Key Words in Computer Architecture

• Chap.1
• Performance Evaluation
• Trace
• SPEC(Standard Performance Evaluation Corporation)/ Perfect Club
• Chap.2
• MCM (Multichip module)
• PIM (Processing in Memory)
• Decoupled architecture
• Superscalar
• Vector Processor, Array Processor
• Out-of-order execution
• Multithreaded Architecture
• Simultaneous multithreading
• Instruction-level parallelism, ILP
• Aynchronous processor
• SUN's UltraSparc, HP's PA-8000, Intel's MMX
• Hitachi's SH3, ARM7TDMI, MIPS' R10000
• Chomatic's Mpact, Media Processor, TI's C82, Software Modem
• Chap.3&4
• Register File
• Reservation table, forbidden latencies, collision vector
• Stalling
• Structural hazard
• Branch prediction, Branch Effect
• Prefetch buffer, Data prefetch
• Instruction scheduling (Dynamic scheduling)
• Chap.5
• Write-back (Copy-back) / Write-through
• Data/Instruction cache
• Associative cache
• L2 (Level 2) Cache
• EDO DRAM
• RDRAM (RAMBUS DRAM)
• IRAM (Intelligent RAM)
• Address translation
• Virtual addressed cache
• Chap.6
• DMA (Direct Memory Access)
• RAID (Redundant Array of Inexpensive Disks)
• Chap.7
• Interconnection Network
• Direct interconnect
• Hypercube / Fat-tree / Mesh / Torus / DeBruijin / k-ary n-cube
• Cluster-based Multiprocessor
• Node degree, Diameter
• Reconfigurable networks
• MIN (Multistage interconnection network)
• Permutation
• Partitioning
• Hot Spot
• Wormhole routing, e-cube routing
• Routing chip
• Virtual channel
• Latency hiding
• PCI, Myrinet, ATM, Ethernet
• Chap.8
• Shared-memory multiprocessor
• UMA / NUMA (Non-uniform memory access)
• Coherent Caches
• Snooping cache coherency
• COMA (Cache Only Memory Architecture), Attraction memory
• Shared Virtual Memory
• Memory Consistency (Lazy Release Consistency)
• PVM (Parallel Virtual Machine), MPI (Message Passing Interface)
• Coscheduling
• Fetch&add Synchronization
• NOW (Network of Workstations)
• TreadMarks, SUIF