Array
(
    [content] => 
    [params] => Array
        (
            [0] => /forum/index.php?threads/intels-weak-2021-sales-expectations.14628/
        )

    [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
)

Intel's Weak 2021 Sales Expectations

Daniel Nenni

Admin
Staff member
3Q21 Earnings Outlooks Bode Well for Most Leading Semi Suppliers
Memory suppliers, Sony, and TSMC benefitting from strong demand and supply shortages

IC Insights’ recently released its compilation of third-quarter sales growth expectations for the top-25 semiconductor suppliers. For the third quarter of this year (ending in September), sales growth outlooks for the top-25 suppliers range from 16th-ranked Sony’s 34% increase at the high end, to Intel’s 3% decline on the low end.

Third-quarter growth expectations for the top-15 suppliers are shown in Figure 1. In advance of an expected surge in demand for 5G smartphones during the upcoming holiday season, Qualcomm and Apple anticipate significant increases in their 3Q21 semiconductor sales. Also, the big three memory IC suppliers—Samsung, SK Hynix, and Micron—are each expected to post a 10% increase and Kioxia is anticipated to show an 11% jump in 3Q21 sales as demand remains strong for memory in data center servers, enterprise computing, and for 5G smartphones and related infrastructure.

In 2Q21, TSMC was the third-largest semiconductor company in terms of sales and the largest semiconductor foundry in the world. It is the go-to manufacturer for ICs based on leading-edge 7/5nm process technologies, which are in high demand and collectively accounted for about half of TSMC’s second-quarter revenue. The company reported that 18% of its 2Q21 revenue was from ICs made using its 5nm process technology while 7nm device technology accounted for 31% of its 2Q21 sales.

TSMC expects its 3Q21 sales to climb 11%. IC Insights forecasts its sales will rise another 4% in 4Q21. IC Insights believes TSMC’s second-half sales will be 14% greater than its first-half sales, yielding a full-year increase of 24%. If achieved, it would mark back-to-back years of more than 20% revenue growth for the foundry giant. TSMC’s annual sales increased a whopping 31% in 2020.

AMD was ranked as the 11th largest semiconductor company in 2Q21 and anticipates its sales will rise 6% in 3Q21. For the year, AMD expects to register an enormous 60% surge in sales. AMD’s sales have been rising quickly since 2Q20. The company is forecast to post sales of $4.10 billion and $4.23 billion in the third and fourth quarters respectively this year, resulting in second half sales that are 14% more than in 1H21. AMD uses TSMC to fabricate its new, Zen 3 processors using 7nm technology.

Top 15 Semiconductor Sales leaders 2021.png
Figure 1

A few leading companies—notably Intel and Texas Instruments—do not expect to see such rosy 3Q21 sales gains. TI is guiding for flat 3Q21 sales results. Intel, now the second-largest semiconductor supplier in the world, put its 3Q21 sales guidance at -3%, and its full-year sales guidance at -1% during its most recent earnings presentation (Figure 2). In a year when total semiconductor sales are forecast to grow 24%, this would be an especially weak showing for the processor giant. Moreover, Intel’s 2H21/1H21 sales expectations are also for a 1% decline. Intel’s 4Q21 sales are currently forecast to be 3% less than its 1Q20 sales, resulting in a two-year period of essentially flat results.

Collectively, the top 15 companies are forecast to experience 7% growth in 3Q21. Semiconductor sales are forecast to remain robust through the end of the year, which supports IC Insights’ current forecast of 24% growth for worldwide semiconductor sales this year.

Intel Weak Sales Expectations.png
Figure 2

Report Details: The 2021 McClean Report

The 2021 edition of The McClean Report—A Complete Analysis and Forecast of the Integrated Circuit Industry was released in January 2021 and the Mid-Year Update in July 2021. A subscription to The McClean Report includes free monthly updates from March through November (including a 180+ page Mid-Year Update), and free access to subscriber-only pre-recorded webcasts through November. An individual user license to the 2021 edition of The McClean Report is available for $5,390 and a multi-user worldwide corporate license is available for $8,590. The Internet access password and the information accessible to download will be available through November 2021.

https://www.icinsights.com/services/mcclean-report/pricing-order-forms/

More Information Contact
For more information regarding this Research Bulletin, please contact Bill McClean, President at IC Insights. Phone: +1-480-348-1133 email: bill@icinsights.com

PDF Version of This Bulletin
A PDF version of this Research Bulletin can be downloaded from our website at https://www.icinsights.com/news/bulletins/
 
The worldwide semiconductor sales will grow a huge 24% in 2021 but Intel and TI are the only two companies who won't see any growth among the top 15 players.

