kristianp

7 ms·

91.
Intel Makes Chipmaking History with High-NA EUV Panther Lake Production(hothardware.com)
4 points by kristianp 3mo ago|0 comments
92.
by kristianp 3mo ago
Apparently Apple is "in talks" with the PrismML: https://www.cnbc.com/2026/07/14/apple-prismml-ai-compression...
93.
by kristianp 3mo ago
Interesting, surf is based off WebKit2. https://surf.suckless.org/
94.
by kristianp 3mo ago
At least the repos are still there, for now. Time to get forking.
95.
by kristianp 3mo ago
Who was it that said: "worse is better"? This guy: https://dreamsongs.com/WorseIsBetter.html
96.
by kristianp 3mo ago
> Numbers like that buy a model a real migration effort. Such a silly choice of words. I wish the human directing the LLM writing the article put some effort into rewriting the worst examples of LLM style. > But it did extremely well,
97.
by kristianp 3mo ago
Ok. but why are you posting a spoiler?
98.
A Brief History of the Crazy Old 7-Segment Display(hackaday.com)
6 points by kristianp 3mo ago|3 comments
99.
WolfIP Doesn't Allocate(hackaday.com)
2 points by kristianp 3mo ago|0 comments
100.
by kristianp 3mo ago
> The increased nuclear mass causes orbiting electrons to speed up to a significant fraction of the speed of light, where the rules of Einstein’s theory of relativity are important. > In the relativistic regime, an electron’s spin — t
101.
Go Pop quiz: what time was it?(dave.cheney.net)
2 points by kristianp 3mo ago|3 comments
102.
by kristianp 3mo ago
I you look at https://arxiv.org/pdf/2401.04088 table 5 on page 8, you'll see that expert(s) used can change from token to token. The experts aren't divided along predictable lines.
103.
by kristianp 3mo ago
This is a little site I've had a for a while now. I've added a new page that compares the prices and memory bandwidth of machines with RAM shared between CPU and GPU. I've only listed machines based on DGX Spark, AMD Max+ 39
104.
GPU Quicklist – AI PC/Mac(gpuquicklist.com)
3 points by kristianp 3mo ago|1 comments
105.
by kristianp 3mo ago
Just about every release of go has incremental performance improvements [1]. It's a very stable language, so don't expect anything mind-blowing. For example the latest 1.26 moved a new garbage collector from experimental to prod
106.
by kristianp 3mo ago
https://news.ycombinator.com/item?id=48840113 seems to answer that question. I.e. an api for the compiler is being worked on.
107.
by kristianp 3mo ago
Yes, the way the URLs are interspersed between short passages makes reading it difficult. It's like reading an article on a news site without an adblocker.
108.
by kristianp 4mo ago
I'm not familiar with .net on Mac, what binding did you use to make this? .NET MAUI? https://learn.microsoft.com/en-us/dotnet/maui/developer-tool...
109.
by kristianp 4mo ago
> where it's easy to install a current OS and browser in 15+ year old computers That's true unless you're using Windows or Mac!
110.
by kristianp 4mo ago
https://en.wikipedia.org/wiki/Id_Tech
111.
by kristianp 4mo ago
If you look at https://openrouter.ai/z-ai/glm-5.2#providers there's about 28 providers, including z.ai and Alibaba. Most outside of China. I've never seen so many providers for a model on there before, glm 5
112.
by kristianp 4mo ago
> It loses to Intel in CPU Not really, their CPU market share has dropped in the last decade.
113.
by kristianp 4mo ago
The Mac Studio M3 Ultra with 96GB of RAM with 1Tb SSD costs $6799, but has triple the memory bandwidth [1]. It's almost twice the price of Strix Halo and when the M5 Ultras come out it will be interesting to see how much they go for.
114.
by kristianp 4mo ago
Interesting insight into Notch's creativity before Minecraft was released, although a public alpha was released beforehand in 2009. Kind of a shame that Java applet support was removed from browsers. Obviously the security problems mad
115.
by kristianp 4mo ago
Surely it's not actually 16 hours a day, that would leave maybe 6 hours for sleep after other needs are taken care of.
116.
by kristianp 4mo ago
Everything outside of the AI data centre build-out is not growing. And the hyperscalers are having to cut costs to help pay for those datacentres full of GPUs.
117.
by kristianp 4mo ago
The little crosses around the em-dashes? If I comment out the BitGeneva12 font in the css body they go away.
118.
by kristianp 4mo ago
LLM tells: > Fyve faded. Gin kept going. > the honest version of web programming: no magic, explicit control flow, and a handler shape you can hold in your head. > keep the request path explicit, run no reflection there, and put th
119.
by kristianp 4mo ago
There was a calculation done, but it had some serious flaws. There are some other discussions of this topic on HN too. Apple Silicon costs more than OpenRouter (williamangel.net) 355 points by datadrivenangel 44 days ago | 311 comments ht
120.
by kristianp 4mo ago
They've compared the Studio with 256GB of RAM when the 128 would have been better for comparison purposes. Neither config can be purchased right now, until the rumoured M5 Ultra comes out. Also how much does it cost for a PC that can t

More ›