Overview¶
MPI for Python provides an object oriented approach to message passing which grounds on the standard MPI-2 C++ bindings. The interface was designed with focus in translating MPI syntax and semantics of standard MPI-2 bindings for C++ to Python. Any user of the standard C/C++ MPI bindings should be able to use this module without need of learning a new interface.
Communicating Python Objects and Array Data¶
The Python standard library supports different mechanisms for data persistence. Many of them rely on disk storage, but pickling and marshaling can also work with memory buffers.
The pickle modules provide user-extensible facilities to
serialize general Python objects using ASCII or binary formats. The
marshal module provides facilities to serialize built-in Python
objects using a binary format specific to Python, but independent of
machine architecture issues.
MPI for Python can communicate any built-in or user-defined Python
object taking advantage of the features provided by the pickle
module. These facilities will be routinely used to build binary
representations of objects to communicate (at sending processes), and
restoring them back (at receiving processes).
Although simple and general, the serialization approach (i.e., pickling and unpickling) previously discussed imposes important overheads in memory as well as processor usage, especially in the scenario of objects with large memory footprints being communicated. Pickling general Python objects, ranging from primitive or container built-in types to user-defined classes, necessarily requires computer resources. Processing is also needed for dispatching the appropriate serialization method (that depends on the type of the object) and doing the actual packing. Additional memory is always needed, and if its total amount is not known a priori, many reallocations can occur. Indeed, in the case of large numeric arrays, this is certainly unacceptable and precludes communication of objects occupying half or more of the available memory resources.
MPI for Python supports direct communication of any object exporting the single-segment buffer interface. This interface is a standard Python mechanism provided by some types (e.g., strings and numeric arrays), allowing access in the C side to a contiguous memory buffer (i.e., address and length) containing the relevant data. This feature, in conjunction with the capability of constructing user-defined MPI datatypes describing complicated memory layouts, enables the implementation of many algorithms involving multidimensional numeric arrays (e.g., image processing, fast Fourier transforms, finite difference schemes on structured Cartesian grids) directly in Python, with negligible overhead, and almost as fast as compiled Fortran, C, or C++ codes.
Communicators¶
In MPI for Python, Comm is the base class of communicators. The
Intracomm and Intercomm classes are sublcasses of the Comm
class. The Comm.Is_inter method (and Comm.Is_intra, provided for
convenience but not part of the MPI specification) is defined for
communicator objects and can be used to determine the particular
communicator class.
The two predefined intracommunicator instances are available:
COMM_SELF and COMM_WORLD. From them, new communicators can be
created as needed.
The number of processes in a communicator and the calling process rank
can be respectively obtained with methods Comm.Get_size and
Comm.Get_rank. The associated process group can be retrieved from a
communicator by calling the Comm.Get_group method, which returns an
instance of the Group class. Set operations with Group objects
like like Group.Union, Group.Intersection and Group.Difference
are fully supported, as well as the creation of new communicators from
these groups using Comm.Create and Comm.Create_group.
New communicator instances can be obtained with the Comm.Clone,
Comm.Dup and Comm.Split methods, as well methods
Intracomm.Create_intercomm and Intercomm.Merge.
Virtual topologies (Cartcomm, Graphcomm and Distgraphcomm
classes, which are specializations of the Intracomm class) are fully
supported. New instances can be obtained from intracommunicator
instances with factory methods Intracomm.Create_cart and
Intracomm.Create_graph.
Point-to-Point Communications¶
Point to point communication is a fundamental capability of message passing systems. This mechanism enables the transmission of data between a pair of processes, one side sending, the other receiving.
MPI provides a set of send and receive functions allowing the communication of typed data with an associated tag. The type information enables the conversion of data representation from one architecture to another in the case of heterogeneous computing environments; additionally, it allows the representation of non-contiguous data layouts and user-defined datatypes, thus avoiding the overhead of (otherwise unavoidable) packing/unpacking operations. The tag information allows selectivity of messages at the receiving end.
