China's SMIC Foundry Fumbles, 80% Decline In Profit

News
By
published

China's premier fab hits a snag.

SMIC Shenzhen
(Image credit: SMIC)

Semiconductor Manufacturing International Corporation (SMIC) just posted its Q3 earnings report, and its profit declined 80% compared to last year (via CNBC). As China's largest foundry, SMIC represents the cutting-edge of the country's semiconductor fabbing industry, and it's clear that things aren't going super well.

Global factors are partly to blame for SMIC's poor showing in Q3. The global economy isn't in great health due to high inflation, which has resulted in lower consumer spending. In China, inflation hasn't been a problem, but consumers are decreasing spending as they focus on paying off debt. SMIC also has to contend with U.S. sanctions over Chinese silicon.

Still, SMIC's decline in profit is particularly notable compared to its rival TSMC. While TSMC also experienced a decline in profit, it was a mere 12% as opposed to 80%. Regarding technology, SMIC is significantly behind TSMC, as the former is only just getting to 7nm while TSMC is working on its upcoming 2nm node. A sharp decline in profitability will likely impede SMIC's attempt to catch up.

SMIC hasn't ever been a cutting-edge foundry, launching its 28nm node in 2015, the same year TSMC reached 16nm. When SMIC launched its 14nm process in 2019, TSMC had been offering 7nm for about a year. Today, although SMIC is apparently going to be able to progress to 7nm and 5nm, by the time that happens it's likely that American-friendly foundries like TSMC, Intel, and Samsung will have sub-2nm processes. 

It's also bad news for China's ambitions to build up a native semiconductor industry. If China wants to be able to produce chips as fast as those designed by companies like AMD and Nvidia, and fabbed at TSMC or Samsung, then cutting-edge nodes are going to be needed at foundries like SMIC.

Matthew Connatser
Matthew Connatser

Matthew Connatser is a freelancing writer for Tom's Hardware US. He writes articles about CPUs, GPUs, SSDs, and computers in general.

16 Comments Comment from the forums
  • Order 66
    I can't say I'm all that surprised, with the massive reduction in sales that the whole PC industry had from its pandemic highs.
    Reply
  • George³
    Of course, this site is not about economics, and maybe because of that, when it publishes news related to finance, it misses some words. Which change the information. The cash flow is not reduced by 80%, there are simply increased development and implementation costs and what has decreased is the net profit. The term "net profit" should be present here instead of just profit.
    Reply
  • The Historical Fidelity
    George³ said:
    Of course, this site is not about economics, and maybe because of that, when it publishes news related to finance, it misses some words. Which change the information. The cash flow is not reduced by 80%, there are simply increased development and implementation costs and what has decreased is the net profit. The term "net profit" should be present here instead of just profit.
    Profit is the amount of revenue minus operating costs and any reinvestment costs. No need for “net profit” as it’s a given. Only when “gross profit” is used, then you must differentiate with “net profit”. Otherwise “net profit” is assumed.
    Reply
  • shady28
    Order 66 said:
    I can't say I'm all that surprised, with the massive reduction in sales that the whole PC industry had from its pandemic highs.

    Yep, consumer electronics in general, smartphones in particular.

    "Oct 16 (Reuters) - The global smartphone market contracted by 8% to its lowest third-quarter level in a decade on subdued demand for major brands including Apple (AAPL.O) and Samsung (005930.KS) in most developed markets, according to data from Counterpoint Research.

    The data, shared exclusively with Reuters, showed that the share of the top five brands, which also include Chinese firms Xiaomi (1810.HK), Oppo and Vivo, had fallen to a three-year low."

    https://www.reuters.com/technology/global-smartphone-market-slumps-lowest-q3-level-decade-counterpoint-2023-10-16/
    Reply
  • George³
    It is good that there is a link to the original news, which is longer and contains information that is missing in the material published here. Economics is a complex subject and can lead to incorrect conclusions when there are certain shortages.
    Reply
  • ivan_vy
    "Revenue: $1.621 billion, vs. $1.625 billion expected"
    so almost hit target, could be lower Net income due higher than expected cost of BOM and equipment caused by sanctions and bigger spending in R&D. Still on good numbers.
    Also not having advance node is not so big deal as tech can be backported to some extent and also depends where is their TAM, I mean, if they sell billions of mid-range chips don´t need cutting edge tech but good enough will suffice.
    The important thing is not lose the cadence, is not so important to be a generation behind as long as the gap remains at 3-4 years top, not everyone buys the cutting edge, it makes no sense in Economies of scale, best bang-for-the-buck will determine if the keep going.
    Reply
  • bit_user
    In China, inflation hasn't been a problem, but consumers are decreasing spending as they focus on paying off debt.
    I've hear China has been in a deflationary spiral, lately. That explains reduced spending due to people sitting back and waiting for prices to fall some more. Of course, it's triggered by reduced spending to begin with, so there are pretty much always external factors involved.

    According to the CNBC article, revenue is only down 15%. So, that 80% figure is probably misleading to some.
    Reply
  • George³
    ivan_vy said:
    Revenue: $1.621 billion, vs. $1.625 billion expected

    bit_user said:
    According to the CNBC article, revenue is only down 15%.
    I think you can work out for yourself if it's 15%. I don't understand numbers, but it doesn't seem like more than 0.25% to me.
    Reply
  • bit_user
    ivan_vy said:
    "Revenue: $1.621 billion, vs. $1.625 billion expected"
    so almost hit target,
    Hitting their target doesn't mean their fortunes are good - just that they're not worse than they previously expected.

    The key things to watch are like year-on-year revenue, and then compare this with their competition.

    ivan_vy said:
    not having advance node is not so big deal as tech can be backported to some extent
    Backporting results in expensive chips that run hot and/or at reduced clockspeeds. It's hardly a good solution.

    ivan_vy said:
    not everyone buys the cutting edge, it makes no sense in Economies of scale, best bang-for-the-buck will determine if the keep going.
    While there's plenty of business on older nodes, if China's access to newer nodes is restricted, then so is the competitiveness of products requiring those chips which do benefit from being on the latest, greatest node.

    Being behind the curve has another downside, which is that they'll have to compete with yields and wafer pricing on uber-mature, refined nodes like TSMC's N7. Margins on such nodes are clearly lower. I'd bet TSMC makes most of its revenues on the couple generations at the leading edge. That probably funds most of their R&D, and then the business they do on older nodes is probably very profitable for them.
    Reply
  • bit_user
    George³ said:
    I think you can work out for yourself if it's 15%. I don't understand numbers, but it doesn't seem like more than 0.25% to me.
    You're confusing year-on-year revenues with expected revenues.

    They don't make revenue projections by always blindly assuming they'll be the same as that quarter from last year. Revenue projections are their best guess, made (I think) last quarter. If trends are pointing to lower revues, their projection could indeed be down. If business is looking good, they'll be up.
    Reply