Array
(
    [content] => 
    [params] => Array
        (
            [0] => /forum/index.php?threads/samsung-electronics-to-make-%E2%80%98meaningful%E2%80%99-cut-to-chip-production-in-effort-to-support-prices.17736/
        )

    [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
)

Samsung Electronics to make ‘meaningful’ cut to chip production in effort to support prices

Daniel Nenni

Admin
Staff member
Samsung Electronics said on Friday it would make a “meaningful” cut to chip production, following the lead of smaller rivals, as it grapples with a sharp global downturn in semiconductor demand that has sent prices plummeting.

The unusual output cut by the world’s biggest memory chipmaker — with no previous announcement recalled by Samsung officials and analysts — came after it flagged a worse-than-expected 96% plunge in first-quarter profit.

Investors brushed off the profit miss, betting the move by the industry leader would support chip prices that had fallen by about 70% over the last nine months.
Samsung jumped 4.5% in early trading in the biggest one-day rise since September, while rival SK Hynix’s shares surged 5.6%.

Smartphone and personal computer makers had stocked up on chips during the pandemic when demand for consumer devices surged, but they are now running down inventories as shoppers cut back on purchases amid rising inflation.

Samsung said memory demand had dropped sharply because of a weak global economy and customers slowing purchases as they focused on using up their stocks.
“We are lowering the production of memory chips by a meaningful level, especially that of products with supply secured,” it added, in a reference to those with sufficient inventories.

Read more about tech and crypto from CNBC Pro​

Samsung did not disclose the size of the planned production cut, but it sent a strong signal for a company that had previously said it would make small adjustments like pauses for refurbishing production lines but not a full-blown cut.

“The fact that the No. 1 market share firm is joining production cuts lifted shares... SK Hynix and Micron have declared production cuts, but only Samsung had not, so the market was watching for it,” said John Park, an analyst at Daishin Securities.

“Today’s production cut signal casts a positive outlook for a memory chip rebound in the second half of the year.”
Although cutting short-term production, Samsung said it was still making long-term investments in infrastructure and research to secure needed clean rooms for chip production and expand its technological lead.

It did not say how its 2023 investment plans would be affected, having previously flagged capital spending similar to the 53.1 trillion won investment in 2022.

Expect Samsung to focus more on U.S. than China in terms of demand and supply, analyst says


Samsung may focus more on U.S. than China in terms of demand and supply: Analyst

SK Hynix said in October it would more than halve its capital spending in 2023 versus 2022, while Micron cut fiscal 2023 investment plans by more than 30% in September.

Record chip loss​

Samsung estimated its operating profit fell to 600 billion won ($455.5 million) in January-March, from 14.12 trillion won a year earlier, in a short preliminary earnings statement. It was the lowest profit for any quarter in 14 years.

The first-quarter profit fell short of a 873 billion won Refinitiv SmartEstimate, weighted toward analysts who are more consistently accurate. Multiple estimates were revised down earlier this week.

Its chip division is likely to report a record loss of 2.1 trillion won ($1.6 billion), according to an average of analyst forecasts, and post another 2 trillion won loss in the current quarter, a major divergence for what had been Samsung’s most important cash cow, generating about half of its profits in better years.

Analysts said Samsung’s production cut might improve its performance slightly in the current quarter and could also cement or hasten the rebound of memory chip prices.

“Samsung talking about production cuts is evidence of how bad the current slump really is,” said Greg Roh, head of research at Hyundai Motor Securities.

The company is due to release detailed earnings, including divisional breakdowns, later this month.

 
This is the same company that just weeks ago insisted they will not cut a penny in their capex and they will not reduce production. They seem like to flip flop. I guess could be part of their strategies to deceive competitions while leveraging their size and deep pockets (including gov help)?
 
In my opinion, it is pessimistic sign from Samsung. The production cut announcement is just a stimulation to slow down the price drop in short term. I believe Samsung might think it will be short term correction (2-3 Q) last year but it seems realize the correction will be longer and deeper now, even Samsung can not afford the loss and need to take actions.
 
Very significant news in my estimation. Turns out Samsung isn’t immune and can’t just chug through the cycle incurring losses all the while. TSMC’s earnings on the 20th will be very interesting to say the least. I wonder what they are seeing in their business atm and for the rest of the year.
 
