December 2015 - discuss - mailman3.cels.anl.gov

Re: [mpich-discuss] mpi_scatter
by Jen B 04 Dec '15

04 Dec '15

Thanks. Best, Jen On Fri, Dec 4, 2015 at 9:08 AM, Thakur, Rajeev <thakur(a)mcs.anl.gov> wrote: > Even with MPICH2, there were several releases of the code at which things > can change since the publication of the paper. > > Rajeev > > > On Dec 3, 2015, at 4:26 PM, Jen B <jenbna(a)gmail.com> wrote: > > > > Thanks, Rajeev. Last question, by old do you mean MPICH2? > > > > Best, > > Jen > > > > On Thu, Dec 3, 2015 at 3:29 PM, Thakur, Rajeev <thakur(a)mcs.anl.gov> > wrote: > > Scatter uses a binomial tree algorithm. Scatterv uses a linear algorithm. > > > > The paper is old. See the comments in the bcast.c file for what is > currently used. > > > > > > > On Dec 3, 2015, at 3:07 PM, Jen B <jenbna(a)gmail.com> wrote: > > > > > > Thanks all. So it uses linear algorithm with send/recv. > > > > > > For my second question, i.e., mpi_bcast for long messages the paper > says scatter+allgather. In the case when the message size is not evenly > divisible by number of processes, I guess it must be using > scatter+allgatherv? > > > > > > Best, > > > Jen > > > > > > > > > On Thu, Dec 3, 2015 at 2:50 PM, Thakur, Rajeev <thakur(a)mcs.anl.gov> > wrote: > > > See the comments in the scatter.c file. > > > > > > Rajeev > > > > > > > On Dec 3, 2015, at 2:47 PM, Jen B <jenbna(a)gmail.com> wrote: > > > > > > > > Here is the sample of information that I need. For example, in that > paper for long messages (>=512KB) and any number of processes "allgather" > uses the ring algorithm. Can anyone give me similar answer (algorithm used > in MPICH2) for mpi_scatter (when we have long messages, and for both power > of two and non power of two number of processes)? Thanks. > > > > > > > > Best, > > > > Jen > > > > > > > > On Thu, Dec 3, 2015 at 2:25 PM, Jeff Hammond <jeff.science(a)gmail.com> > wrote: > > > > I don't know exactly what information you want, but you will find a > lot of useful comments in the source files here: > http://git.mpich.org/mpich.git/tree/HEAD:/src/mpi/coll. In particular, > if you run 'grep -i "we use" *.c', you will see statements about > Rabenseifer's algorithm, Bruck's algorithm, recursive doubling, pairwise > exchange, etc. > > > > > > > > Best, > > > > > > > > Jeff > > > > > > > > On Thu, Dec 3, 2015 at 8:24 AM, Jen B <jenbna(a)gmail.com> wrote: > > > > Hello, > > > > > > > > I am trying to understand how the mpi_scatter is implemented > according to MPICH2 for long messages. I went through the paper of R. > Thakur, et al., titled "Optimization of Collective Communication Operations > in MPICH." It is stated there that for long messages mpi_bcast is composed > of scatter and allgather (where latter two are based on the Van de Geijn's > algorithm). However I wasn't able to find the mpi_scatter details. Could > you please explain this? Also, the another thing is that I think for long > messages (that are not divisible exactly by the number of processes) this > part would use a scatterv + allgatherv. Am I right? Thank you! > > > > > > > > Best, > > > > Jen > > > > > > > > > > > > > > > > > > > > > > > > _______________________________________________ > > > > discuss mailing list discuss(a)mpich.org > > > > To manage subscription options or unsubscribe: > > > > https://lists.mpich.org/mailman/listinfo/discuss > > > > > > > > > > > > > > > > -- > > > > Jeff Hammond > > > > jeff.science(a)gmail.com > > > > http://jeffhammond.github.io/ > > > > > > > > _______________________________________________ > > > > discuss mailing list discuss(a)mpich.org > > > > To manage subscription options or unsubscribe: > > > > https://lists.mpich.org/mailman/listinfo/discuss > > > > > > > > _______________________________________________ > > > > discuss mailing list discuss(a)mpich.org > > > > To manage subscription options or unsubscribe: > > > > https://lists.mpich.org/mailman/listinfo/discuss > > > > > > _______________________________________________ > > > discuss mailing list discuss(a)mpich.org > > > To manage subscription options or unsubscribe: > > > https://lists.mpich.org/mailman/listinfo/discuss > > > > > > _______________________________________________ > > > discuss mailing list discuss(a)mpich.org > > > To manage subscription options or unsubscribe: > > > https://lists.mpich.org/mailman/listinfo/discuss > > > > _______________________________________________ > > discuss mailing list discuss(a)mpich.org > > To manage subscription options or unsubscribe: > > https://lists.mpich.org/mailman/listinfo/discuss > > > > _______________________________________________ > > discuss mailing list discuss(a)mpich.org > > To manage subscription options or unsubscribe: > > https://lists.mpich.org/mailman/listinfo/discuss > > _______________________________________________ > discuss mailing list discuss(a)mpich.org > To manage subscription options or unsubscribe: > https://lists.mpich.org/mailman/listinfo/discuss > _______________________________________________ discuss mailing list discuss(a)mpich.org To manage subscription options or unsubscribe: https://lists.mpich.org/mailman/listinfo/discuss

