INTL 3:770/1 3:770/3
REPLYADDR dennis@none.none
REPLYTO 3:770/3.0 UUCP
MSGID:  848b0ddb
REPLY:  5d72408a
PID: SoupGate-Win32 v1.05
On 12/24/24 09:35, Carl Fink wrote:
> I'm interested in playing with this, but I'm not interested in paying $3000
> for a high-end nVidia NPU. It seems to be in principle possible to split the
> task into smaller subtasks that could be distributed to a Beowulf-style Pi
> cluster. Several enclosures exist that let you connect multiple Compute
> Modules over a high-speed bus, I hear.
>
> I have minimal experience with Raspberry Pi (not zero, but minimal). I have
> none with setting up Beowulf clusters, and none with decomposing machine
> learning tasks and distributing them among processors. Thus, I wonder if
> there might be an existing project I could learn from and maybe even
> eventually contribute to, even if only as a tester.
>
> Thanks.
There were some projects that used multiple PIs. From a cost and
complexity standpoint they were more "because we can" rather than
practical. A multicore AMD or INTEL processor would be a better option.
If you skip the high power GPU the system cost is lower. Also the MPI
cluster code is "off the shelf" for those processors.

https://mpitutorial.com/tutorials/mpi-hello-world/

--- SoupGate-Win32 v1.05
 * Origin: Agency HUB, Dunedin - New Zealand | Fido<>Usenet Gateway (3:770/3)
SEEN-BY: 4/0 19/10 88/0 90/0 105/81 106/201 128/187 129/305 153/757
SEEN-BY: 153/7715 218/700 840 220/70 221/1 6 360 226/17 30 100 227/114
SEEN-BY: 229/110 111 114 200 206 300 317 400 426 428 470 550 616 664
SEEN-BY: 229/700 705 266/512 267/800 282/1038 291/111 292/854 301/1
SEEN-BY: 310/31 320/219 322/757 335/364 341/66 342/200 396/45 460/58
SEEN-BY: 633/280 712/848 770/1 3 100 330 340 772/210 220 230 880/1
SEEN-BY: 900/0 102 106 902/0 6 19 26 905/0 930/1 2320/105 5020/400
SEEN-BY: 5075/35
PATH: 770/3 1 218/840 221/6 341/66 902/26 229/426