Cluster Examples

• The Berkeley Network of Workstations (NOW) is an important representative of cluster systems.
• In 1997, the NOW project achieved over 10 Gflops on the Linpack benchmark, which made it one of the top 200 fastest supercomputers in the world.
• The hardware/software infrastructure for the project included 100 SUN Ultrasparcs and 40 SUN Sparcstations running Solaris, 35 Intel PCs running Windows NT or a PC Unix variant, and between 500 and 1000 disks, all connected by a Myrinet switched network.
• The programming environments used in NOW are sockets, MPI, and a parallel version of C, called Split C.
• Active Messages is the basic communication primitive in Berkeley NOW.

• The idea of the Beowulf cluster project was to achieve supercomputer processing power using off-the-shelf commodity machines.
• One of the earliest Beowulf clusters contained sixteen 100 MHz DX4 processors that were connected using 10 Mbps Ethernet.
• The second Beowulf cluster, built in 1995, used 100 MHz Pentium processors connected by 100 Mbps Ethernet.
• The third generation of Beowulf clusters was built by different research laboratories JPL and Los Alamos National Laboratory each built a 16-processor machine incorporating Pentium Pro processors.
• These machines were combined to run a large N-body problem, which won the 1997 Gordon Bell Prize for high performance.

• The communication between processors in Beowulf has been done through TCP/IP over the Ethernet internal to the cluster.
• Multiple Ethernets were also used to satisfy higher bandwidth requirements.
• Channel bonding is a technique to connect multiple Ethernets in order to distribute the communication traffic.
• Channel bonding was able to increase the sustained network throughput by 75% when dual networks were used.
• Two of the early successful Beowulf clusters are Loki and Avalon.
• In 1997, Loki was built using 16 Pentium Pro Processors connected using Fast Ethernet switches. It achieved 1.2 Gflops.
• In 1998, the Avalon was built using one hundred and forty 533 MHz Alpha Microprocessors connected. Avalon achieved 47.7 Gflops.

• In April 2004, the University of San Francisco hosted the first Flash Mob Computing computer; FlashMob I, with the purpose of creating one of the fastest supercomputers on the planet.
• A FlashMob supercomputer was created by connecting a large number of computers via a high-speed LAN, to work together as a single supercomputer.
• A FlashMob computer, unlike an ordinary cluster, is temporary and organized ad hoc for the purpose of working on a single problem.
• It used volunteers and ordinary laptop PCs, and was designed to allow anyone to create a supercomputer in a matter of hours.
• Over 700 computers came into the gym and they were able to hook up 669 to the network.
• The best Linpack result was a peak rate of 180 Gflops using 256 computers; however, a node failed 75% through the computation.
• The best completed result was 77 Gflops using 150 computers.