HNHacker News
TopNewBestAskShowJobs

klelatti

10,089 karma · joined May 1, 2020

Writing a newsletter about computer history at thechipletter.substack.com
submissionscomments
klelatti··on Alibaba/T-HEAD's Xuantie C910
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 lot less useful than the code that was used to generate the rtl?

klelatti··on Alibaba/T-HEAD's Xuantie C910: An open source RISC-V core
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 around the execution cores of various processors. Apply a pinch of salt; points contain varying degrees of opinion.

> It is still my opinion that RISC-V could be much better designed; though I will also say that if I was building a 32 or 64-bit CPU today I'd likely implement the architecture to benefit from the existing tooling.

klelatti··on Alibaba/T-HEAD's Xuantie C910
Open source but my impression is that this contains verilog generated by another program?
klelatti··on SiFive's P550 Microarchitecture
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 ecosystem and they have made amazing progress. Arm had the advantage of incumbency and a lot more resources.

At the same time I think the have oversold progress on application level processors and I don't think this does RISC-V any favours at all. Arm has a lot of experience and the general tone from some of the RISC-V commentary is that they got it 'wrong' with AArch64 which, to be charitable, is unproven.

klelatti··on SiFive's P550 Microarchitecture
Hmmm. I didn’t make any comparison. I just pointed out that AArch64 isn’t 30 odd years old.
klelatti··on SiFive's P550 Microarchitecture
So PowerPC is 70 years old because IBM was designing computers in 1954? :)
klelatti··on SiFive's P550 Microarchitecture
Hello Chester! Thanks for clarifying and for a terrific post.
klelatti··on SiFive's P550 Microarchitecture
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.
klelatti··on SiFive's P550 Microarchitecture
From the SiFive website [1]

> The Performance P550 scales up to four-core complex configurations while delivering 30% higher performance in less than half the area of a comparable Arm® Cortex®-A75.

Dylan Patel wasn't impressed by these comparisons with A75 [2]

> @SiFive is claiming half the area and higher perf/GHz, but they are using 7nm and 100ns memory latency. Choosing to compare to the 10nm A75 on S845, notorious for its high latency at over 200ns. Purposely ignoring iso-node or other A75 comparisons.

And this analysis seems to be borne out in this Chips and Cheese post.

> As a step along that journey, P550 feels more comparable to one of Arm’s early out-of-order designs like Cortex A57. By the time A75 came out, Arm already accumulated substantial experience in designing out-of-order CPUs. Therefore, A75 is a well polished and well rounded core, aside from obvious sacrifices required for its low power and thermal budgets. P550 by comparison is rough around the edges.

So what to make of SiFive's claims? It seems quite an important claim / comparison.

[1] https://www.sifive.com/cores/performance-p550

[2] https://x.com/dylan522p/status/1415395415000817664

klelatti··on How to build your own ZX80/ZX81 (2019)
Out of interest what was it on the ZX81 that impressed your family and convinced them to upgrade?
klelatti··on UK's hardware talent is being wasted
And the author's most recent tweet [1] (apart from one referring to this HN post):

> Is it just me or is UK’s hardware scene really kicking off again?

> Founder friends have just raised millions, moved into massive warehouses, imported CNC machines and some started metal casting.

> Even SaaS VC friends are talking about hardware now

[1] https://x.com/joseflchen/status/1881058447946391848

klelatti··on SQLite: How it works, by Richard Hipp (2024) [video]
Link to slides for this talk:

https://sqlite.org/talks/howitworks-20240624.pdf

klelatti··on A History of Early Microcontrollers, Part 1: Calculator Chips Came First (2022)
I see you’ve chosen not to reply to this question so FYI this article was not AI generated in any way.
klelatti··on A History of Early Microcontrollers, Part 1: Calculator Chips Came First (2022)
Thank you! I had no idea. By the way, the post definitely isn't AI—whether it's 'slop' is a matter of opinion, of course.
klelatti··on A History of Early Microcontrollers, Part 1: Calculator Chips Came First (2022)
I’m genuinely interested in why you think this particular post is ‘AI slop’.
klelatti··on A History of Early Microcontrollers, Part 1: Calculator Chips Came First (2022)
From the HM guidelines

> Please don't post shallow dismissals, especially of other people's work.

klelatti··on A History of Early Microcontrollers, Part 1: Calculator Chips Came First (2022)
May also be of interest on this topic:

https://thechipletter.substack.com/p/tiny-computers-from-tex...

klelatti··on Dinner for One: the British comedy Germans have been laughing at for years(2018)
Mystified as to why NDR (German TV station) would have wanted to record an English language performance of this sketch in 1963.

Would they not have preferred a German version? Was there a lot of English language TV being broadcast in Germany at that time?

klelatti··on Morris Chang and the Origins of TSMC
Kudos to Brian Potter for getting hold of a copy of Chang’s biography and translating it (with some machine assistance) himself!
klelatti··on The Nvidia Way
Hello Tae and thanks for a terrific book! My comment only compared Elon and Jensen, not Isaacson's book and yours. I completely understand that you had a different set of objectives and you did a fantastic job in setting out what makes Nvidia special without turning it into a Jensen hagiography.
klelatti··on The Nvidia Way
> I posit that this was an enormous gamble to boost valuations of both NVIDIA and ARM and that they knew from the beginning that the deal would not go through.

You posit it without saying why the valuations would go up given that you think they - and everyone else given what you say - thought the deal wouldn’t complete.

klelatti··on The Nvidia Way
Fair comment. I'm in the middle of reading the Isaacson Musk bio so maybe the contrast between the two made me think of Jensen as shy and retiring.
klelatti··on The Nvidia Way
That's a good list. I'd also add a willingness to take risks, which was particularly evident in the early years with RIVA 128 but also arguably CUDA was a risk and one that didn't pay off until relatively recently.
klelatti··on The Nvidia Way – Review of Tae Kim's New Book
I guess it just shows that Jensen has had a (perhaps admirably) low profile until recently.
klelatti··on The Nvidia Way
I think that the blocking of the deal by competition authorities shows the potential benefits of the deal (for Nvidia) in having control of Arm, in addition to making Arm-based SoCs, were real.

Also, Rene Haas (ex Nvidia) became CEO of Arm after the deal fell through. Simon Segars was CEO when the deal was announced.

klelatti··on The Nvidia Way
Jensen Huang was their first CEO!
klelatti··on The Nvidia Way
In addition to Tae Kim's interview with Asianometry's Jon Y [1] mentioned in the post, readers might also enjoy his more recent interview on The Riff podcast [2]. Both are terrific.

[1] https://www.youtube.com/watch?v=3vjGIZOSjLE

[2] https://www.youtube.com/watch?v=I5qub-IS2gY

klelatti··on Ask HN: Pinboard Alternatives for 2025
Thank you! I'll definitely check it out.
klelatti··on CUDA Moat Still Alive
AMD doesn’t just have to fix these issues it has to build up a record of fixing issues like those discussed here.

Otherwise who will bet their firm / cash / career on new hardware without a successful track record.

klelatti··on Tenstorrent and the State of AI Hardware Startups
No, the parent comment is correct (on the relative fees at least). Qualcomm wants to sell Nuvia-derived cores under its own ALA at a lower fee than Nuvia would have had to pay (which makes sense as server cores cost more than smartphone cores).
← PreviousPage 4 of 34Next →