With 13 out of the top 15 semiconductor companies enjoying good growth in 2021, I believe it means the demand is strong and the growth goes across multiple semiconductor segments. But why Intel's 2021 revenue will go downward with a -1% growth year over year? Intel's is facing a lot of challenges, but this negative revenue growth is a very serious situation considering 2021 is another prosperous year for the whole semiconductor industry.
 
Last edited:
The worldwide semiconductor sales will grow a huge 24% in 2021 but Intel and TI are the only two companies who won't see any growth among the top 15 players.

With 13 out of the top 15 semiconductor companies enjoying good growth in 2021, I believe it means the demand is strong and the growth goes across multiple semiconductor segments. But why Intel's 2021 revenue will go downward with a -1% growth year over year? Intel's is facing a lot of challenges, but this negative revenue growth is a very serious situation considering 2021 is another prosperous year for the whole semiconductor industry.

Do you think IDM 2.0 will change Intel's prospects?
 
Why won't TI grow? The world is going more analog IC.
I think analog EDA tools are improving more than digital.
 
Do you think IDM 2.0 will change Intel's prospects?
No, I don't think so. IDM 2.0 may improve people's perception on Intel, but it's creating more challenges than Intel can handle.

It's not because Intel is lacking capital and government subsidies. It's not because Intel can't advance their manufacturing technology. It's not because Intel doesn't have smart engineers.

What Intel needs is the right leadership. I mean Intel needs the right CEO, board of directors, and Chairman. Intel's board are full of nice people but several of them are, IMHO, more suitable for a mid sized company or a Internet startup.

Although I have to admit the current board is better than before. But Intel is running out of time to allow thier leadership to learn or to improve the quality of their decision.

Out of my head, the following is a short list of Intel's wrong decision (endorsed by Intel's board and probably recommended by the past or current CEO):

1. Around 2004 and 2005, Intel rejected Apple's proposal to manufacture processors for iPhones. It probably costed Intel $50 to 100 billion of lost revenue.

2. Intel sold its own ARM product division to Marvel for $600 million in 2006. One of the reason was Intel believed its x86 technology will produce better performance and low power usage processors for mobile devices such as tablets and smartphones. Really?

Intel started ARM development with the purchase of DEC's StrongARM division in 1998. By the time Intel sold its ARM division to Marvel, iPhone and other smartphones had been selling crazily for a year already. Why Intel started early on ARM technology but kicked it out while other companies were making huge revenue and profit from ARM based products?

This easily can be another $100 billion mistake.

3. In 2010, Intel paid $7.7 bullion to acquire anti virus software company McAfee. Why? I didn't understand the reason behind this acquisition back then and I don't understand it even today.

4. Around 2015 and 2016, in order to push slow and power hungry x86 based mobile processors to manufacturers, Intel paid them cash to accept those inferior products for free. It resulted astonishing negative revenue for Intel's mobile division for several quarters. Intel called it "Contra Revenue". Once Intel's subsidies stopped, so did customers' purchase. This can be easily a $10 billion mistake.

5. In 2021, Intel released advertisment to attack Apple and its "inferior" products. Somehow Intel's leadership forgot Apple is a major Intel customer, even today.

Then new Intel CEO Pat Gelsinger started attacking TSMC publicly about how unreliable and unsafe the situation is. Somehow he forgets TSMC is not Intel's enemy. Especially for the next several years TSMC will be a critical partner of Intel's revival.

Intel is probably the only major corporation who is willing to spend time and money to attack their own customer, partner, and supplier. How bizarre it is!

Why Intel's board and CEOs allowed these billion dollar mistakes happened repeatedly? Without the right leadership, Intel's biggest enemy is itself.
 
