klelatti

27 ms·

361.
by klelatti 3y ago
> To be able to write performant programs Wait we’re teaching kids to write performant assembly language?
362.
by klelatti 3y ago
I’ll add it to my wish list!
363.
by klelatti 3y ago
This isn’t what the comment I replied to was suggesting. Also, I can see there is a case for using a simple modern design such as Cortex-M as a base but - as the peer comment also says - it’s really not the case that a Cortex M7 is conceptu
364.
by klelatti 3y ago
Are you suggesting that 'Deep OS knowledge', syscalls, C, assembly, packets and networking protocols should be taught in schools? What I think the author is getting at with '8-bit' is that presenting children with a comp
365.
by klelatti 3y ago
Know of but don’t (yet) own. Strong recommendation?
366.
by klelatti 3y ago
Same here! For those not similarly blessed, the Internet archive has a copy. https://archive.org/details/CompleteSpectrumROMDisassemblyTh... Of particular note, the use of a stack based VM for floating point.
367.
60-Bit Computing (2022)(thechipletter.substack.com)
77 points by klelatti 3y ago|73 comments
368.
by klelatti 3y ago
> I stole copiously from The Complete Spectrum ROM Disassembly by Ian Logan and Frank O’Hara This was an actual book, readily available at your local bookstore, that had a full, independently produced, Z80 disassembly listing (with great
369.
by klelatti 3y ago
Absolutely. Claiming you’ve found a solution to this problem with such a shaky premise is actually inviting others to join your own cargo cult.
370.
by klelatti 3y ago
Errr SaaS caused the great stagnation. Or non text UIs. Or consulting firms. I’m confused! And this is just factually incorrect. > Intel’s road to ruin was when the CFO became the CEO and managed the business by the numbers and abstracti
371.
Unboxing the Tenstorrent Grayskull AI Developer Kit(morethanmoore.substack.com)
3 points by klelatti 3y ago|0 comments
372.
by klelatti 3y ago
100%! Also in a way reversing what Moore/Grove did when they abandoned commodity memories. Such a hard thing to do.
373.
by klelatti 3y ago
They are terrific. Thank you for sharing and licensing under CC.
374.
by klelatti 3y ago
Except they weren't real pivots. Mobile / GPU were x86 centric, foundry half hearted without buying into what needed to be done. Buying a company is the easy bit.
375.
by klelatti 3y ago
> Nope. Intel crushed them. The problem is that Intel has had a defensive strategy for a long time. Yes, they crushed many attempts to breach the x86 moat but failed completely and then gave up attempts to reach beyond that moat. Mobile,
376.
by klelatti 3y ago
The RPi foundation is a charity.
377.
Reverse Engineering a Soviet CMOS Chip(twitter.com)
63 points by klelatti 3y ago|14 comments
378.
Motorola, Intel, IBM Make a Mainframe in a PC – The PC XT/370(thechipletter.substack.com)
2 points by klelatti 3y ago|0 comments
379.
Moore on Moore: The past, present and uncertain future of Moore's Law(thechipletter.substack.com)
2 points by klelatti 3y ago|0 comments
380.
by klelatti 3y ago
No criticism of AdaFruit intended. I meant the HN link should be to the original article - the post that I say the AdaFruit post links to.
381.
by klelatti 3y ago
Would we have been able to train the LLM in the first place though? Guessing that that would have been completely infeasible?
382.
by klelatti 3y ago
Oh wow! Thanks, that's a great discussion.
383.
by klelatti 3y ago
Nothing against AdaFruit, but this really should link to Roy Longbottom's detailed comparisons [1] that this very short post links to. [1] http://www.roylongbottom.org.uk/Cray%201%20Supercomputer%20P...
384.
by klelatti 3y ago
From the Oral History [1] on the rationale for the move: > [Bill] Lattin: And Jack Henneshen [senior Motorola manager], it is claimed that he wanted to move down there because he wanted a windmill on his property. > [Tom] Bennett: I t
385.
by klelatti 3y ago
The really remarkable thing about the 6502 is how the designers managed to get such strong performance out of such an economical design. It used less than half the transistors of the Z80. As a result it really was the design that helped to
386.
Neural Network Quantization and Number Formats from First Principles(semianalysis.com)
1 points by klelatti 3y ago|0 comments
387.
by klelatti 3y ago
Open source and written in Elixir. As someone who is a heavy user of Pinboard and has had Elixir on my 'to learn properly after dabbling' list for a while, looks like a great project to explore.
388.
by klelatti 3y ago
If anyone is interested in the Transputer instruction set, I wrote a short post on the topic last year. In short few registers, stack based but with a tiny stack, heavily microcoded. https://thechipletter.substack.com/p&#x2F
389.
by klelatti 3y ago
It is an interesting piece of economics / public policy history. I think using the phrase 'throw money at any UK firm' though does a bit of a disservice to how credible Inmos's technology was. They were building at or cl
390.
Tiny Computers from Texas: The Early History of the Microcontroller(thechipletter.substack.com)
97 points by klelatti 3y ago|7 comments

More ›