Blocking Communications¶
MPI provides basic send and receive functions that are blocking. These functions block the caller until the data buffers involved in the communication can be safely reused by the application program.
In MPI for Python, the Comm.Send, Comm.Recv and Comm.Sendrecv
methods of communicator objects provide support for blocking
point-to-point communications within Intracomm and Intercomm
instances. These methods can communicate memory buffers. The variants
Comm.send, Comm.recv and Comm.sendrecv can communicate general
Python objects.
Nonblocking Communications¶
On many systems, performance can be significantly increased by overlapping communication and computation. This is particularly true on systems where communication can be executed autonomously by an intelligent, dedicated communication controller.
MPI provides nonblocking send and receive functions. They allow the possible overlap of communication and computation. Non-blocking communication always come in two parts: posting functions, which begin the requested operation; and test-for-completion functions, which allow to discover whether the requested operation has completed.
In MPI for Python, the Comm.Isend and Comm.Irecv methods
initiate send and receive operations, respectively. These methods
return a Request instance, uniquely identifying the started
operation. Its completion can be managed using the Request.Test,
Request.Wait and Request.Cancel methods. The management of
Request objects and associated memory buffers involved in
communication requires a careful, rather low-level coordination. Users
must ensure that objects exposing their memory buffers are not
accessed at the Python level while they are involved in nonblocking
message-passing operations.
Persistent Communications¶
Often a communication with the same argument list is repeatedly executed within an inner loop. In such cases, communication can be further optimized by using persistent communication, a particular case of nonblocking communication allowing the reduction of the overhead between processes and communication controllers. Furthermore , this kind of optimization can also alleviate the extra call overheads associated to interpreted, dynamic languages like Python.
In MPI for Python, the Comm.Send_init and Comm.Recv_init methods
create persistent requests for a send and receive operation,
respectively. These methods return an instance of the Prequest
class, a subclass of the Request class. The actual communication can
be effectively started using the Prequest.Start method, and its
completion can be managed as previously described.
Collective Communications¶
Collective communications allow the transmittal of data between multiple processes of a group simultaneously. The syntax and semantics of collective functions is consistent with point-to-point communication. Collective functions communicate typed data, but messages are not paired with an associated tag; selectivity of messages is implied in the calling order. Additionally, collective functions come in blocking versions only.
The more commonly used collective communication operations are the following.
Barrier synchronization across all group members.
Global communication functions
Broadcast data from one member to all members of a group.
Gather data from all members to one member of a group.
Scatter data from one member to all members of a group.
Global reduction operations such as sum, maximum, minimum, etc.
In MPI for Python, the Comm.Bcast, Comm.Scatter, Comm.Gather,
Comm.Allgather, Comm.Alltoall methods provide support for
collective communications of memory buffers. The lower-case variants
Comm.bcast, Comm.scatter, Comm.gather, Comm.allgather and
Comm.alltoall can communicate general Python objects. The vector
variants (which can communicate different amounts of data to each
process) Comm.Scatterv, Comm.Gatherv, Comm.Allgatherv,
Comm.Alltoallv and Comm.Alltoallw are also supported, they can
only communicate objects exposing memory buffers.
Global reducion operations on memory buffers are accessible through
the Comm.Reduce, Comm.Reduce_scatter, Comm.Allreduce,
Intracomm.Scan and Intracomm.Exscan methods. The lower-case
variants Comm.reduce, Comm.allreduce, Intracomm.scan and
Intracomm.exscan can communicate general Python objects; however,
the actual required reduction computations are performed sequentially
at some process. All the predefined (i.e., SUM, PROD, MAX, etc.)
reduction operations can be applied.
Support for GPU-aware MPI¶
Several MPI implementations, including Open MPI and MVAPICH, support passing GPU pointers to MPI calls to avoid explict data movement between the host and the device. On the Python side, GPU arrays have been implemented by many libraries that need GPU computation, such as CuPy, Numba, PyTorch, and PyArrow. In order to increase library interoperability, two kinds of zero-copy data exchange protocols are defined and agreed upon: DLPack and CUDA Array Interface. For example, a CuPy array can be passed to a Numba CUDA-jit kernel.