1 0

Re: [mpich-discuss] mpi_scatter
by Thakur, Rajeev 04 Dec '15

04 Dec '15

Even with MPICH2, there were several releases of the code at which things can change since the publication of the paper. Rajeev > On Dec 3, 2015, at 4:26 PM, Jen B <jenbna(a)gmail.com> wrote: > > Thanks, Rajeev. Last question, by old do you mean MPICH2? > > Best, > Jen > > On Thu, Dec 3, 2015 at 3:29 PM, Thakur, Rajeev <thakur(a)mcs.anl.gov> wrote: > Scatter uses a binomial tree algorithm. Scatterv uses a linear algorithm. > > The paper is old. See the comments in the bcast.c file for what is currently used. > > > > On Dec 3, 2015, at 3:07 PM, Jen B <jenbna(a)gmail.com> wrote: > > > > Thanks all. So it uses linear algorithm with send/recv. > > > > For my second question, i.e., mpi_bcast for long messages the paper says scatter+allgather. In the case when the message size is not evenly divisible by number of processes, I guess it must be using scatter+allgatherv? > > > > Best, > > Jen > > > > > > On Thu, Dec 3, 2015 at 2:50 PM, Thakur, Rajeev <thakur(a)mcs.anl.gov> wrote: > > See the comments in the scatter.c file. > > > > Rajeev > > > > > On Dec 3, 2015, at 2:47 PM, Jen B <jenbna(a)gmail.com> wrote: > > > > > > Here is the sample of information that I need. For example, in that paper for long messages (>=512KB) and any number of processes "allgather" uses the ring algorithm. Can anyone give me similar answer (algorithm used in MPICH2) for mpi_scatter (when we have long messages, and for both power of two and non power of two number of processes)? Thanks. > > > > > > Best, > > > Jen > > > > > > On Thu, Dec 3, 2015 at 2:25 PM, Jeff Hammond <jeff.science(a)gmail.com> wrote: > > > I don't know exactly what information you want, but you will find a lot of useful comments in the source files here: http://git.mpich.org/mpich.git/tree/HEAD:/src/mpi/coll. In particular, if you run 'grep -i "we use" *.c', you will see statements about Rabenseifer's algorithm, Bruck's algorithm, recursive doubling, pairwise exchange, etc. > > > > > > Best, > > > > > > Jeff > > > > > > On Thu, Dec 3, 2015 at 8:24 AM, Jen B <jenbna(a)gmail.com> wrote: > > > Hello, > > > > > > I am trying to understand how the mpi_scatter is implemented according to MPICH2 for long messages. I went through the paper of R. Thakur, et al., titled "Optimization of Collective Communication Operations in MPICH." It is stated there that for long messages mpi_bcast is composed of scatter and allgather (where latter two are based on the Van de Geijn's algorithm). However I wasn't able to find the mpi_scatter details. Could you please explain this? Also, the another thing is that I think for long messages (that are not divisible exactly by the number of processes) this part would use a scatterv + allgatherv. Am I right? Thank you! > > > > > > Best, > > > Jen > > > > > > > > > > > > > > > > > > _______________________________________________ > > > discuss mailing list discuss(a)mpich.org > > > To manage subscription options or unsubscribe: > > > https://lists.mpich.org/mailman/listinfo/discuss > > > > > > > > > > > > -- > > > Jeff Hammond > > > jeff.science(a)gmail.com > > > http://jeffhammond.github.io/ > > > > > > _______________________________________________ > > > discuss mailing list discuss(a)mpich.org > > > To manage subscription options or unsubscribe: > > > https://lists.mpich.org/mailman/listinfo/discuss > > > > > > _______________________________________________ > > > discuss mailing list discuss(a)mpich.org > > > To manage subscription options or unsubscribe: > > > https://lists.mpich.org/mailman/listinfo/discuss > > > > _______________________________________________ > > discuss mailing list discuss(a)mpich.org > > To manage subscription options or unsubscribe: > > https://lists.mpich.org/mailman/listinfo/discuss > > > > _______________________________________________ > > discuss mailing list discuss(a)mpich.org > > To manage subscription options or unsubscribe: > > https://lists.mpich.org/mailman/listinfo/discuss > > _______________________________________________ > discuss mailing list discuss(a)mpich.org > To manage subscription options or unsubscribe: > https://lists.mpich.org/mailman/listinfo/discuss > > _______________________________________________ > discuss mailing list discuss(a)mpich.org > To manage subscription options or unsubscribe: > https://lists.mpich.org/mailman/listinfo/discuss _______________________________________________ discuss mailing list discuss(a)mpich.org To manage subscription options or unsubscribe: https://lists.mpich.org/mailman/listinfo/discuss

