Thanks., good ideas. I'm considering mentoring again, BSP-based mat-mult and BSP-based graph partitioning (clustering) seems good candidates.
On Tue, Feb 28, 2012 at 1:16 AM, Thomas Jungblut <[email protected]> wrote: > Hi Edward, > > I'm not directly interested in metoring, however we should start slicing > some issues. > Since we always advertise that we are good for matrix computations we > should start considering some issues regarding it. > e.g: > * Matrix inversion > * Matrix multiplication (HAMA-220) > * Matrix transposition (ideally should be part of multiplication, e.G. like > in the BLAS implementations C = alpha * op(A) * op(B) + beta * C, where A > and B should be transposable). > > For fault tolerance I guess that HAMA-509 (Make BSPMaster fault tolerant) > would be a good GSoC project. > > I would be glad if I can work on one of the issues during GSoC, however I'd > be an unofficial mentor anyways. > > Best Regards, > Thomas. > > 2012/2/27 Edward J. Yoon <[email protected]> > >> FWD, >> >> Anyone interesting in GSoC 2012 mentoring? >> >> >> ---------- Forwarded message ---------- >> From: Andrei Savu <[email protected]> >> Date: Mon, Feb 27, 2012 at 9:11 PM >> Subject: Re: GSoC >> To: [email protected] >> >> >> Here is a short list of relevant issues: >> http://s.apache.org/apache-whirr-gsoc-2012 >> >> What do you think? >> >> On Mon, Feb 27, 2012 at 7:08 PM, Andrei Savu <[email protected]> >> wrote: >> >> > Hi, >> > >> > Today the mentoring organisations can begin submitting applications to >> > Google. >> > >> > Is there anything we need to do? I will look into updating the issue >> > tracker as needed. >> > >> > -- Andrei Savu / andreisavu.ro >> > >> > >> >> >> -- >> Best Regards, Edward J. Yoon >> @eddieyoon >> > > > > -- > Thomas Jungblut > Berlin <[email protected]> -- Best Regards, Edward J. Yoon @eddieyoon
