![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
tcpipI've returned from a journey to The Western Lands, as I used to call WA in reference to William S. Burroughs. There are some similarities, of course, Perth being this oasis-like dot among the deserts of WA, although it is perhaps a little unfair to compare to the ancient Egyptian Land of the Dead. It was certainly anything but during my brief visit, although most of that was taken up by attending the HPC-AI Advisory Council conference. Particularly impressed with the scope of work that DK Panda is carrying out for MVAPICH2, and especially in making MPI aware of GPUs; could be quite handy. My own presentation on the International HPC Certification Forum was an elaboration to the one given at the ARDC summit a month prior, and seem to go across reasonably well. Hopefully, I'll get a few more people involved in the project and contributing. The visit to the Pawsey Supercomputing Centre was quite enjoyable as well; they have been strong backers of the project.
A particular highlight of the trip was dinner on Wednesday night at Bistro Bellavista in East Perth, my old stomping ground in that somewhat troubled 17th year of my life. Despite my appallingly short notice (I started contacting people on Sunday night!), and the fact that a few people slipped off the list altogether, some 20+ people still found it worthwhile to come out for the dinner, including some individuals whom I literally hadn't seen for decades. In addition to this, caught up with Arnold and Cathy H on my first night in Fremantle who were also quite remarkable in accepting my late notice, and treated me to an impressive Thai meal. In both cases, they were nights of good company, good cheer, good food, and quality conversation.
The final day in Perth was a trip out to the Australian Resources Research Centre (ARRC) to attend a GPU programming workshop organised by the Pawsey Supercomputing Centre. The workshop led by Rob Faber, an author on OpenACC and CUDA, and the workshop provided plenty of good examples to work with (and which credited permission was granted to use in my own training). On my return to Perth, wrote an extensive report on the conference for work, including a few items of importance concerning industry some trends that I picked up on the grapevine. Immediate tasks will involve however more optimising on our installs of MVAPICH2, building more Singularity containers, and working on an advanced GPU programming course. I think next week I'll be off to Canberra for the Challenges in High Performance Computing conference in Canberra, where I have a paper on some of the mathematical problems in HPC.
A particular highlight of the trip was dinner on Wednesday night at Bistro Bellavista in East Perth, my old stomping ground in that somewhat troubled 17th year of my life. Despite my appallingly short notice (I started contacting people on Sunday night!), and the fact that a few people slipped off the list altogether, some 20+ people still found it worthwhile to come out for the dinner, including some individuals whom I literally hadn't seen for decades. In addition to this, caught up with Arnold and Cathy H on my first night in Fremantle who were also quite remarkable in accepting my late notice, and treated me to an impressive Thai meal. In both cases, they were nights of good company, good cheer, good food, and quality conversation.
The final day in Perth was a trip out to the Australian Resources Research Centre (ARRC) to attend a GPU programming workshop organised by the Pawsey Supercomputing Centre. The workshop led by Rob Faber, an author on OpenACC and CUDA, and the workshop provided plenty of good examples to work with (and which credited permission was granted to use in my own training). On my return to Perth, wrote an extensive report on the conference for work, including a few items of importance concerning industry some trends that I picked up on the grapevine. Immediate tasks will involve however more optimising on our installs of MVAPICH2, building more Singularity containers, and working on an advanced GPU programming course. I think next week I'll be off to Canberra for the Challenges in High Performance Computing conference in Canberra, where I have a paper on some of the mathematical problems in HPC.