1 0

Re: [mpich-discuss] Redirected links on http://www.mcs.anl.gov/research/projects/perfvis/download/
by Schilling, Herbert W. (GRC-VE00) 03 Dec '15

03 Dec '15

Thanks On Dec 3, 2015, at 5:45 PM, Kenneth Raffenetti <raffenet(a)mcs.anl.gov<mailto:[email protected]>> wrote: Thanks for the report. I went through and cleaned up the download page as best I could. Ken On 12/03/2015 12:44 PM, Schilling, Herbert W. (GRC-VE00) wrote: Hi, Some of the links on http://www.mcs.anl.gov/research/projects/perfvis/download/ ( like the links to the User Guide ) get redirected to the MPICH home page. Which might not be too bad but I can't find the MPE Users Guide there even when using the Search. I think I did find it so I am OK, but I thought you should know! ftp://ftp.mcs.anl.gov/pub/mpi/mpeman.pdf Thanks --------- Herb Schilling Scientific Applications and Visualization Team http://ocio.grc.nasa.gov/gvis Office of the Chief Information Officer NASA Glenn Research Center hschilling(a)nasa.gov <mailto:[email protected]> _______________________________________________ discuss mailing list discuss(a)mpich.org To manage subscription options or unsubscribe: https://lists.mpich.org/mailman/listinfo/discuss _______________________________________________ discuss mailing list discuss(a)mpich.org<mailto:[email protected]> To manage subscription options or unsubscribe: https://lists.mpich.org/mailman/listinfo/discuss --------- Herb Schilling Scientific Applications and Visualization Team http://ocio.grc.nasa.gov/gvis Office of the Chief Information Officer NASA Glenn Research Center hschilling(a)nasa.gov<mailto:[email protected]> _______________________________________________ discuss mailing list discuss(a)mpich.org To manage subscription options or unsubscribe: https://lists.mpich.org/mailman/listinfo/discuss

