Portals network programming api
Encyclopedia
Portals is a low-level network API
for high-performance networking on high-performance computing
systems developed by Sandia National Laboratories
and the University of New Mexico
. Portals is currently the lowest-level network programming interface on the commercially successful XT line of supercomputers from Cray
.
. Instead, the ultimate destination in memory of an incoming message is determined at the receiver by comparing contents of the message header with the contents of structures at the destination. This flexibility allows for efficient implementations of both one-sided and two-sided communications. In particular, Portals is aimed at providing the fundamental operations necessary to support a high-performance and scalable implementation of the Message Passing Interface
(MPI) standard. It was also used as the initial network transport layer for the Lustre
file system.
message passing system used in the SUNMOS
and Puma operating system. It was first implemented for the Intel Paragon
at Sandia, and later ported to the Intel TeraFLOPS machine named ASCI Red
.
There were four building blocks in the first version of Portals: the single block, the dynamic block, the independent block and the combined block. All incoming messages would first pass through a match-list that allowed individual portals to respond to specific groups, ranks, and a set of user specified match-bits.
The MPI library was ported from the retronymed
Portals 2 to the new Portals 3.0.
The Portals 3 specification has been implemented several times, first on ASCI Red
, then on CPlant over Myrinet
, Linux and the Cray XT
family.
(PGAS) languages, several new features have been added to the Portals API as part of Portals 4. Portals 4 also made several changes to improve the interaction between the processor and network interface (NIC) for implementations that provide offload. Finally, an option to support a form of flow-control was added to Portals 4..
Application programming interface
An application programming interface is a source code based specification intended to be used as an interface by software components to communicate with each other...
for high-performance networking on high-performance computing
High-performance computing
High-performance computing uses supercomputers and computer clusters to solve advanced computation problems. Today, computer systems approaching the teraflops-region are counted as HPC-computers.-Overview:...
systems developed by Sandia National Laboratories
Sandia National Laboratories
The Sandia National Laboratories, managed and operated by the Sandia Corporation , are two major United States Department of Energy research and development national laboratories....
and the University of New Mexico
University of New Mexico
The University of New Mexico at Albuquerque is a public research university located in Albuquerque, New Mexico, in the United States. It is the state's flagship research institution...
. Portals is currently the lowest-level network programming interface on the commercially successful XT line of supercomputers from Cray
Cray
Cray Inc. is an American supercomputer manufacturer based in Seattle, Washington. The company's predecessor, Cray Research, Inc. , was founded in 1972 by computer designer Seymour Cray. Seymour Cray went on to form the spin-off Cray Computer Corporation , in 1989, which went bankrupt in 1995,...
.
Overview
Portals is based on the concept of elementary building blocks that can be combined to support a wide variety of upper-level network transport semantics. Portals provides one-sided data movement operations, but unlike other one-sided programming interfaces, the target of a remote operation is not a virtual addressVirtual address
In computer technology, a virtual address is an address identifying a virtual, i.e. non-physical, entity.-Description:The term virtual address is most commonly used for an address pointing to virtual memory or, in networking, when referring to a virtual network address...
. Instead, the ultimate destination in memory of an incoming message is determined at the receiver by comparing contents of the message header with the contents of structures at the destination. This flexibility allows for efficient implementations of both one-sided and two-sided communications. In particular, Portals is aimed at providing the fundamental operations necessary to support a high-performance and scalable implementation of the Message Passing Interface
Message Passing Interface
Message Passing Interface is a standardized and portable message-passing system designed by a group of researchers from academia and industry to function on a wide variety of parallel computers...
(MPI) standard. It was also used as the initial network transport layer for the Lustre
Lustre (file system)
Lustre is a massively parallel distributed file system, generally used for large scale cluster computing. The name Lustre is a portmanteau word derived from Linux and cluster...
file system.
Early years
Portals began in the early 1990s as an extension to the nXNCUBE
nCUBE was a series of parallel computing computers from the company of the same name. Early generations of the hardware used a custom microprocessor...
message passing system used in the SUNMOS
SUNMOS
SUNMOS is an operating system jointly developed by Sandia National Laboratories and the Computer Science Department at the University of New Mexico...
and Puma operating system. It was first implemented for the Intel Paragon
Intel Paragon
The Intel Paragon was a series of massively parallel supercomputers produced by Intel. The Paragon XP/S was a productized version of the experimental Touchstone Delta system built at Caltech, launched in 1992. The Paragon superseded Intel's earlier iPSC/860 system, to which it was closely...
at Sandia, and later ported to the Intel TeraFLOPS machine named ASCI Red
ASCI Red
ASCI Red was the first computer built under the Advanced Strategic Computing Initiative . ASCI Red was built by Intel and installed at Sandia in late 1996. The design was based on the Intel Paragon computer...
.
There were four building blocks in the first version of Portals: the single block, the dynamic block, the independent block and the combined block. All incoming messages would first pass through a match-list that allowed individual portals to respond to specific groups, ranks, and a set of user specified match-bits.
Portals 3
The Portals concept continued to evolve over successive generations of lightweight kernels and massively parallel systems. In 1999, an operational programming interface was given to Portals so that it could be implemented for intelligent and/or programmable network interfaces outside of a lightweight kernel environment.. This standard was designed for systems where the work required to prepare, transmit, and deliver messages is longer than the round-trip to the Portals data structures. For example, in modern systems, this work is dominated by the round-trip through the IO bus to the network interface. The standard has been revised since the initial release to make it more suited for modern high performance, massively parallel computers.The MPI library was ported from the retronymed
Retronym
A retronym is a type of neologism that provides a new name for an object or concept to differentiate the original form or version of it from a more recent form or version. The original name is most often augmented with an adjective to account for later developments of the object or concept itself...
Portals 2 to the new Portals 3.0.
The Portals 3 specification has been implemented several times, first on ASCI Red
ASCI Red
ASCI Red was the first computer built under the Advanced Strategic Computing Initiative . ASCI Red was built by Intel and installed at Sandia in late 1996. The design was based on the Intel Paragon computer...
, then on CPlant over Myrinet
Myrinet
Myrinet, ANSI/VITA 26-1998, is a high-speed local area networking system designed by Myricom to be used as an interconnect between multiple machines to form computer clusters. Myrinet has much lower protocol overhead than standards such as Ethernet, and therefore provides better throughput, less...
, Linux and the Cray XT
Cray XT3
The Cray XT3 is a distributed memory massively parallel MIMD supercomputer designed by Cray Inc. with Sandia National Laboratories under the codename Red Storm. Cray turned the design into a commercial product in 2004...
family.
Portals 4
In light of emerging partitioned global address spacePartitioned global address space
In computer science, a partitioned global address space is a parallel programming model. It assumes a global memory address space that is logically partitioned and a portion of it is local to each processor. The novelty of PGAS is that the portions of the shared memory space may have an affinity...
(PGAS) languages, several new features have been added to the Portals API as part of Portals 4. Portals 4 also made several changes to improve the interaction between the processor and network interface (NIC) for implementations that provide offload. Finally, an option to support a form of flow-control was added to Portals 4..