You are currently viewing SemiWiki as a guest which gives you limited access to the site. To view blog comments and experience other SemiWiki features you must be a registered member. Registration is fast, simple, and absolutely free so please, join our community today!
Yeah I wasn't super clear about that. It wasn't really a jab at technological competitiveness as that is a remains to be seen sort of thing given intel hasn't talked about what 14A brings over 18A and even then we need to wait until 2026 just to see how the different aspects of N2 stack up vs...
Kind of disappointing to hear that TSMC is doing two teams with 5-7 years of development. That would mean the earliest we can expect A14 (or I guess in light of the renaming of N2+BSPD to A16, that they are more likely to go with A12) is 2028, and worst case 2030. If TSMC lets the gaps between...
I have a lot of disorganized thoughts, so I will try to have them listed as coherently as possible:
Originally I had thought that the TSMC AZ stuff was way overblown, but there is to much evidence that I can no longer just write everything off anymore. There is clearly some degree of issues...
A love a good “rumor” that is just public knowledge. TSMC said as much on their last earnings so pretty laughable reporting, thanks for the chuckle Dan! In other hot rumor news I heard that intel’s next generation server chips are made on intel 3 and are called sierra forrest and granite rapids...
My guess for earliest intercept is launching the M4 with a launch during the summer (with the large die ultra and max chips coming at a later time like with M2). Option two is the 2026 iPhone. TSMC mentioned how brutal the process flow durations are getting, so it would not surprise me if Apple...
Probably because memory is the ugly stepchild of the semi industry. It is a sad fate since no memory means the logic is completely worthless, but good luck making any technical arguments to commerce. As for intel in particular they are also building more factories in the US than Micron 5+1...
Not using EUV for 10nm was (and even with the benefit of hindsight still is) objectively the correct decision, no ifs thens or buts. Even if ASML had the tools in 2018 to fully equip intel at the expensive of Samsung and TSMC for volume production in 2019 (Icelake), then 10nm would have been...
My guess is they were expecting more. Being up double digits is kind of the expectation for a firm that has grown every year since it’s inception with the exception of 2023. A similar deal happened with NVIDIA’s last call where they were up a good bit but stock stayed flat because the...
We will have to see what the scale of this R&D is, but this is potentially very interesting/a big shake up. I have to assume that TD will still be done in Hwaseong for logic and memory. But maybe they will alternate logic TD between Korea and Texas (similar to what Micron seem to do with DRAM TD...
Could you give some examples? My understanding is that everything was just serial numbers for a while (4004, 8088, 286, 486, etc) before moving to things in Oregon (Tualatin, Willamette, etc), and finally I assume the current lakes and rapids aren't all in Oregon.
That is derivative nodelet. TSMC almost never converts fabs to new process technologies, they almost always build new fabs for new process node families. The only exceptions in TSMC company history are 20nm fabs since that technology didn't gain widespread adoption, 10FF for the same reason...
TSMC isn't intel was the one who built an advanced packaging fab in the US. Thanks for pointing this out since the wording was terrible, I fixed the comment for clarity.
TSMC is getting 78% the money as intel while intel builds 1.67X more fabs in the US + an advanced packaging site. The US has it's national champion and their based in the based in Hsinchu (part of the mythical 51st state of the ROC) :ROFLMAO:. Jokes aside this is super cool to see. I'm happy to...
Kind of insane that intel is as close to TSMC as they are. I expected it to be a larger gap given intel lacks the free money printers (aka old depreciated fabs). While 2021 is the most extreme year, to think that back in 2021 intel wafer demand (on a dollar basis) was like 30-40% of all of...
Excuse my ignorance Ian, but I guess a few questions. The way NVIDIA makes their new product more valuable than the old one is improving the flops/$, yes? If so then I assume we would agree that the best way to increase the number of flops/$ is to increase energy efficiency because over the...