1 0

Re: [mpich-discuss] Redirected links on http://www.mcs.anl.gov/research/projects/perfvis/download/
by Kenneth Raffenetti 03 Dec '15

03 Dec '15

Thanks for the report. I went through and cleaned up the download page as best I could. Ken On 12/03/2015 12:44 PM, Schilling, Herbert W. (GRC-VE00) wrote: > Hi, > > Some of the links on > > http://www.mcs.anl.gov/research/projects/perfvis/download/ > > ( like the links to the User Guide ) > > get redirected to the MPICH home page. Which might not be too bad but I > can't find the MPE Users Guide there even when using the Search. > > I think I did find it so I am OK, but I thought you should know! > > ftp://ftp.mcs.anl.gov/pub/mpi/mpeman.pdf > > Thanks > > > --------- > Herb Schilling > Scientific Applications and Visualization Team > http://ocio.grc.nasa.gov/gvis > Office of the Chief Information Officer > NASA Glenn Research Center > hschilling(a)nasa.gov <mailto:[email protected]> > > > > > > > _______________________________________________ > discuss mailing list discuss(a)mpich.org > To manage subscription options or unsubscribe: > https://lists.mpich.org/mailman/listinfo/discuss > _______________________________________________ discuss mailing list discuss(a)mpich.org To manage subscription options or unsubscribe: https://lists.mpich.org/mailman/listinfo/discuss

1 0

Re: [mpich-discuss] mpi_scatter
by Jen B 03 Dec '15

03 Dec '15

Thanks, Rajeev. Last question, by old do you mean MPICH2? Best, Jen On Thu, Dec 3, 2015 at 3:29 PM, Thakur, Rajeev <thakur(a)mcs.anl.gov> wrote: > Scatter uses a binomial tree algorithm. Scatterv uses a linear algorithm. > > The paper is old. See the comments in the bcast.c file for what is > currently used. > > > > On Dec 3, 2015, at 3:07 PM, Jen B <jenbna(a)gmail.com> wrote: > > > > Thanks all. So it uses linear algorithm with send/recv. > > > > For my second question, i.e., mpi_bcast for long messages the paper says > scatter+allgather. In the case when the message size is not evenly > divisible by number of processes, I guess it must be using > scatter+allgatherv? > > > > Best, > > Jen > > > > > > On Thu, Dec 3, 2015 at 2:50 PM, Thakur, Rajeev <thakur(a)mcs.anl.gov> > wrote: > > See the comments in the scatter.c file. > > > > Rajeev > > > > > On Dec 3, 2015, at 2:47 PM, Jen B <jenbna(a)gmail.com> wrote: > > > > > > Here is the sample of information that I need. For example, in that > paper for long messages (>=512KB) and any number of processes "allgather" > uses the ring algorithm. Can anyone give me similar answer (algorithm used > in MPICH2) for mpi_scatter (when we have long messages, and for both power > of two and non power of two number of processes)? Thanks. > > > > > > Best, > > > Jen > > > > > > On Thu, Dec 3, 2015 at 2:25 PM, Jeff Hammond <jeff.science(a)gmail.com> > wrote: > > > I don't know exactly what information you want, but you will find a > lot of useful comments in the source files here: > http://git.mpich.org/mpich.git/tree/HEAD:/src/mpi/coll. In particular, > if you run 'grep -i "we use" *.c', you will see statements about > Rabenseifer's algorithm, Bruck's algorithm, recursive doubling, pairwise > exchange, etc. > > > > > > Best, > > > > > > Jeff > > > > > > On Thu, Dec 3, 2015 at 8:24 AM, Jen B <jenbna(a)gmail.com> wrote: > > > Hello, > > > > > > I am trying to understand how the mpi_scatter is implemented according > to MPICH2 for long messages. I went through the paper of R. Thakur, et al., > titled "Optimization of Collective Communication Operations in MPICH." It > is stated there that for long messages mpi_bcast is composed of scatter and > allgather (where latter two are based on the Van de Geijn's algorithm). > However I wasn't able to find the mpi_scatter details. Could you please > explain this? Also, the another thing is that I think for long messages > (that are not divisible exactly by the number of processes) this part would > use a scatterv + allgatherv. Am I right? Thank you! > > > > > > Best, > > > Jen > > > > > > > > > > > > > > > > > > _______________________________________________ > > > discuss mailing list discuss(a)mpich.org > > > To manage subscription options or unsubscribe: > > > https://lists.mpich.org/mailman/listinfo/discuss > > > > > > > > > > > > -- > > > Jeff Hammond > > > jeff.science(a)gmail.com > > > http://jeffhammond.github.io/ > > > > > > _______________________________________________ > > > discuss mailing list discuss(a)mpich.org > > > To manage subscription options or unsubscribe: > > > https://lists.mpich.org/mailman/listinfo/discuss > > > > > > _______________________________________________ > > > discuss mailing list discuss(a)mpich.org > > > To manage subscription options or unsubscribe: > > > https://lists.mpich.org/mailman/listinfo/discuss > > > > _______________________________________________ > > discuss mailing list discuss(a)mpich.org > > To manage subscription options or unsubscribe: > > https://lists.mpich.org/mailman/listinfo/discuss > > > > _______________________________________________ > > discuss mailing list discuss(a)mpich.org > > To manage subscription options or unsubscribe: > > https://lists.mpich.org/mailman/listinfo/discuss > > _______________________________________________ > discuss mailing list discuss(a)mpich.org > To manage subscription options or unsubscribe: > https://lists.mpich.org/mailman/listinfo/discuss > _______________________________________________ discuss mailing list discuss(a)mpich.org To manage subscription options or unsubscribe: https://lists.mpich.org/mailman/listinfo/discuss