MPI for Python provides an experimental support for GPU-aware MPI. This feature requires:
mpi4py is built against a GPU-aware MPI library.
The Python GPU arrays are compliant with either of the protocols.
See the Tutorial section for further information. We note that
Whether or not a MPI call can work for GPU arrays depends on the underlying MPI implementation, not on mpi4py.
This support is currently experimental and subject to change in the future.
Dynamic Process Management¶
In the context of the MPI-1 specification, a parallel application is static; that is, no processes can be added to or deleted from a running application after it has been started. Fortunately, this limitation was addressed in MPI-2. The new specification added a process management model providing a basic interface between an application and external resources and process managers.
This MPI-2 extension can be really useful, especially for sequential applications built on top of parallel modules, or parallel applications with a client/server model. The MPI-2 process model provides a mechanism to create new processes and establish communication between them and the existing MPI application. It also provides mechanisms to establish communication between two existing MPI applications, even when one did not start the other.
In MPI for Python, new independent process groups can be created by
calling the Intracomm.Spawn method within an intracommunicator.
This call returns a new intercommunicator (i.e., an Intercomm
instance) at the parent process group. The child process group can
retrieve the matching intercommunicator by calling the
Comm.Get_parent class method. At each side, the new
intercommunicator can be used to perform point to point and collective
communications between the parent and child groups of processes.
Alternatively, disjoint groups of processes can establish
communication using a client/server approach. Any server application
must first call the Open_port function to open a port and the
Publish_name function to publish a provided service, and next call
the Intracomm.Accept method. Any client applications can first find
a published service by calling the Lookup_name function, which
returns the port where a server can be contacted; and next call the
Intracomm.Connect method. Both Intracomm.Accept and
Intracomm.Connect methods return an Intercomm instance. When
connection between client/server processes is no longer needed, all of
them must cooperatively call the Comm.Disconnect
method. Additionally, server applications should release resources by
calling the Unpublish_name and Close_port functions.
One-Sided Communications¶
One-sided communications (also called Remote Memory Access, RMA) supplements the traditional two-sided, send/receive based MPI communication model with a one-sided, put/get based interface. One-sided communication that can take advantage of the capabilities of highly specialized network hardware. Additionally, this extension lowers latency and software overhead in applications written using a shared-memory-like paradigm.
The MPI specification revolves around the use of objects called windows; they intuitively specify regions of a process’s memory that have been made available for remote read and write operations. The published memory blocks can be accessed through three functions for put (remote send), get (remote write), and accumulate (remote update or reduction) data items. A much larger number of functions support different synchronization styles; the semantics of these synchronization operations are fairly complex.
In MPI for Python, one-sided operations are available by using
instances of the Win class. New window objects are created by
calling the Win.Create method at all processes within a communicator
and specifying a memory buffer . When a window instance is no longer
needed, the Win.Free method should be called.
The three one-sided MPI operations for remote write, read and
reduction are available through calling the methods Win.Put,
Win.Get, and Win.Accumulate respectively within a Win instance.
These methods need an integer rank identifying the target process and
an integer offset relative the base address of the remote memory block
being accessed.
The one-sided operations read, write, and reduction are implicitly
nonblocking, and must be synchronized by using two primary modes.
Active target synchronization requires the origin process to call the
Win.Start and Win.Complete methods at the origin process, and
target process cooperates by calling the Win.Post and Win.Wait
methods. There is also a collective variant provided by the
Win.Fence method. Passive target synchronization is more lenient,
only the origin process calls the Win.Lock and Win.Unlock
methods. Locks are used to protect remote accesses to the locked
remote window and to protect local load/store accesses to a locked
local window.
