Intel’s investment value

[user_2013101]

Despite the evidences of shrinking leads in the process technologies, INTC, in my view, remain bullish short-term and neutral long-term, mainly because no meaningful challenges to Intel’s monopoly in PC/server.

[h=3]Short-term outlook 1-2 years [/h]
Windows 10, if no delays in its release and not as bad as Windows 8, is likely to bring another wave of upgrade demands, starting in the Fall. This is similar to the PC upgrades resulting from the termination of Windows XP support.

Atom X3, previously code-named SoFIA, will start shipping in 2Q or 3Q. Yes, I know X3 delivers only mediocre performance. But, from an investor’s, not technical, point of view, the chip has integrated connectivity and thus, reduces the bill of materials – read: “contra revenue.”

Last year Intel lost $4.2 B in the mobiles. If X3 cutting the contra revenue by a modest 30-50% will significantly improves the company’s profitability.

SoFIA is also Intel’s first real attempt into the smartphones.

A short-term trade: A low-risk buying opportunity may be at hand, if INTC corrects to a summer low, perhaps under $30. And, sell the shares early next years.

[h=3]Long-term outlook 3-5 years [/h]
The picture is not so clearly bullish.

It depends on if Intel penetrates the mobiles and IoT, and if the ARM camp can make inroads into PC/server. Currently, it is a stalemate. Neither camp can advance meaningfully into the other side.

Yes, I am fully aware that several companies, including QCOM, BRCM, AMD, etc., have plans of ARM-based servers. AMCC’s ARM-based X-GEN2 will start shipping in 2Q15. However, none of them look to be creditable threats to Intel.

Similar situation in PCs. The proliferation of increasingly powerful mobile devices limits the growth of PCs. But, nothing can replace Intel’s role in the PC processors – at least not yet.


[h=4]Disclaimer [/h]
The posters are not licensed financial advisers. The forum posts are an expression of personal opinions, and does not constitute recommendation of investments whatsoever.

 

[ta152h]

Any effect from Windows will be relatively minor, as it normally is. I would not at all put this on par with Windows XP, as businesses will almost certainly not be delaying purchases to get Windows 10 (it's a free upgrade, in any case), and I can't see consumers getting excited about a new version of Windows, at least not in material numbers. Again, even if they were considering waiting, it's a free upgrade, so why would anyone?

You're confused by contra-revenue. It's not a phone thing, it's a tablet thing, as given by the fact they shipped almost nothing in phones in 2014, and lost $4.2 billiion. Your estimates are also very optimistic, since Intel believes they will lose 800 million less than last year, or around $3.4 billion. Of course, they might decide to abandon the lower-end tablet market, particularly if it shrinks this quarter like it did last quarter, but so far Intel has given no indication of this.

It's difficult to call it a stalemate, when the ARM camp has not even made an attempt. We'll see in 2016 whether they can gain and market share, but clearly before then there is no real chance, because there are no real products. Products like Seattle are just to seed the market, not generate significant sales.

Obviously, Intel has not made a real attempt in phones, which they also will not until 2016 (SoFIA LTE2 and MID will be the beginning of the story, everything before that will just be a prologue). In tablets they have certainly made a real attempt, and failed so ignominiously, it's difficult to imagine. How does one generate negative revenue, and then blame it on BoM?

Oh, and if you believe that's why the products lost money, I have a bridge in Brooklyn that I'm willing to let go cheap. By Intel's own estimations, their BoM deficit was $15 a chip, with Bay Trail. With Bay Trail-CR, it was significantly lower. Yet, they generated negative revenue in Q4. So, what does that say about the value of the chip? It's worth less than $15?

Then, we are told the company will lose north of 3 billion this year, but Cherry Trail will not have any BoM deficit. So, this makes sense how? Bay Trail-CR didn't lower the losses, and Cherry Trail will still generate deficits (even granted it will not be out the full calendar year), despite no BoM deficit. Simply put, Intel is lying, and there are actually people that believe it. That's scary, since only a cursory examination reveals how false it is.

Intel's 2015 concerns center around OpenPOWER, and POWER8. POWER8 has significantly better performance than anything Intel makes, and is much more open. More than that, the Chairman is from Google, which is a Platinum member. CAPI is way ahead of anything on x86, giving co-processors the same access to cache and memory as the CPU. NVIDIA will be moving NV-LINK, with IBM, to the platform, giving GPGPU on that platform the best of both APUs, and discrete cards.

Other platinum members include Samsung, Tyan, NVIDIA, Sk Hynix, Altera, Powercore, etc... Also, companies like Powercore are supported by the Chinese government, and will be able to make new processors based on the POWER instruction set. So, this is a formidable array of companies, and with the tacit approval of the Chinese government, this looks like it has a chance. Here's a link to an article on it -> IBM's OpenPOWER Foundation: What Is It And Does It Matter? - International Business Machines Corporation (NYSE:IBM) | Seeking Alpha

Also, Intel has to worry about AMD's new processors in 2016. Given they haven't had a good history the last decade, it's still something to be concerned with. K12 will be ARM, so probably will not have much immediate impact due to the immaturity of that platform, but Zen is x86, and if a competent processor, could take a material amount of market share from Intel. Consider also that AMD has visibility into ARM's designs now, and may have learned something that could help with areas they have been historically weak at, like memory controllers, and dreadfully slow caches. Jim Keller has also indicated that development of K12 has yielded new insights they have directly used in improving the Zen design. So, there's hope. One thing seems close to certain - their competitive situation will improve dramatically. Mainly because they are still on 28nm, and the Bulldozer line is poor, at best.