1 0

Re: [mpich-discuss] mpi_scatter
by Thakur, Rajeev 03 Dec '15

03 Dec '15

Scatter uses a binomial tree algorithm. Scatterv uses a linear algorithm. The paper is old. See the comments in the bcast.c file for what is currently used. > On Dec 3, 2015, at 3:07 PM, Jen B <jenbna(a)gmail.com> wrote: > > Thanks all. So it uses linear algorithm with send/recv. > > For my second question, i.e., mpi_bcast for long messages the paper says scatter+allgather. In the case when the message size is not evenly divisible by number of processes, I guess it must be using scatter+allgatherv? > > Best, > Jen > > > On Thu, Dec 3, 2015 at 2:50 PM, Thakur, Rajeev <thakur(a)mcs.anl.gov> wrote: > See the comments in the scatter.c file. > > Rajeev > > > On Dec 3, 2015, at 2:47 PM, Jen B <jenbna(a)gmail.com> wrote: > > > > Here is the sample of information that I need. For example, in that paper for long messages (>=512KB) and any number of processes "allgather" uses the ring algorithm. Can anyone give me similar answer (algorithm used in MPICH2) for mpi_scatter (when we have long messages, and for both power of two and non power of two number of processes)? Thanks. > > > > Best, > > Jen > > > > On Thu, Dec 3, 2015 at 2:25 PM, Jeff Hammond <jeff.science(a)gmail.com> wrote: > > I don't know exactly what information you want, but you will find a lot of useful comments in the source files here: http://git.mpich.org/mpich.git/tree/HEAD:/src/mpi/coll. In particular, if you run 'grep -i "we use" *.c', you will see statements about Rabenseifer's algorithm, Bruck's algorithm, recursive doubling, pairwise exchange, etc. > > > > Best, > > > > Jeff > > > > On Thu, Dec 3, 2015 at 8:24 AM, Jen B <jenbna(a)gmail.com> wrote: > > Hello, > > > > I am trying to understand how the mpi_scatter is implemented according to MPICH2 for long messages. I went through the paper of R. Thakur, et al., titled "Optimization of Collective Communication Operations in MPICH." It is stated there that for long messages mpi_bcast is composed of scatter and allgather (where latter two are based on the Van de Geijn's algorithm). However I wasn't able to find the mpi_scatter details. Could you please explain this? Also, the another thing is that I think for long messages (that are not divisible exactly by the number of processes) this part would use a scatterv + allgatherv. Am I right? Thank you! > > > > Best, > > Jen > > > > > > > > > > > > _______________________________________________ > > discuss mailing list discuss(a)mpich.org > > To manage subscription options or unsubscribe: > > https://lists.mpich.org/mailman/listinfo/discuss > > > > > > > > -- > > Jeff Hammond > > jeff.science(a)gmail.com > > http://jeffhammond.github.io/ > > > > _______________________________________________ > > discuss mailing list discuss(a)mpich.org > > To manage subscription options or unsubscribe: > > https://lists.mpich.org/mailman/listinfo/discuss > > > > _______________________________________________ > > discuss mailing list discuss(a)mpich.org > > To manage subscription options or unsubscribe: > > https://lists.mpich.org/mailman/listinfo/discuss > > _______________________________________________ > discuss mailing list discuss(a)mpich.org > To manage subscription options or unsubscribe: > https://lists.mpich.org/mailman/listinfo/discuss > > _______________________________________________ > discuss mailing list discuss(a)mpich.org > To manage subscription options or unsubscribe: > https://lists.mpich.org/mailman/listinfo/discuss _______________________________________________ discuss mailing list discuss(a)mpich.org To manage subscription options or unsubscribe: https://lists.mpich.org/mailman/listinfo/discuss

