@parkbot.bsky.social
@philparkbot
Followers
3K
Following
7K
Media
2K
Statuses
7K
Burrito technologist. @AMD Infinity Fabric performance and architecture. CTRL+ALT+DEL the GOP. Personal account. Reluctant x86 defender.
British Columbia, Canada
Joined November 2008
My BIL’s name is Moise, but I only call him by his full name of Modified Owned Invalid Shared Exclusive when he’s in real trouble.
M(O)ESI(F|R) Modified, Owned, Exclusive, Shared, Invalid, Forward, Recent Coherency protocol states for a given line in a modern microprocessor cache
3
2
35
Ever wonder why compilers use `xor eax, eax` to zero registers? It's smaller AND faster—CPUs optimise it out entirely! Day 1 of Advent of Compiler Optimisations: https://t.co/VOo6hv2rEX Video: https://t.co/Kcqj07D5Ve
#AoCO2025
14
34
263
"Tachyum Unveils 2nm Prodigy with 21x Higher AI Rack Performance than the Nvidia Rubin Ultra" - 1,024 64-bit cores - up to 6 GHz - 1 GB L2+L3$ - 24x DDR5-17600 - 48 TB per socket - 128 lanes of PCIe 7.0 - 1,600 W TDP But surely ...
3
4
22
The next Frontier is Discovery! Leveraging the @AMD Instinct MI430X: co-designed for high-performance computing. This system will be designed by @ORNL , @HPE and AMD. Coming in 2028. I am very honored to be part of this project. https://t.co/25coFQ12RI
reuters.com
The U.S. has formed a $1 billion partnership with Advanced Micro Devices to construct two supercomputers that will tackle large scientific problems ranging from nuclear power to cancer treatments to...
1
7
21
@philparkbot I think this is all a misunderstanding of the original quote from Shang-Yi Chang, former TSMC VP - he talks about using 3 shifts of *research* engineers 24/7 to speed up R&D and produce one new process node every 1-1.5 yr instead of standard 2 yrs. Not about fab technicians
2
1
13
Is the implication that US fabs don’t run 24/7??? They do. The standard 24/7 manufacturing schedule is to have 4 shifts: 12 hr shifts 3 days/week and 4 days/week + night shift and holiday pay + OT because it’s cheaper to do that than have idle capacity.
At TSMC, PhD-level engineers work in three rotating shifts. The fabs in Taiwan run 24 hours a day without stopping. Do you really think Americans would work like this? Or even want to? And let’s assume, for the sake of argument, that the work culture in the U.S. changes to make
3
6
98
Hello you fine Internet folks, Today we are looking at AMD’s largest client APU to date, Strix Halo. This is an mobile APU able to handle high end CPU and GPU workloads without compromise. Hope y'all enjoy! https://t.co/PS7e1rgbZJ
https://t.co/acUqHqMiCg
old.chipsandcheese.com
Hello you fine Internet folks! Today we are looking at AMD’s largest client APU to date, Strix Halo. This is an APU designed to be a true all-in-one mobile processor, able to handle high end CPU an…
10
32
145
Legend has it gym memberships are particularly hard to cancel here because nobody knows how to exit
293
818
11K
For those interested, @philparkbot The Long Mode Chronicles is surely worth a read! https://t.co/SpLN3a4xHU
Ever wonder how AMD beat Intel with their own instruction set? This is the inside story of X64... RT if you like exclusive pre-release videos on X!
0
1
3
Ever wonder how AMD beat Intel with their own instruction set? This is the inside story of X64... RT if you like exclusive pre-release videos on X!
29
174
1K
When AMD64 was proposed, Athlon was already in the market and was scaling beyond what (some at) Intel believed was possible less than 10 years prior. But by then, Intel had already committed to IA64 with HP
0
1
12
Folks may not realize that when Intel announced IA64, their flagship part was Pentium - superscalar but in-order. While some in the company didn’t believe CISC could continue improving, Pentium Pro was still in active development and showed that superscalar and OoO was possible.
2
1
12
A lot of people rightfully complain that x86 is complex and ugly, but this quote from Linus Torvalds on x86-64 vs IA64 was also revealing
1
1
16
A fact I memory holed goes back to the CISC vs RISC debate. My recollection for IA64’s justification was that Intel believed CPUs were getting too complicated and that they wanted a new ISA. In reality it was because they believed CISC could not scale or be competitive with RISC!
2
1
12
I wrote my first and maybe only Substack post. This is a follow up to my Colwell post where I found a 2007 technical report by two Microsoft employees who interviewed folks from MS, AMD and Intel about the history of the transition, and I learned a lot! https://t.co/5eHDxVzEue
I was looking up some history behind the x86-64 transition, particularly around the Pentium 4 time frame, and I found out that Bob Colwell (Pentium Pro chief architect) has been posting on Quora. Pentium 4 had a version of x86-64 that was fused off.
9
41
223
Those who have followed Jon know there’s only one possible reaction to his tweet.
@geofflangdale @lemire I can’t believe it but I’m defending x86 😉
2
1
8
Everyone knows that the x86 ISA is big. Modern CPUs have ~1000+ mnemonics. Guess how many make up 90% of compiled C/C++ code? TWELVE. I'm not kidding. The question is…what if we shrank it?
1
2
12
There is a family photo in the background. The left is the original, the right is from the image above. I don’t know how else to describe it except that it looks like a Bob’s Burgers + Adventure Time crossover.
1
0
4
ICE has released a video of its raid on Hyundai–LG's Georgia battery plant site, showing Korean workers chained up and led away. South Korea's foreign ministry has confirmed over 300 of the 457 taken into custody are Korean nationals.
US authorities have reportedly detained 450 workers at Hyundai-LG battery plant construction site in Georgia yesterday, including over 30 South Koreans said to have legitimate visas. Seoul has expressed concern and says Korean nationals' rights "must not be unjustly violated."
817
3K
17K
“Speaking of efficiency, the topic of ARM vs. x86 came up here, as it does every year.” 💀 💀 💀
2
1
16