Matt Schlicht made Moltbook, not Peter Steinberger.
4,137 karma · joined December 29, 2013
Matt Schlicht made Moltbook, not Peter Steinberger.
I think you're misunderstanding or misrepresenting them. The fight to have the most jaded or pessimistic take, the hottest flame, the spiciest rant, it's all so predictable and it's just a bunch of the same people saying the same things and agreeing with each other for the nth time. It brings nothing new to the table, and the posts that actually respond to the new information get drowned out or worse downvoted for insufficient vitriol.
my system is running GLM-5 MXFP4 at about 17 tok/s. That’s with a single RTX Pro 6000 on an EPYC 9455P with 12 channels of DDR5-6400. Only 16k context though, since it’s too slow to use for programming anyway and that’s the only application where I need big context.
I think you’re imagining that these hypocrites exist.
I wish I had better numbers to compare with the 2x M3 Ultra setup. My system is a few RTX A4000s on a Xeon with 190GB/s actual read bandwidth, and I get ~8 tok/s with experts quantized to INT4 (for large models with around 30B active parameters like Kimi K2.) Moving to 1x RTX Pro 6000 Blackwell and tripling my read bandwidth with EPYC Turin might make it competitive with the the macs, but I dunno!
There's also some interesting tech with ktransformers + sglang where the most frequently-used experts are loaded on GPU. Pretty neat stuff and it's all moving fast.
I felt people were unwilling to take the responsibility for communicating properly, and so they took the easy route where they could shrug their shoulders and say "I included all the context."
I only ever got complaints from people who were confused by the quoting style or didn't know what the email was about. I'm not sure if it's still true, but at the time, Outlook didn't use threaded view mode by default and most people didn't know about it. FWIW I work in manufacturing and not in tech, I expect the level of competence in tech is a little higher, though I also hear how people moan about having to learn the tools they use every day, so maybe there's little difference.
I'm trying to figure out how to add another card on a riser hanging off a slimsas port, or maybe I could turn the bottom slot into two vertical slots.. the case (fractal meshify 2 xl) has room for a vertical mounted card that wouldn't interfere with the others, but I'd need to make a custom riser with two slots on it to make it work. I dunno, it's possible!
I also have an RTX Pro 6000 Blackwell and an RTX 5000 Ada.. I'd be better off pulling all the A7000s and throwing both of those cards in this machine, but then I wouldn't have anything for my desktop. Decisions, decisions!
Well, duh.
> unless you use the binary-only drivers
Which you also have to do with nvidia cards.
> which only work on a handful of very new cards
So get a new card?
> You can't use it for editing video.
Yes I can.
----
I actually switched from a 7900 XTX to a 4090 BITD because I wanted CUDA, so I get that angle, but that doesn't mean I go around telling people "AMD isn't accelerated," because it's not true and it's a silly thing to try to claim.
This is a bewildering assertion.
Is this more of an aspirational thing, like Windows supports "doing it right", and with time and effort by the right people, more and more applications may be able to be drawn correctly?
[edit] I guess so, I see your comment about setting registry keys to make stuff work in Microsoft's own programs. That aligns more closely with my experience.
Development systems for AI inference tend to be smaller by necessity. A DGX Spark, Station, a single B300 node... you'd work on something like that before deploying to a larger cluster. There's just nothing bigger than what you'd actually deploy to.