1 0

Re: [mpich-discuss] mpi_scatter
by Jen B 03 Dec '15

03 Dec '15

Thanks all. So it uses linear algorithm with send/recv. For my second question, i.e., mpi_bcast for long messages the paper says scatter+allgather. In the case when the message size is not evenly divisible by number of processes, I guess it must be using scatter+allgatherv? Best, Jen On Thu, Dec 3, 2015 at 2:50 PM, Thakur, Rajeev <thakur(a)mcs.anl.gov> wrote: > See the comments in the scatter.c file. > > Rajeev > > > On Dec 3, 2015, at 2:47 PM, Jen B <jenbna(a)gmail.com> wrote: > > > > Here is the sample of information that I need. For example, in that > paper for long messages (>=512KB) and any number of processes "allgather" > uses the ring algorithm. Can anyone give me similar answer (algorithm used > in MPICH2) for mpi_scatter (when we have long messages, and for both power > of two and non power of two number of processes)? Thanks. > > > > Best, > > Jen > > > > On Thu, Dec 3, 2015 at 2:25 PM, Jeff Hammond <jeff.science(a)gmail.com> > wrote: > > I don't know exactly what information you want, but you will find a lot > of useful comments in the source files here: > http://git.mpich.org/mpich.git/tree/HEAD:/src/mpi/coll. In particular, > if you run 'grep -i "we use" *.c', you will see statements about > Rabenseifer's algorithm, Bruck's algorithm, recursive doubling, pairwise > exchange, etc. > > > > Best, > > > > Jeff > > > > On Thu, Dec 3, 2015 at 8:24 AM, Jen B <jenbna(a)gmail.com> wrote: > > Hello, > > > > I am trying to understand how the mpi_scatter is implemented according > to MPICH2 for long messages. I went through the paper of R. Thakur, et al., > titled "Optimization of Collective Communication Operations in MPICH." It > is stated there that for long messages mpi_bcast is composed of scatter and > allgather (where latter two are based on the Van de Geijn's algorithm). > However I wasn't able to find the mpi_scatter details. Could you please > explain this? Also, the another thing is that I think for long messages > (that are not divisible exactly by the number of processes) this part would > use a scatterv + allgatherv. Am I right? Thank you! > > > > Best, > > Jen > > > > > > > > > > > > _______________________________________________ > > discuss mailing list discuss(a)mpich.org > > To manage subscription options or unsubscribe: > > https://lists.mpich.org/mailman/listinfo/discuss > > > > > > > > -- > > Jeff Hammond > > jeff.science(a)gmail.com > > http://jeffhammond.github.io/ > > > > _______________________________________________ > > discuss mailing list discuss(a)mpich.org > > To manage subscription options or unsubscribe: > > https://lists.mpich.org/mailman/listinfo/discuss > > > > _______________________________________________ > > discuss mailing list discuss(a)mpich.org > > To manage subscription options or unsubscribe: > > https://lists.mpich.org/mailman/listinfo/discuss > > _______________________________________________ > discuss mailing list discuss(a)mpich.org > To manage subscription options or unsubscribe: > https://lists.mpich.org/mailman/listinfo/discuss > _______________________________________________ discuss mailing list discuss(a)mpich.org To manage subscription options or unsubscribe: https://lists.mpich.org/mailman/listinfo/discuss