Parallel Input/Output¶
The POSIX standard provides a model of a widely portable file system. However, the optimization needed for parallel input/output cannot be achieved with this generic interface. In order to ensure efficiency and scalability, the underlying parallel input/output system must provide a high-level interface supporting partitioning of file data among processes and a collective interface supporting complete transfers of global data structures between process memories and files. Additionally, further efficiencies can be gained via support for asynchronous input/output, strided accesses to data, and control over physical file layout on storage devices. This scenario motivated the inclusion in the MPI-2 standard of a custom interface in order to support more elaborated parallel input/output operations.
The MPI specification for parallel input/output revolves around the use objects called files. As defined by MPI, files are not just contiguous byte streams. Instead, they are regarded as ordered collections of typed data items. MPI supports sequential or random access to any integral set of these items. Furthermore, files are opened collectively by a group of processes.
The common patterns for accessing a shared file (broadcast, scatter, gather, reduction) is expressed by using user-defined datatypes. Compared to the communication patterns of point-to-point and collective communications, this approach has the advantage of added flexibility and expressiveness. Data access operations (read and write) are defined for different kinds of positioning (using explicit offsets, individual file pointers, and shared file pointers), coordination (non-collective and collective), and synchronism (blocking, nonblocking, and split collective with begin/end phases).
In MPI for Python, all MPI input/output operations are performed
through instances of the File class. File handles are obtained by
calling the File.Open method at all processes within a communicator
and providing a file name and the intended access mode. After use,
they must be closed by calling the File.Close method. Files even
can be deleted by calling method File.Delete.
After creation, files are typically associated with a per-process
view. The view defines the current set of data visible and
accessible from an open file as an ordered set of elementary
datatypes. This data layout can be set and queried with the
File.Set_view and File.Get_view methods respectively.
Actual input/output operations are achieved by many methods combining read and write calls with different behavior regarding positioning, coordination, and synchronism. Summing up, MPI for Python provides the thirty (30) methods defined in MPI-2 for reading from or writing to files using explicit offsets or file pointers (individual or shared), in blocking or nonblocking and collective or noncollective versions.
Environmental Management¶
Initialization and Exit¶
Module functions Init or Init_thread and Finalize provide MPI
initialization and finalization respectively. Module functions
Is_initialized and Is_finalized provide the respective tests for
initialization and finalization.
Note
MPI_Init() or MPI_Init_thread() is actually called
when you import the MPI module from the
mpi4py package, but only if MPI is not already
initialized. In such case, calling Init or Init_thread from
Python is expected to generate an MPI error, and in turn an
exception will be raised.
Note
MPI_Finalize() is registered (by using Python C/API
function Py_AtExit()) for being automatically called when
Python processes exit, but only if mpi4py actually
initialized MPI. Therefore, there is no need to call Finalize
from Python to ensure MPI finalization.
Implementation Information¶
The MPI version number can be retrieved from module function
Get_version. It returns a two-integer tuple(version, subversion).The
Get_processor_namefunction can be used to access the processor name.The values of predefined attributes attached to the world communicator can be obtained by calling the
Comm.Get_attrmethod within theCOMM_WORLDinstance.
Timers¶
MPI timer functionalities are available through the Wtime and
Wtick functions.
Error Handling¶
In order facilitate handle sharing with other Python modules
interfacing MPI-based parallel libraries, the predefined MPI error
handlers ERRORS_RETURN and ERRORS_ARE_FATAL can be assigned to and
retrieved from communicators using methods Comm.Set_errhandler and
Comm.Get_errhandler, and similarly for windows and files.
When the predefined error handler ERRORS_RETURN is set, errors
returned from MPI calls within Python code will raise an instance of
the exception class Exception, which is a subclass of the standard
Python exception RuntimeError.
Note
After import, mpi4py overrides the default MPI rules governing
inheritance of error handlers. The ERRORS_RETURN error handler is
set in the predefined COMM_SELF and COMM_WORLD communicators,
as well as any new Comm, Win, or File instance created
through mpi4py. If you ever pass such handles to C/C++/Fortran
library code, it is recommended to set the ERRORS_ARE_FATAL error
handler on them to ensure MPI errors do not pass silently.
Warning
Importing with from mpi4py.MPI import * will cause a name
clashing with the standard Python Exception base class.