sylware

18 ms·

361.
by sylware 7mo ago
The current hardware used is self-hosting mini-server grade, and certainly not on the latest silicon process. "Slow" is expected. It is not the ISA, but the implementations and those horrible SDKs which needs to be adjusted for RI
362.
by sylware 7mo ago
What??? microsoft openai is Big Tech. Are you ok?
363.
by sylware 7mo ago
Where have you been in the last 2 decades?
364.
by sylware 7mo ago
Is this a drone or a transport for delta force hit and run?
365.
by sylware 7mo ago
If, for even 1s, they get in a position which is threatening, in any way, Big Tech AI (mostly US based if not all), they will be raided by international finance to be dismantled and poached hardcore with some massive US "investment fun
366.
by sylware 7mo ago
I am almost sure it was done so carefully that you can extract it from the abominations which are the whatng cartel web engines with a direct to OS abstraction layer that with only some little amount of work.
367.
by sylware 7mo ago
Then specialized models would be even better, more so if the ML/inference uses a formal solver and maybe some scratch memory.
368.
by sylware 7mo ago
Exactly why I do ask for non LLM, since it is not specialized for maths.
369.
by sylware 7mo ago
What is the model used for maths since it is not LLM? This model would have to be strap on a formal solver as a tool. Are those models training to use "scratch memory"? So many question about the models used for real maths.
370.
by sylware 8mo ago
There are tons of RISC-V SOCs and mini-boards. Ez and inexpensive native port...
371.
by sylware 8mo ago
Which models? Which nasty issues did you encounter?
372.
by sylware 8mo ago
Yep, licensing fee and IP usage restrictions is a massive decision point on some silicon markets. The other massive point: RISC-V integrates a lot of CPU "we know now" in a very elegant "sweet spot". And it is not china
373.
by sylware 8mo ago
But with hardware IP locks like x86_64. Better favor as much as possible RISC-V implementations. But, I don't know if there are already good modern-desktop-grade RISC-V implementations (in the US, Sifive is moving fast as far as I know
374.
by sylware 8mo ago
Anybody is aware of a public token (severely limited) I can use to test claude coding ability? You know using CURL. I am itching at testing claude for assembly coding and c++ to plain and simple C ports.
375.
by sylware 8mo ago
Anybody: can I test claude code without a whatng cartel web engine? web API using curl with an "public" token? Anything else? I am itching at testing its ability to code assembly.
376.
by sylware 8mo ago
Wait, I can get such a key and perform gemini API requests with curl? (probably limited in some ways)
377.
by sylware 8mo ago
Wallet codes bought at local monetary terminals to increase the balance of an account seems to be the less too much 'insecure' way to do things and to stay away from a hard dependency on whatng cartel web engines.
378.
by sylware 8mo ago
And of course, this is not a web app but a web site with basic HTML forms not requiring a whatng cartel web engine?
379.
by sylware 8mo ago
This is what I implied: this is not against the US, which have actually the most control and sovereignty on critical software. It is much cheaper and easier to have control and sovereignty on less complex software, including the SDK. Usuall
380.
by sylware 8mo ago
From an applications point of view: They want web apps only running in whatng cartel web engines? libreoffice? A massive piece of software you can build only with US c++ compilers (MIT and mostly apple)? (the mistake was to use c++ in the f
381.
by sylware 8mo ago
What's the article is missing: this is directly related to the complexity of file formats and protocols. There are 2 webs: - the web site, to serve noscript/basic (x)html, namely basic HTML forms which can be augmented with <vi
382.
by sylware 8mo ago
Try to talk about noscript/basic (x)html browser interoperability on HN, namely the web without the engines from the whatng cartel. You'll see...
383.
by sylware 8mo ago
This is part of the awakening: IPv6 is the only way to make work the real, symmetric, internet at scale. The next step is to assess the damage on DNS and online payment systems, basically now gated by whatng cartel web engines. What a sad s
384.
by sylware 8mo ago
Isn't that linux DMA buf?
385.
by sylware 8mo ago
That makes me think about the whatng cartel apocalypse. People lost themselves, forgetting how important noscript/basic (x)html (aka basic HTML forms, nowdays which could be augmented with <audio> and <video>)) has been for
386.
by sylware 8mo ago
And noscript/basic (x)html for web sites? They broke them to force the usage of whatng cartel web engines (or they were incompetent and/or malicious). At least on critical sites. GAFAM and big tech do not like protocols and file f
387.
by sylware 8mo ago
ARM should drop its ISAs and move to 'licensing' RISC-V implementations. IP-locked ISA is something nobody sane would want.
388.
by sylware 8mo ago
They are still trying.... insane...
389.
by sylware 8mo ago
Allright, the ISA itself is probably protected under some copyrights(+patents), which in the US last for ~one century (cf Disney). In the end, nobody sane would try its luck, better go for something non "IP-locked". Aka RISC-V, no
390.
by sylware 8mo ago
As far as I can remember, it is not only a "patent" issue. It seems there are other legal mechanisms. That said, I would not use a x86_64 CPU without AVX nowadays.

More ›