Cray-2 เป็น vector supercomputer ซึ่งถูกนำเสนอโดย Cray Research ในปี 1985 โดยจัดว่าเป็นเครื่องที่มีความเร็วที่สุดในโลกในขณะนั้น แทนที่รุ่น X- MP Cray Research ModelBackground processorsMemory Mwords 264 SRAM MOS 2126 SRAM MOS 4128 SRAM MOS 4256 DRAM MOS 4512 DRAM MOS รุ่นต่างๆที่ทำการผลิตออกมา
Cray Research Machine Type:Cray Y-MP EL (Entry Level) CPU Type:Cray Y-MP EL (Vector processor, CMOS technology) Number of processors:4 Main memory:128 MWords -> 1024 MByte Word length:64 bit Binary compatibility:up to the Cray-1 (1976) Local storage: 2x DD4 Hard Disks - 1 GB each One IBM 3740 (?) tape drive Network connectivity: 10 MBit Ethernet 125 MBit FDDI Specified performance: 133 MFlops per CPU peak 4-8 times a Cray-1 1/4 of a Cray-2 Measured peak performance: 460 MFlops (A 2.4 GHz P4/RDRAM is 770 MFlops on the same benchmark problem, huge matric multiply, with Linux gnu cc -O3) Operating system:Cray UNICOS (Unix-based) Commercial products installed: C Compiler, F77 Compiler Free software installed (/usr/local): Bash, GNU Make, Readline, GZip, less, vim, unzip, GNU tar, xaos, perl (in /opt)
Cray Research Machine Type: Cray T3D Serial number: 6223 Year of introduction: 1993 CPU Type: DEC ALPHA (Microprocessor) Number of processors: 128 Main memory: 8 MWord -> 64 MByte per CPU 1 Gword -> 8 GByte total Word length: 64 bit Binary compatibility: none Local storage: none (C90 used as frontend) Network connectivity: none (C90 used as frontend) HISP (200 MByte/sec) channel to C90 LOSP ( 20 MByte/sec) channel to C90 Specified performance: 150 MFlops per CPU peak 19 GFlops total peak Measured peak performance: not yet measured Operating system: Cray UNICOS max (Unix-based) Commercial products installed: C Compiler, F77 Compiler ข้อมูลจากจาก
Cray Research Cray T3E
Cray Research Cray T3E เป็น Supercomputer แบบ massively parallel มีการจัด จำหน่ายโดย Cray research ตั้งแต่ปี 1995 มีการออกแบบให้ใช้หน่วยประมลผลจำนวน ตั้งแต่ 8 จนถึง 2178 หน่วย o MHz Alpha CPUs o 600 MFLOPS peak per CPU o 8 KB L1 data cache, direct-mapped, 4-word line o 96 KB L2 data cache, 3-way set associative, 8-word line o 32.5 GB of RAM ข้อมูลจาก manila.cats.rwth-aachen.de/mech525/crayt3e manila.cats.rwth-aachen.de/mech525/crayt3e
Cray X1E System Highlights High Performance CPU The Cray X1E CPU uses vectorization and streaming to deliver peak performance of 18 GFLOPS per CPU. High memory bandwidth allows the processor to achieve a higher percentage of peak performance for applications than in most HPC systems. Scalable Interconnect and Memory The Cray X1E system is designed to scale beyond what any vector supercomputer has before. This scalability is made possible by extremely highperformance interconnect and memory subsystems. While the Cray X1E memory is physically distributed on individual modules,anyMSP can logically share any memory. Using standard load and store instructions, an MSP can use a remote memory address to read or write memory located on a separate module. The system supports very high concurrency to allow applications to tolerate global network latencies.
Cray Inc True Single System Image OS UNICOS/mp™, the Cray X1E operating system, is a true single system image operating system (OS). Regardless of system size, system administrators need only manage and configure a single OS — significantly reducing the time and effort needed for software upgrades and other tasks. High Performance Application Support The Cray X1E programming environments support multiple levels of parallel programming, with a wide choice of programming models. Optimizing Fortran, C, and C++ compilers perform automatic vectorization and streaming, and support OpenMP programming within a node, and Co-Array Fortran or UPC across the system. Optimized libraries are included for MPI and SHMEM, as well as for a large number of key scientific functions. To assist programmers in developing and optimizing applications, the TotalView® debugger and powerful performance analysis tools are availablefor the Cray X1E system.
Cray Inc Cray XT3
Scalable Application Performance The Cray XT3 supercomputer’s high speed 3D torus interconnect, x86 64-bit AMD Opteron™ processors, high speed global I/O, and advanced MPP operating system ensure that applications scale steadily from 200 to 30,000 processors without performance losses from communications bottlenecks, asynchronous processing, or memory access delays. Scalable Reliability and Management Each Cray XT3 component, from industrial cooling fans, to disk drives, to the Cray Reliability, Availability and Serviceability (RAS) and Management System, is engineered to operate as part of a highly reliable system at immense scale, ensuring that large, complex jobs run to completion. Tightly integrated operating and management systems allow administrators to manage hundreds or thousands of processors as a single system, eliminating the administrative effort and problems associated with loosely coupled cluster systems. Cray XT3 System Highlights
Cray Inc Scalable Programmability The Cray XT3 supercomputer lets programmers focus on their applications instead of designing around processing inefficiencies such as asymmetric processor performance, memory access algorithms, and communication delays. Fully scalable performance analysis and debugging tools enable programmers to rapidly test and fine-tune their applications. Scalable I/O The Cray XT3 I/O system uses the highly scalable, open source Lustre™ parallel file system. Highly reliable Fibre Channel disks and controllers provide up to 100 GB/s global I/O performance, ensuring high efficiency for I/O intensive applications and providing the I/O capacity needed for rapid data dumps and user level checkpointing. Scalable System Upgrades Cray XT3 systems can be expanded by adding cabinets or by upgrading Opteron processors with faster or dual- core models, or upgrading the Cray SeaStar processor to increase interconnect speeds. This flexible expansion ensures a long system life, maximizing return on investment.
Cray Inc Cray XD1
Cray XD1 System Highlights Compute Processors 12 AMD Opteron™ 64-bit single or dual core processors run Linux and are organized as six nodes of 2 or 4-way SMPs to deliver up to 106 GFLOPS* per chassis. Matching memory and I/O performance removes bottlenecks and maximizes processor performance. RapidArray™ Interconnect The industry’s fastest embedded switching fabric, the RapidArray interconnect uses 12 custom communications processors and a 96 GigaBytes (GB) per second nonblocking switching fabric per chassis to deliver 8 GB per second bandwidth between nodes with 1.7 microsecond MPI latency. Each chassis presents 24 RapidArray links externally with an aggregate 48 GB per second bandwidth between chassis.
Cray Inc Application Acceleration Six Xilinx Virtex-4 Field Programmable Gate Arrays (FPGAs) per chassis attach to the RapidArray fabric for massively parallel execution of critical algorithm components, promising orders of magnitude performance improvement for target applications. Active Management A management processor in each chassis, a realtime operating system, distributed software and an independent supervisory network operate together to monitor, control and manage every aspect of the computer. The Cray XD1 system enables single system command and control and provides extensive high availability features.