Posts by azonenberg@ioc.exchange
(DIR) Post #B4gTRNuc72P8HJfO5Y by azonenberg@ioc.exchange
2026-03-27T12:28:09Z
0 likes, 0 repeats
Is there a way in mastodon or activitypub in general to use a hash character as a literal and *not* make the following word a hashtag (e.g. when posting about C/C++ preprocessor directives I don't necessarily intend for my post to be hashtag ifdef)
(DIR) Post #B4hPqrPXMuX0Ldz0VM by azonenberg@ioc.exchange
2026-03-26T13:59:33Z
0 likes, 0 repeats
The existence of toblerone implies the existence of related triangular hydrochocolate chains such as the base alkane (toblerane), as well as the closed ring form (cyclotoblerone)
(DIR) Post #B4iEQgEXOmHxob6zGC by azonenberg@ioc.exchange
2026-03-28T00:07:13Z
0 likes, 1 repeats
Zombie: "Braaaaains"Fedi user: "Traaaaaains"
(DIR) Post #B4rQ7G7tBiy7DQ9AtE by azonenberg@ioc.exchange
2026-04-01T20:13:17Z
0 likes, 0 repeats
@foone @whitequark I've designed or contemplated several PCBs as a result of dating people but I don't think I have ever had to write a web app lol. You must be dating interesting people.
(DIR) Post #B4sqna5y7rs1HF3A7k by azonenberg@ioc.exchange
2026-04-02T12:37:28Z
1 likes, 0 repeats
@0xabad1dea The only way it makes sense is if you already (before stealing it) have a buyer lined up who really wants it for themself, understands they'll never be able to show it off, and is willing to pay a high price.I'm sure these exist but they're rare.
(DIR) Post #B4xTfVSg8y7fQxeWu0 by azonenberg@ioc.exchange
2026-04-04T14:39:25Z
0 likes, 0 repeats
cursed image: nvidia RTX GPU but the "RTX" logo is the raytheon version
(DIR) Post #B54o2dYLaHkqNZKXz6 by azonenberg@ioc.exchange
2026-04-08T05:18:43Z
0 likes, 0 repeats
The 100baseT1 decode benchmark is dropping below real time when I have a lot of traffic on the link. Do not want.Sooo now to figure out how to either double the speed of the decode or steal some more time from something else along the critical path like the edge detector or CDR filters that I've spent a ton of time optimizing.It looks like the workload scheduler might be a weak point here: the greedy algorithm makes the eye pattern run as soon as data is available, which makes the I/Q demux (on the critical path for the decode) compete with the eye pattern for GPU time. Hypothetically if the eye pattern were later in the filter graph, it might run during the CPU-bound part of the baseT1 decode, which would start earlier and thus the overall filter graph would finish sooner. But fixing this is... decidedly nontrivial.
(DIR) Post #B54o2eJqjfkgktYU1g by azonenberg@ioc.exchange
2026-04-08T05:26:22Z
0 likes, 0 repeats
As a test, to verify the hypothesis that the eye is stealing GPU time from the demux, I deleted the eye pattern from the filter graph.Refresh rate went from about 9.8 WFM/s to 11.5, not *quite* real time (that would be 12.5) but pretty close and certainly a big improvement. But I don't want to delete it, and I have no easy way to delay it or de-prioritize it: while Vulkan queues have priority associated with them, the assignment of worker threads (and thus queues) to filter blocks is essentially random (whichever one catches the condition variable first) and thus not usable as a precedence mechanism.Messing with the scheduler also seems like a bad idea here, because any changes to tune for this particular workload unless very well tested might cause problems for other filter graphs. So I don't want any hacky special casing.
(DIR) Post #B54o2lSyAACqvCsMiG by azonenberg@ioc.exchange
2026-04-08T05:37:27Z
0 likes, 0 repeats
Also I have more margin than I thought... I left validation layers on during the previous test oops. It's still sub realtime but by a lot less: averaging around 11.5 WFM/s with the eye pattern active, and pretty comfortably real time with it removed.That explains some of the weirdness I was seeing in the NSight trace with long gaps on submits and such.So the actual overall graph run time is on the order of 76 ms, much closer to where we want to be, but I still do want to speed it up just a bit.
