Array
(
    [content] => 
    [params] => Array
        (
            [0] => /forum/index.php?threads/tsmc-reports-first-quarter-eps-of-nt-7-98.17800/page-2
        )

    [addOns] => Array
        (
            [DL6/MLTP] => 13
            [Hampel/TimeZoneDebug] => 1000070
            [SV/ChangePostDate] => 2010200
            [SemiWiki/Newsletter] => 1000010
            [SemiWiki/WPMenu] => 1000010
            [SemiWiki/XPressExtend] => 1000010
            [ThemeHouse/XLink] => 1000970
            [ThemeHouse/XPress] => 1010570
            [XF] => 2021370
            [XFI] => 1050270
        )

    [wordpress] => /var/www/html
)

TSMC Reports First Quarter EPS of NT$7.98

I think that statement makes it clear there is significant N2 volume shipping in 2025.
Depends. If ramping happens in the second half of the year, then high volume products can’t launch until the first half of 2026 (as the chips would only just start leaving TSMC’s door at the end of the year). If ramping happens in the first half then I see no reason why we wouldn’t see iPhones in 2025. 2 quarters is a sufficient amount of time for wafers to go to Apple’s OSATs then Foxconn, then get shipped to key markets for a September launch.

Assuming Apple continues to be the lead customer, it does sound probable that ramp will happen in the first half of the year. If risk starts are still scheduled for late ‘24, N2 must have been moved up, and that is one commendable achievement. By my count this would also be the fastest ramp TSMC’s had in years (maybe even ever). N5 (what TSMC claims was their fastest ramping process to date) was if memory serves 5 quarters from the beginning of risk starts to iPhones on store shelves. N2 is sounding like it might be 3-4 quarters.
 
Last edited:
It only takes two months to package a chip and put it into a product. However, to launch an iPhone they need to build up a significant amount of inventory, and *that* takes more time. So you are correct - if ramp up only happens in Q3 then there is no 2025 2nm iPhone. However, the HPC customer mentioned by Wei could theoretically launch it's product in 2025.
 
I think that statement makes it clear there is significant N2 volume shipping in 2025.

We will know more next week at the TSMC Symposium. It sounds like there will be N2 revenue in 2025 for Apple and the CPU/GPU people. Most of the tape-outs I see today are still N5 but N3 is coming up quick, absolutely.
 
Same. Stock is very cheap IMO. Like i've said before, I cant understand why people think they can derisk their portfolios from an invasion of Taiwan. Not owning TSMC isnt going to spare you from economic armageddon.
True, but TSMC will get hit harder than almost every other company.

Intel probably will benefit, assuming Samsung gets dragged into it too. If not, Samsung could benefit the most. Intel will lose sales to China and Taiwan, but being the only game in town, their sales everywhere else will depend almost entirely on their fab capacity, which will not be impacted, and is growing aggressively.

I wouldn't buy TSMC, especially after this dismal report. Not that they don't make compelling technology, but the price is largely dependent on them remaining clearly superior to their competitors in technology, which is far from certain. To put it another way, they are worth almost 4x what Intel is. They have much less potential, given they don't sell their own products. On the other hand, they are currently very successful, despite this miserable quarter, whereas Intel's recent quarter would make TSMC's look spectacularly good.

But, their landscape is going to get a lot more complicated as IFS builds up, and it already is. Plus, even TSMC is careful to not to mention performance superiority with their process, they are already inferior in this metric, and it's clearly going to get worse. Whether their tech remains ahead in density and power is another story, it's certainly conceivable, but hardly certain with Intel's nodes that will be available by 2025. And Samsung isn't exactly going away.

And even if Intel doesn't take big market share quickly, which is almost a certainty they will not, just having them around and with all the extra capacity being built by everyone, will give companies using fabs more leverage. And then there's Samsung lurking. And IBM licensing their tech. Just seems like too many things can happen, and none of them good for TSMC; they are pretty much already in best case. What else can they do, have the leading tech and win huge orders from Apple, NVIDIA, AMD and even Intel? Already there. And they are priced accordingly.

Almost four times Intel just doesn't make sense to me. It's a nice company, in a nice position, but it's much more difficult to see them doubling than Intel, for example. Much more difficult. And so many risks, especially considering geo-political, which even is going to weigh on them even if nothing does happen. Because, it can, and companies are going to limit their risks if they can.
 
True, but TSMC will get hit harder than almost every other company.

Intel probably will benefit, assuming Samsung gets dragged into it too. If not, Samsung could benefit the most. Intel will lose sales to China and Taiwan, but being the only game in town, their sales everywhere else will depend almost entirely on their fab capacity, which will not be impacted, and is growing aggressively.

I wouldn't buy TSMC, especially after this dismal report. Not that they don't make compelling technology, but the price is largely dependent on them remaining clearly superior to their competitors in technology, which is far from certain. To put it another way, they are worth almost 4x what Intel is. They have much less potential, given they don't sell their own products. On the other hand, they are currently very successful, despite this miserable quarter, whereas Intel's recent quarter would make TSMC's look spectacularly good.

But, their landscape is going to get a lot more complicated as IFS builds up, and it already is. Plus, even TSMC is careful to not to mention performance superiority with their process, they are already inferior in this metric, and it's clearly going to get worse. Whether their tech remains ahead in density and power is another story, it's certainly conceivable, but hardly certain with Intel's nodes that will be available by 2025. And Samsung isn't exactly going away.

And even if Intel doesn't take big market share quickly, which is almost a certainty they will not, just having them around and with all the extra capacity being built by everyone, will give companies using fabs more leverage. And then there's Samsung lurking. And IBM licensing their tech. Just seems like too many things can happen, and none of them good for TSMC; they are pretty much already in best case. What else can they do, have the leading tech and win huge orders from Apple, NVIDIA, AMD and even Intel? Already there. And they are priced accordingly.

Almost four times Intel just doesn't make sense to me. It's a nice company, in a nice position, but it's much more difficult to see them doubling than Intel, for example. Much more difficult. And so many risks, especially considering geo-political, which even is going to weigh on them even if nothing does happen. Because, it can, and companies are going to limit their risks if they can.
Intels P/E is higher than TSMC’s. TSMC is valued outright more because it’s considerably more profitable……
 
I think that statement makes it clear there is significant N2 volume shipping in 2025.
Yes, indeed 2nm technologies and volumes are strongly planned by TSMC but not yet hugely possible in very short term, like Rapidus the new Japanese Foundry startup created by 8 Japanese companies to build 2nm technology and recently announced related fab in Hokkaido, requiring around 51.4B$ investment and in collaboration for technology development with IBM and European research institute IMEC in Belgium in order that,
(i) prototypes line starts in 2025…
(ii) and mass production around 2027…
 
Back
Top