Re: SHRIMP talk at UCI 11/20

Joe Kiniry (
Tue, 19 Nov 96 13:42:44 -0800

You wrote:
> This sounds similar to the Infospheres project -- perhaps it would
> be worth a trip down for the talk?
> - Jim

um, actually, this has little or nothing to do with what we do.
they're creating a new high-performance multi-computer out of
commodity parts (pentiums and what sounds like the mesh-connect from
a paragon, though i don't know if that is commodity or not). they
are interested in operating system-level work (memory mapping,
communication, performance monitoring, etc.) for such a machine.

from their web page (
> SHRIMP (Scalable, High-Performance, Really Inexpensive
> Multi-Processor) is a parallel machine being designed and
> built in the _Computer Science Department_ at _Princeton
> University_. Shrimp is built from highly-integrated,
> commodity parts. The computing nodes of SHRIMP are Pentium
> PCs, and the routing network is the same one used in the
> Intel Paragon. We are building a network interface card that
> connects the PCs to the routing network, and we are creating
> the software to make SHRIMP a fully usable multicomputer.

to clarify; we (the infospheres work) are interested in services,
structures, models, etc. to support reasoning on compositional
distributed systems on a global scale. we don't build hardware, we
hardly build any software, and we certainly stay out of the
operating system space.

from our web page(
> The Caltech Infospheres Project deals with the theory and
> implementation of compositional systems that support
> peer-to-peer communication among persistent multithreaded
> distributed objects. Though our implementation of a generic
> distributed system uses Java, TCP/IP, and the Web, the ideas
> are directly applicable to any distributed (possibly
> object-oriented) system that supports messages and threads. A
> key concern is the development of reliable distributed
> applications by composing existing and created software
> components in structured ways. We are solving the problems
> and influencing the standards of distributed component
> technology for the user and developer for the Internet of the
> 21st century.

i don't mean to bite, i just want to make clear what we work on to
this group of folks.