(DIR) Post #B55hvOW0gsCP99MPdg by azonenberg@ioc.exchange
2026-04-08T13:54:48Z
0 likes, 0 repeats
@ignaloidas That is a possibility, although "the same" workload may not be strictly true due to noise bursts, different amounts of packet activity, etc. Also consider GPU and CPU power management if you are doing single shot trigger vs continuous (single shot trigger often makes the filter graph run slower than continuous, because the GPU isn't saturated for long enough to spin up to a higher p-state).The other question is just how to even come up with concepts for different schedules. Like, the goal is to have the batch finish as quickly as possible, normally you would not expect starting a filter late to be a net benefit for this. Having a CPU-heavy filter at the end of a long linear path with GPU-heavy stuff before it is kind of a special case and I'm wondering if I might be better off just making it more GPU-heavy instead lol
(DIR) Post #B5IK7lS3Mt7fubM1yK by azonenberg@ioc.exchange
2026-04-14T19:43:21Z
0 likes, 0 repeats
@foone @alice You live an interesting life, lol. I've probably interacted with a lot more FBI agents than the average American, but the only folks with suits and radio earpieces showing up at my wedding were the groomsmen (long story)
(DIR) Post #B5ILYlV1OkCDpjBJgm by azonenberg@ioc.exchange
2026-04-14T19:59:26Z
0 likes, 0 repeats
@foone @alice The officiant at mine was carrying a tommy gun but it was a non-functional replica.It was an interesting wedding.
(DIR) Post #B5ILmeRERpkvzJ1woq by azonenberg@ioc.exchange
2026-04-14T20:01:58Z
0 likes, 0 repeats
@foone @alice Mine was the week before halloween. We invited guests to come in costume.I was dressed as a mad scientist and my wife was a witch, so of course we had a brain in a jar on the table in front of us.
(DIR) Post #B5ILpjjxBxDjIgG61A by azonenberg@ioc.exchange
2026-04-14T19:59:42Z
0 likes, 0 repeats
@foone @alice No feds but the angel of death made an appearance
(DIR) Post #B5ILyV4H717IOwwUa0 by azonenberg@ioc.exchange
2026-04-14T20:04:09Z
0 likes, 0 repeats
@foone @alice oh no, the grim reaper was a guest he didn't mess with the other guests. At least two attendees have died since mine as well but I am not aware of any new gender changes (the only trans guest I can remember was already out by that time)
(DIR) Post #B5IrVMZongrUCYqZA8 by azonenberg@ioc.exchange
2026-04-15T01:57:08Z
1 likes, 0 repeats
@n3tcat (recruiter walks up to you at emfcamp meowing session and says hi)
(DIR) Post #B5IrcotHvawq2jVdeS by azonenberg@ioc.exchange
2026-04-15T01:58:36Z
0 likes, 0 repeats
@n3tcat oh? I wouldn't expect Canada - UK travel to be problematic?
(DIR) Post #B5ItPg8rLQDh3KecdM by azonenberg@ioc.exchange
2026-04-15T02:08:10Z
1 likes, 0 repeats
@n3tcat thats quite the hangover.Then again I once left a day in Manhattan with a dozen or so, so...
(DIR) Post #B5ItPmbRK4e95i4X0y by azonenberg@ioc.exchange
2026-04-15T02:09:31Z
1 likes, 0 repeats
@n3tcat (as coinventor not solo... The customer basically had a brainstorming session with the team and patented everything that came out of it however impractical)
(DIR) Post #B5N9wsNzDb3YTwQAi0 by azonenberg@ioc.exchange
2026-04-17T03:33:10Z
1 likes, 0 repeats
57.4 kWh today, a new record by a massive margin. Exported 12.1 of that because I was generating 10.9 kW at peak while burning closer to 5.Tomorrow is looking to be a nice day and I'm also gonna be hopping up on the roof to wipe some of the pollen off the array which should buy me a few percent more generation as well, so we'll see how much I end up getting then.