shown in Fig. 3. The die contains four multi-threaded high performance 64 bit cores. Associated with each core is 6 MB of level three cache implementing the Intel Cache Safe Technology [3]. A system interface is designed around a 12 port crossbar router that allows
communication between the four cores, two home agents, and six IO channels. Associated with each home agent is a 1MB directory cache in support of a directory-based cache coherence protocol. Dual integrated memory controllers allow communication to system memory through four full duplex FBD2 channels with a peak bandwidth of 34 GB/s. Four full width and two half width Intel QuickPath Interconnects (QPI) [4] allow
processor to IO and processor-to-processor communication at a peak bandwidth of 96 GB/s. To connect the system interface to the core and IO physical layer, Tukwila implements a synchronizer and routing architecture that is distributed across the die. Finally, the charge rationing (QR) controller monitors chip activity factor, and together with the Tukwila clock system, allows dynamic modulation of the core voltage and frequency within a fixed power envelope.