As previously mentioned, Samsung semiconductor revenues are still mainly depending of the memory market representing around 3/4 of their business (76.1% in 2021 then 69,6% in 2022 despite their new pure foundry business still a bit marginal well below TSMC…)…
This strongly contributed in 2021 to lead Samsung to become top 1 semiconductor company above Intel, but then in 2022, as the memory was already strongly decreasing, it allows TSMC to become top 1 semiconductor company above Samsung now going back to top 2…
But in 2023, the semiconductor market is expected to globally decrease/collapse even with some stabilisation in the memory market which has always been very volatile since the last century with regularly/often very strong collapses which the biggest related companies, especially Samsung, have always wanted to strongly reboot with a model very specific to this segment, systematically pushing huge capacity limitations in order to decrease their costs and also regularly contribute to boost back their price and customers’ demand…
Concerning the other/global semiconductor markets, the situation in Q1 2023 is already quite negative, including for example,
(i) UMC revenue already decreasing -14.5% YoY in Q1 2023…
(ii) TSMC still remaining positively growing but with now already much smaller YoY growth at only +3.6% in Q1 2023 (as their main supplier Apple has decreased their computers shipments by -40% in this trimester Q1 2023 which are now mainly using TSMC most advanced technologies which are at the highest price/cost per wafer…)…
 
It's a bit surprising because AI training will likely need many many exaflops of compute, and fast memory, in the near future.

By the way how big could a 'meaningful' cut be? Going from 24/7 to 16/7?
 
Samsung Semi's performance over the past 5 years has been on the wane, despite the record profits last year. Before the current crisis they already lost their lead on DRAM and NAND cost & technology and they've leaked key production talent to Hynix & China.
Their decision to remain aggressive in production in 2H '22 has turned out to be a disaster. True, this was often their playbook in the past but it didn't work this time and they have been too slow to accept the mistake and correct course.
We are not quite at the "Change everything except your wife and kids" stage but if there is not a major clear out & reset at Samsung for '24, then the position they have held with all the customers as the premier memory supplier will disappear and much of their profits with it.
 
As EUV revived logic scaling, a breakthrough capacitor is needed in DRAM. I think it's hard to produce breakthroughs in old organizations like Samsung, or Micron, and look more to YMTC for leadership. Yet the US has hobbled that rising group.

I think we need to revive Sematech, as poor of a solution as that was, at least it offers the possibility to address physics and chemistry challenges before it is too late.
 
I would never count out Samsung long term. From leapfrogging the Japanese in DRAM by skipping a node and the first foundry finFET, every single time their backs were against the wall they always pulled a rabbit out of a hat. The nearterm does look rough, and to put it mildly I am skeptical that Samsung will achieve their goal and surpass TSMC on the foundry front this decade given how distant their 2nd place position is/having intel nipping at their heels. Memory is more of a coin flip. If they can leverage their extra EUV expense over SK and Micron and actually succeed on their plan to skip ahead a node for DRAM, then things will start looking really nice for SE. In spite of all of the headwinds, their massive scale and manufacturing excellence is still allowing them to be cost competitive even with inferior technology (a big accomplishment in my opinion). Long term I would love see (and also be terrified of seeing) a Samsung that not only has mastery of manufacturing but also technology.
 
I would never count out Samsung long term. From leapfrogging the Japanese in DRAM by skipping a node and the first foundry finFET, every single time their backs were against the wall they always pulled a rabbit out of a hat. The nearterm does look rough, and to put it mildly I am skeptical that Samsung will achieve their goal and surpass TSMC on the foundry front this decade given how distant their 2nd place position is/having intel nipping at their heels. Memory is more of a coin flip. If they can leverage their extra EUV expense over SK and Micron and actually succeed on their plan to skip ahead a node for DRAM, then things will start looking really nice for SE. In spite of all of the headwinds, their massive scale and manufacturing excellence is still allowing them to be cost competitive even with inferior technology (a big accomplishment in my opinion). Long term I would love see (and also be terrified of seeing) a Samsung that not only has mastery of manufacturing but also technology.
Samsung's EUV set being used for both DRAM and foundry? That could be tricky..
 
Samsung's EUV set being used for both DRAM and foundry? That could be tricky..
You would be correct.

If you take Samsung at their word they’ve been using EUV DRAM for a couple of years now (since like 1z or something like that). But both Samsung and SK seem to be struggling to get large volumes of their EUV nodes to market. Meanwhile Micron went full speed ahead with SAQP on 1-alpha and 1-beta before the others, with EUV supposedly coming with the next generation. Given how Samsung has been working on EUV logic and DRAM longer than SK or Micron, I would give Samsung the advantage for cracking EUV DRAM HVM first, but you never know.
 
Back
Top