Last edited:
Look at the decline of Microsoft office. Microsoft is a cloud company now.
 
Also, I think Intel took too much risk with the hyperscale-strategy (basically 10nm Cannon Lake). Management took far too much risk; thinking they can pull of 4 (or so) simultaneous new technologies.

It included too many changes at once which all had to go right, like jumping through 13 misaligned burning hoops; and then the core-design being totally dependent on that very excercise.
Release date is set by successfully reaching the technological targets.

Compare to TSMC: They have the agile-methodology; where the technological targets are set by the release date. That way, you never have much delays.

This basically means Intel has lost their unique selling point. Basically, 14nm has been Intels cash cow the last few years; and was accepted because products were better than the competition.
If you check 10Q's the last few quarters, there's always explanation of "less profit" blamed of ramping up 10nm (and unqualified stock). Intel doesn't want to make 10nm.

I think Intel has learned from their mistakes here; design is a bit decoupled from the process, and there are TSMC-agreements for Intel process-problem mitigations. The latter thanks to Bob Swan I believe.

However, because Intel has lost its unique selling point (process advantage) and slowly 14nm-products are less and less accepted by clients for the price Intel wants to sell them for, margin compression is the only result if a company fails to diversify.

Sure, Intel tried; almost all diversification-projects failed. I think the transformation from "CPU-centric to data-centric strategy" is pretty much dead.
Mobileye and FPGA's amount to <3,5% of Intel's profits.

Apple and Microsoft successfully transform from hardware / software companies towards platform / service companies, Intel didn't.

It's like a ship with passengers and the captain being a financial engineer in a storm while there's a hole in the ship:
To stuff up quarterly results, the captain will throw passengers overboard to keep the company floating for stockholder perception. But the hole isn't gone, and after Intel withdrew from mobile, IoT, McAfee, Windriver, 5G and NVM-business; there are no passengers left.

Also, when TSMC was upping capex, Intel was upping the share buyback program; that didn't help for an Intel-favourable distribution of EUV-tools either.
 
1. Around 2004 and 2005, Intel rejected Apple's proposal to manufacture processors for iPhones. It probably costed Intel $50 to 100 billion of lost revenue.

2. Intel sold its own ARM product division to Marvel for $600 million in 2006. One of the reason was Intel believed its x86 technology will produce better performance and low power usage processors for mobile devices such as tablets and smartphones. Really?

Intel started ARM development with the purchase of DEC's StrongARM division in 1998. By the time Intel sold its ARM division to Marvel, iPhone and other smartphones had been selling crazily for a year already. Why Intel started early on ARM technology but kicked it out while other companies were making huge revenue and profit from ARM based products?

This easily can be another $100 billion mistake.
To be pedantic, iPhone launched in 2007, one year after this sale closed. Intel's largest customer for their XScale + modem business at the time was RIM/BlackBerry with PXA901/905 (Hermon), with some side business at Samsung. I'm not sure if their PXA270 (Bulverde) processor as a stand-alone AP ever saw much volume. Both of those decisions rested on Otellini, who then turned around and purchased another cellular modem solution (Infineon) in 2010/11 ostensibly to get back into the iPhone socket, just through a different vector. Unfortunately by that time Apple had switched to Qualcomm for modem solutions.
 
To be pedantic, iPhone launched in 2007, one year after this sale closed. Intel's largest customer for their XScale + modem business at the time was RIM/BlackBerry with PXA901/905 (Hermon), with some side business at Samsung. I'm not sure if their PXA270 (Bulverde) processor as a stand-alone AP ever saw much volume. Both of those decisions rested on Otellini, who then turned around and purchased another cellular modem solution (Infineon) in 2010/11 ostensibly to get back into the iPhone socket, just through a different vector. Unfortunately by that time Apple had switched to Qualcomm for modem solutions.
Thank you for the correction about the first iPhone release date was actually after Intel sold its ARM division to Marvel.

It's really strange to me that despite the early start and all the money and talents Intel had, they couldn't figure out a way to do business in the world of ARM. There are many companies, except Intel, built big business around ARM and made huge profit from ARM based products. Is it because Intel's culture or something else?
 
