03:31:12 would a risc-v based computer be found there? 06:19:17 They assemble it, so yes they can find the needed components 08:22:27 have we got any stats on what kind of hashrates can be squeezed out of those RISC-V SBCs yet? I'm assuming though bitmain rigs are basically just aluminium cases packed full of them with some fans and a PSU 08:59:30 My nonce analysis showed 400-500 h/s per CPU core 09:02:20 thats not terrible :o anyway to tell what CPU was in use? guessing it wasn't the JH7110 09:26:53 xuantie c910 is a likely candidate 09:33:30 not cheap nor easy to get my hands on an SBC using that chip :/ assuming they went straight to the manufacturer. I guess buying one of those X5's could at least mean you ended up with a boatload of SBCs (if thats how they managed to jam so much hashrate into a single unit) 16:30:35 you can get those JH7110 chips in a VisionFive board pretty cheaply anyway 16:32:11 this Sipeed SBC has C910 quadcore https://www.aliexpress.com/item/1005005532736080.html 16:33:11 looks like 2x faster perf than JH7110 16:39:06 I might get one to mess around with 16:40:47 around £150gbp ... for the novelty, definitely worth a play with 16:41:16 but for a little over the price of 2 of them, could probably build a bog standard Ryzen5 3600 rig :) 16:42:34 true 16:42:42 hm, the case is out of stock. phooey 16:43:00 I can't run bare boards in a house with 2 cats 16:44:23 not ideal :P I can barely run my 7950x rig as the radiator vents out through the top of the case. Damn cat gets too comfortable 16:45:25 lol 16:46:07 I wonder if the mounting holes would line up with my Pi1 acrylic case. that thing's been offline for years now 16:47:13 was just looking at the form factor, seems pretty shallow, but quite a bit more square than the pi's 16:47:34 might be a 3d printer job 16:48:06 or, if you've bricked any android boxes, the case might be repurposed 16:50:31 12v 2a supply though, would probably switch out that jacked socket for a usb-c mount. Should be able to push 12v 2a through a decent cable 17:25:40 it'll run on 5V 2A USB-C 17:25:55 interesting review https://liliputing.com/sipeed-lichee-pi-4a-is-a-modular-risc-v-mini-pc-with-up-to-16gb-ram/ 17:26:45 note that the actual SOC & memory is on its own LM4A card 17:27:24 and see the photo near bottom: 7 slot carrier itx board 17:28:07 I think we've found our ASIC... 17:28:30 yeah... looking at that... 17:29:47 its just a carrier board with all of the peripherals ... bit like the pi compute modules I'm guessing 17:31:36 Lichee Cluster 4A - 7 slots for the LM4A 17:32:04 https://www.cnx-software.com/2022/12/14/sipeed-lm4a-t-head-th1520-risc-v-module-to-power-raspberry-pi-4-competitor-and-cluster-board/ 17:33:33 so at around 600h/s per core, 4 codes per module, 7 modules, 17kh/s-ish per board 17:33:44 s/codes/cores 17:36:47 https://sipeed.com/licheepi4a 17:37:02 "on sale next week" 17:38:26 a shame to build a cluster like that, but no hardware crypto instructions 17:38:39 I love the whole "Not going to specify which week it is... just next" method of marketing :D 17:39:40 yeah, the announcement was a couple weeks ago. oh well 17:40:03 have no doubts they can produce it tho 17:41:40 I was looking at getting a bulk buy of the rpi compute modules early last year, then stacking them up to try and build an expandable mining rig in a little itx case 17:43:35 now if I could get a bulk buy of the LM4A modules and some cluster boards, well its basically an expandable X5 :D 17:54:32 lol 17:54:39 https://rwmj.wordpress.com/2023/07/25/heads-up-lichee-pi-4a-vs-visionfive-2-vs-hifive-unmatched-vs-raspberry-pi-4b/ 17:55:26 lack of PCIe/NVMe makes it unappealing for general use 17:55:47 but for pure compute it's prob ok 17:56:43 BeagleV will have same chip 17:59:13 the BeagleV-Ahead? hmm 18:01:10 similar price to the lichee pi, however I can order from my usual electronics provider in the UK :D 18:02:33 wonder if xmrig can pick up the chip type, doesn't seem to be any any benchmarks for its hashrate 18:09:25 might grab one for shits and giggles 18:35:25 too bad beagleV only has 4GB RAM 18:47:01 looking at the global hashrates on the monero blockchain, do you think Bitmain were doing some tests in July to make sure the kit they had still worked prior to starting to sell it off? 18:47:34 spent a week over 3GH/s, then just dropped right off 18:48:06 they've been testing since a year ago December at least 18:48:07 No, when they switch it on, they keep it. This is how it happened in 2021 18:48:35 Which is logical - if the money printer is working, why turn it off? 18:49:17 unless they started work on a 400kh/s setup when some new tech came out, so time to offload the old kit 18:56:35 quite unlikely that they could justify the investment now, with tail emission and stagnant XMR price 18:56:57 in 2017 it was worthwhile to develop an XMR ASIC because both price and block reward were high 19:00:57 looking at historical price (and hashrate) https://bitinfocharts.com/comparison/hashrate-price-xmr.html#alltime 19:01:20 If they're looking for ROI profits on the short term, certainly true - been a difficult few years. Though I guess they've picked a decent enough time to release their 'ASIC' as XMR is certainly picking up on the questionable side of the internet 19:01:31 the price in 2017 was an ATH. it was comparable in 2021, but block reward was much smaller. 19:03:03 this chart is more to the point: mining profitability https://bitinfocharts.com/comparison/mining_profitability-price-xmr.html#alltime 19:03:35 you would have to be a fool to invest in the R&D now. 19:03:58 hmm, was that peak in June last year related to the Russian's offloading their FIAT into a virtual currency perhaps? :D