The Blue Gene project by IBM aims to create a series of massively parallel supercomputers capable of performing petaflop-level calculations with low power consumption. Key iterations include Blue Gene/L, Blue Gene/C, Blue Gene/P, and Blue Gene/Q, each advancing in architecture and performance capabilities. The project focuses on enhancing scientific computing through novel machine designs to tackle complex biological and computational problems.

2. WHAT IS BLUE GENE
•Blue Gene is an ambitious project to expand the
horizons of supercomputing, with the ultimate goal
of creating a system that can do one quadrillion
calculations per second, or perform one *petaflop.”
A massively parallel supercomputer using thousands of
embedded PowerPC processors supporting a large
memory space.

3. CONTD…
• With standard compilers and message
passing environment.
• Blue Gene is an IBM project aimed at designing
supercomputers that can reach operating speeds in the
PFLOPS(petaFLOPS) range, with low power
consumption.
 The first supercomputer to beat human in chess.
*A petaflop is a measure of a computer's processing speed and can be expressed as a thousand trillion floating point operations per second.

4. HISTORY
In December 1999 , IBM announced to build a massively
parallel computer, to be applied to study the protein gene
sequence.
Major areas of investigation included:
 The use of this novel platform to meet scientific goals
 Making of parallel machines more usable
 Achieving performance targets at reasonable cost through a
novel machine architecture.

5. RESULTS
 Linpack Top 500 Supercomputers

6. BLUE GENE PROJECTS
 Four Blue Gene projects :
 BlueGene/L
 BlueGene/C
 BlueGene/P
 BlueGene/Q

7. BLUE GENE/L
 The first computer in the Blue Gene series .
 Designed to deliver the most performance per kilowatt of
power consumed.
 It is a 16 rack system, with each rack holding 1024
compute nodes and a LINPAC performance of 70.72
*TFLOPS.
 Theoretical peak performance of 360 TFLOPS .
*TFLOPS -A tflop , or teraflop, is a parallel supercomputing system that has the ability to
compute one trillion floating point operations per second.

8. MAJOR FEATURES
 Trading the speed of processors for lower power
consumption.
 Dual processors per node with two working
modes: co-processor mode where one processor
handles computation and the other handles
communication.
 System-on-a-chip design. All node components
were embedded on one chip, with the exception
of 512 MB external DRAM.

9. 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

10. BLUE GENE/LARCHITECTURE CONTD…
 Double-pipeline-double-precision Floating Point Unit
 A cache sub-system with built-in DRAM controller
 Node CPUs are not cache coherent with one another
 FPUs and CPUs are designed for low power consumption

11. BLUE GENE/L ARCHITECTURE
CONTD…
1024 nodes
System Overview

12. BLUE GENE/L NETWORKS
 Each node is attached to 3 main parallel
communication networks
 3D Torus network - peer-2-peer between compute
nodes
 Collective network – collective & global
communication
 Ethernet network - I/O and management (such as access
to any node for configuration, booting and diagnostics )

13. BLUE GENE/L SYSTEM SOFTWARE
 System software supports efficient execution of parallel
applications
 Compiler support for *DFPU (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 a running
process
 Because of single-process nature, no context switching required
* DFPU - Double Floating Point Unit

14. BLUE GENE/L SYSTEM SOFTWARE
CONTD…
 To allow multiple programs to run concurrently
 Blue Gene/L system can be partitioned into electronically isolated sets of
nodes
 The number of nodes in a partition must be a positive integer power of 2
 To run program – reserve this partition
 No other program can use till partition is done with current program
 With so many nodes, component failures are inevitable. The system is
able to electrically isolate faulty hardware to allow the machine to
continue to run

15. BLUE GENE/L SYSTEM SOFTWARE
CONTD…
 Parallel Programming model
 Message Passing – supported through an
implementation of MPI
 Only a subset of POSIX calls are supported
 Green threads are also used to simulate local
concurrency

16. BLUE GENE/C
 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.

17. ARCHITECTURE OVERVIEW
 Each 64-bit Cyclops64 chip (processor) will run
at 500 megahertz and contain 80 processors.
 Each processor will have
two thread units and
a floating point unit.
 Five processors share a
32 kB instruction cache.

19. BLUE GENE/P
 Architecturally similar to BlueGene /L .
 Expected to operate around one petaflop.
 Launched in 2008.
 In here, the cores are cache coherent and the
chip can operate as a 4-way symmetric multi-
processor.
 The memory subsystem on the chip consist of
small private L2 caches , a central shared 8
MB cache , and dual DDR2 memory
controllers.

20. BLUE GENE/Q
 Third and the Last known supercomputer in
the Blue Gene series .
 Expected to reach 20 petaflops in 2012.
 Enhancement to the blue gene/L and P
architecture.

21. RESOURCES
 IBM
website(www.03.ibm.com/servers/deepcomputi
ng/bluegene.html)
 www.supercomp.org/sc2002/paperpdfs/pap.pap
207.pdf
 http://en.wikipedia.org/wiki/Blue_Gene
 http://community.anitaborg.org/wiki/images/9/9
2/GHC07-BlueGene_salapura.pdf