The thing that Intel investors miss out on is that in order to transform Intel it has to bleed a lot of money. The ARC GPU's won't be very profitable for a few years, the CPU's will have lower margins in order to compete. All that is happening while Intel is investing in new fabs, tools and reorganizing itself. I don't think it will be able to compete with TSMC, but maybe the new generation of EUV machines will save it.
 
The thing that Intel investors miss out on is that in order to transform Intel it has to bleed a lot of money. The ARC GPU's won't be very profitable for a few years, the CPU's will have lower margins in order to compete. All that is happening while Intel is investing in new fabs, tools and reorganizing itself. I don't think it will be able to compete with TSMC, but maybe the new generation of EUV machines will save it.

From what I have heard Intel and TSMC will get the new EUV machines at the same time so I see no advantage there for Intel. In fact I would bet TSMC has them in production a year or two before Intel.

But I do agree that Intel stock will get worse before getting better. They have been propping up the stock with buybacks which will have to stop. Huge amounts of investment will be required to compete with TSMC.

Samsung did well against TSMC with 14nm and 8/7nm but at 5nm and 3nm TSMC is killing it. It will be interesting to see what TSMC, Intel ,and Samsung do for an encore but TSMC certainly has the advantage now with the shear amount of customer traction, my opinion.

The only big variable I see is if Intel or Samsung acquire GF. That would be disruptive, absolutely.
 
It's really strange to me that despite the early start and all the money and talents Intel had, they couldn't figure out a way to do business in the world of ARM. There are many companies, except Intel, built big business around ARM and made huge profit from ARM based products. Is it because Intel's culture or something else?

From the stuff I read, Intel mainly was incompetent at delivering integrated SoCs where Samsung and Qualcomm succeeded.

Back at the time, you had Intel Core CPU for desktop, barely able to decode h.264.

At the same time, smartphones had video accelerators in the SoC, so the smartphone could better play video than desktops; even first Raspberry Pi was better than some Intel CPU's

AMD was first with memory controller integration, took Intel years to catch up.

Intel didn't lose to ARM, but to the whole SoC / ARM / TSMC ecosystem.

Intel has and probably will always be a CPU-centric company.
 
Intel didn't lose to ARM, but to the whole SoC / ARM / TSMC ecosystem.

A long time ago, maybe around 2009, I saw it as a battle of Intel (x86) vs ARM (SoC) and predicted ARM would eventually win. However, realized, with the help of this site and Daniels wonderful book Fabless, that is really a battle between the IDM and Fabless business models. No IDM can compete against the strength of the fabless ecosystem.
 
From the stuff I read, Intel mainly was incompetent at delivering integrated SoCs where Samsung and Qualcomm succeeded.

Back at the time, you had Intel Core CPU for desktop, barely able to decode h.264.

At the same time, smartphones had video accelerators in the SoC, so the smartphone could better play video than desktops; even first Raspberry Pi was better than some Intel CPU's

AMD was first with memory controller integration, took Intel years to catch up.

Intel didn't lose to ARM, but to the whole SoC / ARM / TSMC ecosystem.

Intel has and probably will always be a CPU-centric company.
"Intel didn't lose to ARM, but to the whole SoC / ARM / TSMC ecosystem.

Intel has and probably will always be a CPU-centric company."


Is it still possible for Intel to build a competitive, relevant, and financially sustainable ecosystem under their IDM 2.0 strategy?

If not, what should Intel do? Join the crowd?
 
Wintel died and Microsoft became a cloud company. If Intel reinvents itself it will be in israel. Look at Chromebooks there are samsung, mediatek, amd, and soon google. Apple has their own. They still dominate the wintel workplace platform but there are other options for word, excel or powerpoint.
 
"Intel didn't lose to ARM, but to the whole SoC / ARM / TSMC ecosystem.

Intel has and probably will always be a CPU-centric company."


Is it still possible for Intel to build a competitive, relevant, and financially sustainable ecosystem under their IDM 2.0 strategy?

If not, what should Intel do? Join the crowd?
I think that IDM 2.0 is very smart strategy if Intel can execute flawlessly. But I don't think they will execute flawlessly and their window of opportunity will close in like 5 years.
 
Back
Top