1 0

Re: [mpich-discuss] mpi_scatter
by Thakur, Rajeev 03 Dec '15

03 Dec '15

See the comments in the scatter.c file. Rajeev > On Dec 3, 2015, at 2:47 PM, Jen B <jenbna(a)gmail.com> wrote: > > Here is the sample of information that I need. For example, in that paper for long messages (>=512KB) and any number of processes "allgather" uses the ring algorithm. Can anyone give me similar answer (algorithm used in MPICH2) for mpi_scatter (when we have long messages, and for both power of two and non power of two number of processes)? Thanks. > > Best, > Jen > > On Thu, Dec 3, 2015 at 2:25 PM, Jeff Hammond <jeff.science(a)gmail.com> wrote: > I don't know exactly what information you want, but you will find a lot of useful comments in the source files here: http://git.mpich.org/mpich.git/tree/HEAD:/src/mpi/coll. In particular, if you run 'grep -i "we use" *.c', you will see statements about Rabenseifer's algorithm, Bruck's algorithm, recursive doubling, pairwise exchange, etc. > > Best, > > Jeff > > On Thu, Dec 3, 2015 at 8:24 AM, Jen B <jenbna(a)gmail.com> wrote: > Hello, > > I am trying to understand how the mpi_scatter is implemented according to MPICH2 for long messages. I went through the paper of R. Thakur, et al., titled "Optimization of Collective Communication Operations in MPICH." It is stated there that for long messages mpi_bcast is composed of scatter and allgather (where latter two are based on the Van de Geijn's algorithm). However I wasn't able to find the mpi_scatter details. Could you please explain this? Also, the another thing is that I think for long messages (that are not divisible exactly by the number of processes) this part would use a scatterv + allgatherv. Am I right? Thank you! > > Best, > Jen > > > > > > _______________________________________________ > discuss mailing list discuss(a)mpich.org > To manage subscription options or unsubscribe: > https://lists.mpich.org/mailman/listinfo/discuss > > > > -- > Jeff Hammond > jeff.science(a)gmail.com > http://jeffhammond.github.io/ > > _______________________________________________ > discuss mailing list discuss(a)mpich.org > To manage subscription options or unsubscribe: > https://lists.mpich.org/mailman/listinfo/discuss > > _______________________________________________ > discuss mailing list discuss(a)mpich.org > To manage subscription options or unsubscribe: > https://lists.mpich.org/mailman/listinfo/discuss _______________________________________________ discuss mailing list discuss(a)mpich.org To manage subscription options or unsubscribe: https://lists.mpich.org/mailman/listinfo/discuss

