Array
(
    [content] => 
    [params] => Array
        (
            [0] => /forum/index.php?threads/tsmc-1q22-discussion.15885/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 1Q22 Discussion

I'm in the camp that think there is likely going to be a recession around the corner - maybe 6-12 months away. Right now we have both semiconductor capex and demand growth in the 30% range. Even if demand growth were to fall back to a still very healthy 10%, we'd likely see a glut. If demand shrinks due to a recession there will be a lot of underutilized fabs in 2 years.
The idea we are headed for a recession seems quite speculative. But demand for chips will likely continue to grow even if there is a slowdown in growth. Everyone I hear agrees the chip shortage will continue into next year. So it's simply hard to imagine we will see a glut in chips in the short or medium term.
 
I'm in the camp that think there is likely going to be a recession around the corner - If demand shrinks due to a recession there will be a lot of underutilized fabs in 2 years.

Barring some worldwide near-extinction event, any recession will be short lived for TSMC.

In 2021 the global automotive industry “alone” lost $210 billion. Ask the automotive industry "today" if they would rather have an under-supply, or have the luxury of throwing away any unusable or obsolete over-supply.

In a few years, light bulbs will have more compute power than Voyager 2.
 
You might be right about that. But isn't that putting the cart before the horse? Is there any evidence that is the case. Last I heard, everyone is talking continued shortages into next year.

I don’t think we have a wafer shortage. We do have systems level bill of material shortages. Not everything that is needed is readily available like before the pandemic.

Inflation should ease demand in 2H 2022 and into 2023. Getting people back to the office will help as well. Less time and less money to spend.

Unfortunately Covid is not going away anytime soon. Businesses near me are being shut down for days at a time as entire offices are getting sick. Entire cities in China are being shut down in an effort to contain new variants. At home Covid testing and not needing hospitalization if you do get sick is skewing the numbers.

I frequently sail past the port of Oakland in Northern CA. I have never seen it busier. I have radar on my boat so I can monitor traffic from my slip in Alameda.

Bottom line: Wafer manufacturing shortages are not going to be our biggest worry in the coming quarters.
 
TSMC 2Q22 revenue guidance is $17.6B to $18.2B (midpoint $17.9B).... You're range is a billion low.

Such guidance equates to QOQ revenue growth rate of 4.4% to 8.0% (midpoint 6.2%), and a YOY revenue growth rate of 37.8% to 42.5% (midpoint 40.2%).

1Q22 was 3.5% over the high end of revenue guidance.
Thank you for pointing out the error. I have modified my original post in order to avoid any confusion, The correct one:


"While Q1 2022 36% YoY revenue growth is impressive, the YoY revenue growth for Q2 2022 is significant as well. TSMC Q2 2021 revenue was US13.29 billion and Q2 2022 revenue will be between US$17.6-18.2 billion, according to the guidance given yesterday. It means Q2 2022 YoY growth will be between 32.43% and 36.95%.

Typically TSMC's Q3 and Q4 contribute the largest share to their whole year revenue. Using TSMC US$56.82 billion 2021 revenue as the base and assume a 30% YoY growth, TSMC can achieve US$73.866 billion 2022 revenue. If we reduce the growth rate to 25%, TSMC possibly can reach US$71.025 billion 2022 revenue.

Intel 2022 revenue guidance is set at US$76 billion currently."
 
I don’t think we have a wafer shortage. We do have systems level bill of material shortages. Not everything that is needed is readily available like before the pandemic.

Inflation should ease demand in 2H 2022 and into 2023. Getting people back to the office will help as well. Less time and less money to spend.

Unfortunately Covid is not going away anytime soon. Businesses near me are being shut down for days at a time as entire offices are getting sick. Entire cities in China are being shut down in an effort to contain new variants. At home Covid testing and not needing hospitalization if you do get sick is skewing the numbers.

I frequently sail past the port of Oakland in Northern CA. I have never seen it busier. I have radar on my boat so I can monitor traffic from my slip in Alameda.

Bottom line: Wafer manufacturing shortages are not going to be our biggest worry in the coming quarters.

This should be a new thread, the end of the world.
 
I have a feeling TSMC will be seeing even greater earnings than previously thought.

Samsung foundry is on life support it seems, as 3nm yields at 10%, and 4nm yields at 30%. We’ll soon see all Samsung chips being made by TSMC.

Just imagine what Intel yields are, given Intel is technologically behind Samsung, in EUV production experience.

https://wccftech.com/samsung-3nm-gaa-worse-yields-than-4nm/
 
  • Like
Reactions: VCT
I have a feeling TSMC will be seeing even greater earnings than previously thought.

Samsung foundry is on life support it seems, as 3nm yields at 10%, and 4nm yields at 30%. We’ll soon see all Samsung chips being made by TSMC.

Just imagine what Intel yields are, given Intel is technologically behind Samsung, in EUV production experience.

https://wccftech.com/samsung-3nm-gaa-worse-yields-than-4nm/
Samsung may not use TSMC directly to make Samsung designed chips, but they are using and will use more Qualcomm and Mediatek products that are made by TSMC. In a way it's a testimony of TSMC's "Everyone's Foundry" spirit.
 
Samsung may not use TSMC directly to make Samsung designed chips

I understand that Samsung may continue to use Qualcomm chips and perhaps MediaTek chips, but I’m referring to “greater earnings” from Samsung in-house designed chips as well, such as their Exynos line.
 
I understand that Samsung may continue to use Qualcomm chips and perhaps MediaTek chips, but I’m referring to “greater earnings” from Samsung in-house designed chips as well, such as their Exynos line.
Samsung is the combination of IDM, foundry, fabless company, design house, and consumer product (smartphones, TV, washers, PCs etc.) company. Samsung is in a situation trickier than Intel. At least Intel generally doesn't sell PCs or servers directly to the consumers. HP, Dell, Lenovo, Asus, and Acer don't treat Intel as a competitor.

In my opinion Samsung's strategy seems powerful but is full of conflict of interest.
 
Back
Top