fulafel

5 ms·

1.
by fulafel 4d ago
Transparent scheduling of superscalar execution was a later advance in microprocessors, termed out of order execution. Apart from micros both did come out around the same time in the mid-1960s. In the x86 microarchitectures superscalar came
2.
by fulafel 4d ago
What browser are you using an what does the back button do for you instead of going back?
3.
by fulafel 4d ago
Population-weighted, the country is more European (80% of pop on Europe side). That 80% counts towards population of Europe.
4.
by fulafel 9d ago
Seems incoherent. What exactly is the bad idea and why? Are they rediscovering "don't run stuff needlessly as root"?
5.
by fulafel 13d ago
128k tokens is about 12-20k lines of code/prompt right? Or more than an average novel. And you don't need all the source code or entire files in the context after all. I wonder what the approximate context window of a human progra
6.
by fulafel 13d ago
Isn't context size one of the central motivations of the whole agent / orchestration business - fanning out increasingly detailed work to a tree of subagents. Orchestrator vs worker, hierarchical multitier trees, etc. I guess this
7.
by fulafel 13d ago
There have been recent attempts. Apparently not.
8.
by fulafel 13d ago
What are the best benchmarks/leaderboards that compare task completion time between provider+model combos?
9.
by fulafel 14d ago
Yes it's chronological but no it doesn't have unresolved items nearer the present: The last 5 ones in the list read "Wrong" and seem reasonable. The first thing after the list is "After this point, most further pred
10.
Antirez Brings Vision to DS4, DeepSeek V4 Flash Runs Locally on an M5 Max(pasqualepillitteri.it)
3 points by fulafel 15d ago|0 comments
11.
by fulafel 15d ago
I thought in the post Ed was being evaluated based on all his predictions, and it turned out all of them were wrong. Most people will read the post as going over all the falsifiable predictions and none of them panning out, since after the
12.
by fulafel 15d ago
Data retention still sounds bad: "Claude Fable 5.1 and Claude Mythos 5.1 carry 30-day data retention and aren't available under zero data retention unless expressly authorized by Anthropic." Anyone know who the ZDR special tr
13.
by fulafel 15d ago
Can you run several MkLinux kernels (one concurrently with a Darwin kernel) at once?
14.
by fulafel 17d ago
Also several other places: https://en.wikipedia.org/wiki/District_cooling
15.
by fulafel 18d ago
Async ruins everything, as usual.
16.
by fulafel 18d ago
It has been a prominent "next major shift" idea in computer architecture since the 90s, to deal with the memory wall. Eg David Patterson advocating it in the 1990 and 1997 articles. or CRAM [1]. [1] https://www.eecg.tor
17.
by fulafel 19d ago
What % of global electricity production is in too small markets? Seems to be working well in EU at least, markets and grids being multinational. I'd guess many places in Africa could be potential markets that have underdeveloped grids
18.
by fulafel 20d ago
Where are their users coming from? Besides the few manually putting 1.1.1.1 in their settings.
19.
by fulafel 21d ago
And IPv6.
20.
by fulafel 21d ago
Microsoft has many products that people hate but are still the reason why they exist (eg Windows, Teams, Sharepoint etc).
21.
by fulafel 21d ago
We need a new kind of "view source" on the web.
22.
by fulafel 23d ago
Product companies have been talking this up for a year, eg: https://www.sandisk.com/en-ap/company/newsroom/blogs/2025/me...
23.
by fulafel 23d ago
The voters bear much of the blame too. The oil, gas and coal producing countries that still have democracies need to stop voting for short term profits and star voting for rapid rampdown of fossils production.
24.
by fulafel 23d ago
Sandboxes are speed bumps. Even the serious ones have spectator sports for compromising them (eg pwn2own vs Chrome). In addition, proprietary GPU sw stacks are notoriously crashy and lacking in robustness against hostile inputs, which the i
25.
by fulafel 24d ago
Even if you get past these problems, those models are available only under the condition that Anthropic retains your data.
26.
by fulafel 24d ago
Wouldn't that be the default assumption for Kimi or GLM? Only 1 out of 20 ppl live there after all.
27.
by fulafel 27d ago
Aaron's case was wrong, but we shouldn't turn coats about "information wants to be free" just because it's gaining more momentum? It would be petty and counterproductive to do it just for consistency of enforcing th
28.
by fulafel 27d ago
Microsoft are doing many things to help the AI coding surge, I'm not sure if this'd be a cost effective thing towards it.
29.
by fulafel 27d ago
In the README, the C version gets "452× faster than Python", the 2500x one is using OpenBLAS which uses lower level things than what's in C[1]. [1] see eg https://github.com/OpenMathLib/OpenBLAS/tree
30.
by fulafel 29d ago
They might mean static typing in queries.

More ›