1 0

Re: [mpich-discuss] mpi_scatter
by Jen B 03 Dec '15

03 Dec '15

Here is the sample of information that I need. For example, in that paper for long messages (>=512KB) and any number of processes "allgather" uses the ring algorithm. Can anyone give me similar answer (algorithm used in MPICH2) for mpi_scatter (when we have long messages, and for both power of two and non power of two number of processes)? Thanks. Best, Jen On Thu, Dec 3, 2015 at 2:25 PM, Jeff Hammond <jeff.science(a)gmail.com> wrote: > I don't know exactly what information you want, but you will find a lot of > useful comments in the source files here: > http://git.mpich.org/mpich.git/tree/HEAD:/src/mpi/coll. In particular, > if you run 'grep -i "we use" *.c', you will see statements about > Rabenseifer's algorithm, Bruck's algorithm, recursive doubling, pairwise > exchange, etc. > > Best, > > Jeff > > On Thu, Dec 3, 2015 at 8:24 AM, Jen B <jenbna(a)gmail.com> wrote: > >> Hello, >> >> I am trying to understand how the mpi_scatter is implemented according to >> MPICH2 for long messages. I went through the paper of R. Thakur, et al., >> titled "Optimization of Collective Communication Operations in MPICH." It >> is stated there that for long messages mpi_bcast is composed of scatter and >> allgather (where latter two are based on the Van de Geijn's algorithm). >> However I wasn't able to find the mpi_scatter details. Could you please >> explain this? Also, the another thing is that I think for long messages >> (that are not divisible exactly by the number of processes) this part would >> use a scatterv + allgatherv. Am I right? Thank you! >> >> Best, >> Jen >> >> >> >> >> >> _______________________________________________ >> discuss mailing list discuss(a)mpich.org >> To manage subscription options or unsubscribe: >> https://lists.mpich.org/mailman/listinfo/discuss >> > > > > -- > Jeff Hammond > jeff.science(a)gmail.com > http://jeffhammond.github.io/ > > _______________________________________________ > discuss mailing list discuss(a)mpich.org > To manage subscription options or unsubscribe: > https://lists.mpich.org/mailman/listinfo/discuss > _______________________________________________ discuss mailing list discuss(a)mpich.org To manage subscription options or unsubscribe: https://lists.mpich.org/mailman/listinfo/discuss

1 0

Re: [mpich-discuss] mpi_scatter
by Jeff Hammond 03 Dec '15

03 Dec '15

I don't know exactly what information you want, but you will find a lot of useful comments in the source files here: http://git.mpich.org/mpich.git/tree/HEAD:/src/mpi/coll. In particular, if you run 'grep -i "we use" *.c', you will see statements about Rabenseifer's algorithm, Bruck's algorithm, recursive doubling, pairwise exchange, etc. Best, Jeff On Thu, Dec 3, 2015 at 8:24 AM, Jen B <jenbna(a)gmail.com> wrote: > Hello, > > I am trying to understand how the mpi_scatter is implemented according to > MPICH2 for long messages. I went through the paper of R. Thakur, et al., > titled "Optimization of Collective Communication Operations in MPICH." It > is stated there that for long messages mpi_bcast is composed of scatter and > allgather (where latter two are based on the Van de Geijn's algorithm). > However I wasn't able to find the mpi_scatter details. Could you please > explain this? Also, the another thing is that I think for long messages > (that are not divisible exactly by the number of processes) this part would > use a scatterv + allgatherv. Am I right? Thank you! > > Best, > Jen > > > > > > _______________________________________________ > discuss mailing list discuss(a)mpich.org > To manage subscription options or unsubscribe: > https://lists.mpich.org/mailman/listinfo/discuss > -- Jeff Hammond jeff.science(a)gmail.com http://jeffhammond.github.io/ _______________________________________________ discuss mailing list discuss(a)mpich.org To manage subscription options or unsubscribe: https://lists.mpich.org/mailman/listinfo/discuss

1 0