IBM BLUE GENE
Guided By: Ms. Sashikala Mishra Asst. Professor Comp. Science Dept.
Submitted By: Akash Paltasingh 0811012093 CSE, Sec-C
What is Blue Gene ?
Its a Massively Parallel Processing supercomputer using tens of thousands of embedded PowerPC processors supporting a large memory space. Aims on achieving speeds on PetaFLOPS scale, using unprecedented
levels of parallelism.
Why the name Blue Gene?
Blue: The corporate color of IBM Gene: The intended use of the Blue Gene clusters Computational biology, specifically, protein folding and DNA Computing.
History
Dec99, IBM Research announced $100M US effort to build a PetaFLOPS scale supercomputer.
The Blue Gene Project has two primary goals: Advance the state of the art Bio-Molecular Simulation Advance the state of the art in software and architecture design in extremely large scale systems.
November 2001, IBM announced partnership with Lawrence Livermore National Laboratory (LLNL) for the Blue Gene Project and the first system in the series IBM BG/L.
Blue Gene Projects
Four Blue Gene projects :
BlueGene/L BlueGene/C BlueGene/P BlueGene/Q
Blue Gene/L
The first computer in the Blue Gene series IBM first announced the Blue Gene/L project, Sept. 29, 2004 Originally had a theoretical peak performance of 360 TFLOPS, and after an upgrade in 2007 the performance increased to 478 TFLOPS sustained and 596 TFLOPS peak.
Blue Gene/L Architecture
Can be scaled up to 65,536 compute or I/O nodes, with 131,072 processors
Each node is a single ASIC with associated DRAM memory chips Each ASIC has 2 700 MHz IBM PowerPC processors PowerPC processors
Low-frequency, low-power embedded processors, superior to today's high-frequency, high-power microprocessors by a factor of 2 or more.
Blue Gene/L Architecture contd
1024 nodes
System Overview
Blue Gene/L Architecture cont.
1 rack holds 1024 nodes or 2048 processors Nodes optimized for low power consumption ASIC based on System-on-a-chip technology
Large numbers of low-power system-on-a-chip technology allows it to outperform commodity clusters while saving on power
Aggressive packaging of processors, memory and interconnect Power Efficient & Space Efficient Allows for latencies and bandwidths that are significantly better than those for nodes typically used in ASC scale supercomputers
Blue Gene/L System Architecture contd..
The BG/L ASIC supports 4 types of networks:
3D Torus Intra-Processor Network Tree Network Ethernet Network Global Interrupts
3D Torus Intra Processor Network
Its a three dimensional mesh comprising of numerous computing nodes. All nodes are capable of intra node communications, and so divide the work and process it faster and more efficiently.
Each node provides about 4.2Gbps, providing a total of about 134Tbps bandwidth.
Global Tree Network
Provides One-to-all or all-all broadcast. Arithmetic operation conducted in tree. Forms a bandwidth of about 1.44Gbps from any node to another.
Ethernet Network
This is incorporated into each ASIC Chip. Provides I/O Control, Host control, Booting and Diagnostics.
Blue Gene/L System Software
System software supports efficient execution of parallel applications Compiler support for languages like C, C++, Fortran. Compute nodes use a minimal operating system called BlueGene/L compute node kernel
A lightweight, single-user operating system Supports execution of a single dual-threaded application compute process Kernel provides a single and static virtual address space to one running compute process Because of single-process nature, no context switching required
Blue Gene/C
Sister-project to BlueGene/L Renamed to Cyclops64 Massively parallel, supercomputer-on-a-chip cellular architecture Cellular architecture gives the programmer the ability to run large
numbers of concurrent threads within a single processor.
BG/C- Cellular Architecture
Implements Thread-level Parallelism. Each cell comprises of a compute node, containing thread units, memory, and communication module.
Used in PlayStation 3 .
Blue Gene/P
Architecturally similar to BlueGene/L .
Configures over 294912 processors in 72 racks, to work at 1petaFlops,
and over 884,736 processors over 214 racks, to achieve 3 petaFLOPS.
Blue Gene/Q
Last known supercomputer in the Blue Gene series Aimed to achieve over 20petaFLOPS . Uses similar architecture as BG/L, with a improved power efficiency of 1684 MFLOPS/Watt.
Latest Advancements : IBM BG/Q has been implemented in various low yield prototypes as,
IBM Sequoia IBM Watson IBM Mira
The full yield system is to set to be released by Nov2012.
Resources
www.wikipedia.org IBM website
(www.03.ibm.com/servers/deepcomputing/bluegene.html) (www.research-ibm.com)
www.supercomp.org/sc2002/paperpdfs/pap.pap207.pdf www.scribd.com
THANK YOU ALL.
ANY QUESTIONS PLEASE ?
