klelatti

10 ms·

121.
by klelatti 2y ago
Are you really saying that first hand evidence of Zuck’s character is irrelevant to our understanding of him because it relates to a period when he was relatively young? We can’t use that together with his more recent actions to create a mo
122.
Modern Baby: A pioneering computer from Manchester(thechipletter.substack.com)
74 points by klelatti 2y ago|11 comments
123.
by klelatti 2y ago
If it's so cheap and easy then why haven't they done it already? ChatGPT was announced Nov 22. The opportunity has been clear for two years and still essential software breaks.
124.
by klelatti 2y ago
Since 2007 Nvidia has built with CUDA and compatible hardware - a large and growing ecosystem - a massive installed base of backwards compatible hardware - the ability to massively scale delivery of new systems - and lots lots more They now
125.
by klelatti 2y ago
Because we already see firms competing effectively with OpenAI. There is as yet no indication that AMD can match Nvidia's execution for the very good reason that doing so is extremely difficult. The head start is just the icing on the
126.
by klelatti 2y ago
> AMD could have ROCm perfected if they really want to. It's not an act of will or CEO dictat. It's about hiring and incentivising the right people, putting the right structures in place etc all in the face of competing demands
127.
AMD's RDNA4 Architecture [video](chipsandcheese.com)
2 points by klelatti 2y ago|0 comments
128.
by klelatti 2y ago
> The article would be equally fact-free non-news ... It may not be surprising but it's not fact-free or non-news. The pace at which Alibaba is developing its RISC-V designs is significant. Removing one dependency on western supply
129.
Why it's so hard to build a jet engine(construction-physics.com)
5 points by klelatti 2y ago|1 comments
130.
The Engine : Jonathan Swift (Sort of) Predicts Large Language Models in 1726(thechipletter.substack.com)
1 points by klelatti 2y ago|0 comments
131.
by klelatti 2y ago
Inexpensive personal computers weren’t shipped with EPROMs, they were shipped with mask programmable ROMs. EPROMs were used in development but they were nowhere near as important as the microprocessor.
132.
The Birth of Chipzilla(abortretry.fail)
4 points by klelatti 2y ago|0 comments
133.
by klelatti 2y ago
I worked in that 1950s office block (Temple Court) and was involved in the sale of the site to Bloomberg. The Temple of Mithras sat outside the office a few feet from the pavement with free public access at all times. It attracted very litt
134.
Intels Battlemage Architecture(chipsandcheese.com)
2 points by klelatti 2y ago|0 comments
135.
Work at the Mill Or, the Story of Digital Equipment Corporation(abortretry.fail)
6 points by klelatti 2y ago|0 comments
136.
by klelatti 2y ago
Completely agree with all these points and that without the 6502 the ARM1 would not have looked like it did. One might say ARM1 was inspired by the 6502 and it prevented them going down the CISCy 68k route. But I don't think that ARM &
137.
by klelatti 2y ago
Could you say why? It doesn’t feel much like it to me at all!
138.
by klelatti 2y ago
> ARM borrowed most of its design from 6502 Not really (perhaps leaving mnemonics aside). The 6502 has little in common with the first ARM. The ARM's designers liked the simplicity and speed of the 6502 and it was their favourite 8-
139.
by klelatti 2y ago
Thank you, that's very helpful.
140.
by klelatti 2y ago
All the RTL basically. It’s in a directory called gen_rtl (generated RTL?) and has remarkably few comments for such a complex code base. Also although technically it's open source if it's generated verilog then isn't that a l
141.
by klelatti 2y ago
For clarity from the original analysis > This document was originally written several years ago. At the time I was working as an execution core verification engineer at Arm. The following points are coloured heavily by working in and aro
142.
by klelatti 2y ago
Open source but my impression is that this contains verilog generated by another program?
143.
The Engine – Jonathan Swift (Sort of) Predicts Large Language Models in 1726(thechipletter.substack.com)
3 points by klelatti 2y ago|0 comments
144.
by klelatti 2y ago
No worries and thanks for sharing your experiences with RISC-V MCUs - really interesting. FWIW I agree that AArch64's launch was not the same as the RISC-V launch. There has clearly been a lot more work required to boot the RISC-V ecos
145.
A RISC-V Progress Check: Benchmarking P550 and C910(chipsandcheese.com)
6 points by klelatti 2y ago|1 comments
146.
by klelatti 2y ago
Hmmm. I didn’t make any comparison. I just pointed out that AArch64 isn’t 30 odd years old.
147.
by klelatti 2y ago
So PowerPC is 70 years old because IBM was designing computers in 1954? :)
148.
Back to the Future with Gordon Moore(thechipletter.substack.com)
2 points by klelatti 2y ago|0 comments
149.
by klelatti 2y ago
Hello Chester! Thanks for clarifying and for a terrific post.
150.
by klelatti 2y ago
I think it’s a bit problematic to say ARM is 30-ish years old. The company is 34 years old but 64 bit Arm (AArch64) which is really very different to its predecessors was announced in 2011 so arguably only 14 years old.

More ›