15 Jan 2020 |
swedishhat | I was thinking that the distributed datacenter part could be based on IPFS so that neighbors could peer the internet to get faster speeds and lower latency | 18:05:13 |
swedishhat | There could also be apps for phones, desktops, and personal clouds (Synology, Unraid, WD MyCloud, etc.) that further subsidize the cost of customer internet access by providing more storage and compute. | 18:06:33 |
swedishhat | You could have really good georedundancy | 18:08:20 |
31 Jan 2020 |
di0_thedestroyer | oi. it's a new year. i fizzled out a bit last year. i'm going to be careful about creating contracts this year, but my new cluster platform based on Rust is coming along. Also, the comphaq.com build has made a tiny amount of progress. | 09:56:43 |
di0_thedestroyer | i've wondered whether my ideal platform for clusterfights is more of a 24/7 event year-round, with maybe quarterly events | 09:57:34 |
di0_thedestroyer | trust me, i'm going to keep building clusters that brandonb can spank in 2 days, but i'm just worried about my commitments when my funded projects need work during the year. I'm not quite as fast at coming up with cool builds as some competitors will be. I think we need a platform that accomodates competitors building at any speed. If they want to submit logs once per month, or once per hour -- I think that danielle.wis 's idea of having that constant competition is key. Beyond that, a standard system of regionals and nationals could be held once per year, as per other sporting events. | 10:01:46 |
19 Feb 2020 |
@gnarlsmarley:matrix.org | https://psteinb.github.io/hpc-in-a-day/ | 23:50:50 |
27 Feb 2020 |
@danielle.wis:matrix.org | Redacted or Malformed Event | 04:21:21 |
@danielle.wis:matrix.org | Redacted or Malformed Event | 04:23:14 |
@danielle.wis:matrix.org | Redacted or Malformed Event | 04:23:24 |
3 Mar 2020 |
swedishhat | https://web.stanford.edu/group/mast/cgi-bin/drupal/system/files/isca2017-prabhakar.pdf | 07:20:36 |
| CouldBeARobot changed their display name from jeff_comstockisthebest to CouldBeARobot. | 21:10:25 |
7 Mar 2020 |
bobsmith-dpi | How a cluster can beat a GPU in NN learning:
https://www.sciencedaily.com/releases/2020/03/200305135041.htm | 02:42:49 |
@gnarlsmarley:matrix.org | https://www.allaboutcircuits.com/news/founder-of-fpgas-swooping-in-with-comprehensive-smartnic/ | 13:07:32 |
8 Apr 2020 |
bobsmith-dpi | KiCad does Altium?????
https://www.cnx-software.com/2020/04/05/how-to-build-kicad-on-ubuntu-18-04-and-import-altium-pcb-files/ | 03:39:51 |
15 Apr 2020 |
swedishhat | That's pretty awesome | 02:03:01 |
swedishhat | I bet I could start using that code as a basis for doing library extraction | 02:03:28 |
swedishhat | It looks like they are focusing on the whole PCB at once, pulling out all the altium primitives, rather than treating it like components connected together with primitives | 02:04:30 |
18 Apr 2020 |
@gnarlsmarley:matrix.org | Probably there had been others, but this article about our friend @turingpi:matrix.org just came across my feed. https://betanews.com/2020/04/16/raspberry-pi-edge-server/ | 03:01:44 |
swedishhat | I saw that article too! It's awesome to see him getting some news coverage | 03:49:46 |
25 Apr 2020 |
swedishhat | https://www.dinofizzotti.com/blog/2020-04-10-raspberry-pi-cluster-part-1-provisioning-with-ansible-and-temperature-monitoring-using-prometheus-and-grafana/ | 06:48:17 |
swedishhat | this cluster was featured on hackaday | 06:48:59 |
27 May 2020 |
| @gnarlsmarley:matrix.org left the room. | 07:13:02 |
7 Jul 2020 |
| @lillianna7359:matrix.org joined the room. | 17:31:43 |
8 Jul 2020 |
| @lillianna7359:matrix.org left the room. | 00:36:06 |
16 Jan 2021 |
| @danielle.wis:matrix.org joined the room. | 22:57:54 |
| @danielle.wis:matrix.org joined the room. | 22:58:06 |
| @danielle.wis:matrix.org left the room. | 22:58:08 |
7 May 2021 |
| @amon_morgan:matrix.org joined the room. | 12:30:42 |
| @amon_morgan:matrix.org left the room. | 12:31:24 |