<unnamed>

Public Key

npub1vtrlmcawlvmctzpq6awgdh5j495pgcrwr7yw8h347rdk0z62a68sd6e50z

Profile Code

nprofile1qqsx93lauwh0kdu93qsdwhyxm6f2j6q5vphplz8rmc6lpkm83d9warcprpmhxue69uhhyetvv9ujumt0d4hhxarj9ecxjmntqyxhwumn8ghj7mn0wvhxcmmvqy28wumn8ghj7un9d3shjtnyv9kh2uewd9hsuyhgg8

Publishing to

relay.momostr.pink

nos.lol

relay.damus.io

Last Notes

2026-07-12 23:15:41 CEST

why is this article from 2013 making the rounds? did it get commercialized?
From: @npub1edu…xhkr
https://mastodon.world/@jblue/116908292371948193

2025-11-30 22:39:51 CET

If I don't make it out alive from FOSDEM... it's because of this talk:
https://mast.hpc.social/system/media_attachments/files/115/640/809/488/157/846/original/5aba866a44afe6ec.png

2025-11-30 01:03:30 CET

For my #ISA friends, a preview of what I'm submitting to #FOSDEM
#riscv
https://mast.hpc.social/system/media_attachments/files/115/635/713/346/505/794/original/8d719cf82d4ea698.png

2025-08-09 15:56:32 CEST

A CPU without a floating point unit is like a demon without its horns 😔

2025-04-22 19:40:53 CEST

Hey friends! My team is looking for a Chiplet Product Manager. Looking for somebody with ASIC PM experience who's technical, ideally who's gone from concept to launch.
Chiplet experience a bonus!
LMK if you're interested 😊
Link https://job-boards.greenhouse.io/tenstorrent/jobs/4638886007

2025-04-22 00:56:51 CEST

- reply

Was partially a pun on people in the north east US being called Yankees 😅

2025-04-22 00:21:27 CEST

you can tell that Vim was influenced by Americans; they do call copying "Yanking" after all

2025-03-04 17:25:00 CET

Really confused why Apple is launching the #iPadAir with M3? AFAIK that was an expensive chip on an expensive process, and yields weren’t exactly great?
Must be to avoid canabalizing the iPad Pro M4? (Lovely device btw)

2024-11-12 02:23:58 CET

- reply

yup, exactly what I’m getting at!

2024-11-11 22:02:47 CET

- reply

there’s very little that’s heroic in saying “we forked the main FOSS project, maintain a downstream version, and push patches upstream”
Only way that works is incentives on delivering a solution faster. But then you reward people for not creating new IP/technologies, while also not having “full control” over a load bearing portion of your stack

2024-11-11 22:00:29 CET

- reply

the first order response is “pay them in stock; the work they do must better the company”
But that falls flat when you consider that it incentivizes people to want more stock, something they achieve by moving up in the org, which is classically done by delivering a big, internal project; in other words, back to square 1.

2024-11-11 21:59:03 CET

- reply

my org seems very much directed towards the “senior people that nip it”, and as a junior who’s being trust towards senior quickly, I’m trying to learn that lesson and internalize it.
Part of what I’m grappling with is designing systems such that you always align the individual to the goals of the org, and from the org to the company;
How do you lay a foundation that fundamentally disincentivizes the individual from empire building?

2024-11-11 18:44:19 CET

- reply

Sort of have to ask; how do you align incentives across levels to minimize that sort of approach?

2024-11-11 18:41:40 CET

Found myself discussing FOSS & chiplets with a senior person.
They remarked that part of the problem at a lot of firms is that you’re incentive at all levels in most companies is to build the thing, instead of license/use the thing.
“You don’t make {vendor} fellow by licensing a {redacted} controller; you do it by building a 50 person Si design, FW & SW team, spending 50+M, then complaining that sales should license what you built”

2024-09-28 17:12:58 CEST

The efficient market theorem is just an excuse for economists to avoid talking to people as a means of gathering market information