I would also expect Apple to move from x86 to ARM at some point, maybe to K12 if it's very good. Staying on x86 seems very implausible, as they certainly do not wish to maintain two software stacks, and they also would prefer more control over the product line. It's not clear if the Mac sells enough devices to warrant the investment in a higher-end chip, which is why I think there's a small chance they'll use K12 and possibly move to their own devices later on, if the situation warrants it. It would also allow them to use their lower-end chips in some of their lower performance Macs, while using the K12 in bigger devices. So, it's possible. What's likely is them moving from x86, one way or another.

I wouldn't touch Intel, not until they show some kind of vision, and a substantial and profitable position in a growing market. I'm also not at all sanguine on their fabs going forward, as the growing mobile market, which they are not participating in, fuels the development resources of her competitors. Put another way, if Intel could not win in mobile when they were on FinFets, and the rest of the market was on planar, how will they close the gap with a processors (Airmont) that is essentially unchanged, while their opponents have newer processor designs, built on FinFets? It's not clear to me where their opportunity is, unless I actually believed their BoM nonsense. Which I don't.

 

[hist78]

Both of you have good points. But what are the options available to Intel?

First: Intel might already accept the fact that they can't be a dominant player in every product category. They may also recognized that they can't just walk away from certain product categories even they will have a smaller market share or a smaller profit margin compare to the X86 processors.

Second: Intel's senior leadership team is full of MBAs. They are probably debating what kind of Intel should be in the next ten years. Can Intel become a hybrid IDM one day? They can have in house design teams but pick and choose internal or external manufacturing providers based on time to delivery, scale of economy, CAPEX, cost, profit, risk, technology, IP, capacity, competition, and efficiency.

Intel is a great company because they have great products. Does that really matter who actually "manufactured" their products? At least in the mind of many MBAs, it's not that important as long as the product can make a lot of money for Intel.

TSMC is already doing production work for Intel on the SoFIA line. But I believe it's mainly due to the Infineon acquisition.

Looking forward, can a foundry like TSMC partner with Intel in a much bigger and significant role?

 

[user_2013101]

Any effect from Windows will be relatively minor ....

Thanks for your inputs.

The free upgrade to Windows 10 is only for the users of Windows 8, if I understand it correctly. A lot of people delayed or avoided PC purchases because they don’t like Windows 8. IF the new Windows can improve over the old one, additional demands will materialize.

I mean SoFIA’s “integrated connectivity,” instead of “integrated baseband.” Such integration does reduce BoM, even among tablets.

It’s up for debates just how much of Intel’s mobile losses result from the BoM deficits, and the contra revenue to compensate such deficits.

Thanks for the POWER8 info. It’s interesting, but still remote. AMD’s past records are not good. As an investor, AMD and POWER8 are not big factors in the equation – not yet. I wish AMD well; but it needs to deliver.

It IS a concern that Apple may use its own AP in the Mac lines. But, I don’t think it will happen in the next two years, and thus does not affect my short-term bullish outlook.

INTC still generates enormous profits, from its dominant PC position, and pays dividends. I think, hi-tech investors cannot ignore INTC, even though the company probably should not be the top holding.

 

[benb]

I think the cross-roads in the coming 10 years will be the end of scaling, at least economically beneficial scaling, at which time semiconductor manufacturing becomes a commodity. Having scale will matter. We see that already with scale that Samsung + GF offers partners being more attractive than the single advanced fab TSMC offers. The Intel virtual factory (two HVM and one low-volume leader fab) anticipates this. The economics will change. Once fabs are expensive but static in their technology, there will be a rush of new entrants and the more expensive manufacturing models will have to change. Intel is poorly equipped for this.

I think the question isn't ARM vs. x86 vs. Power vs....there will always be a new platform, the next new thing. It's whether advanced manufacturing can survive in the US, or other developed countries such as Taiwan and Korea. It could all end up in China, and we all know what that means will happen to IP and innovation as a source of competitive advantage. But then again, IP may not matter, once scaling ends, and there may be a race among chipmakers to set up manufacturing in the lowest cost country.

If you are Samsung or Intel, you already have a fab in China. But TSMC is in an awkward position, being in "China" but not China.

 

[hist78]

If you are Samsung or Intel, you already have a fab in China. But TSMC is in an awkward position, being in "China" but not China.

I think certain things need to be clarified:

1. TSMC has majority of their fabs in Republic of China (ROC, founded in 1911, moved to Taiwan after communists took over mainland China in 1949, ROC is frequently and informally called Taiwan).
2. TSMC also has a wholly owned fab in Shanghai of People's Republic of China (PROC, started in 1949).

So no matter which China you are talking about or how you want to define or agree the word of "China", TSMC is there in "China".

 

[benb]

Hopefully I didn't inadvertently make an offensive statement, that was not my intention. If I did, I apologize.

I have noticed Samsung's investment in mainland China grow in the recent years, while TSMC and Intel's have been dormant, so much so that I forgot about TSMC's fab in Shanghai altogether. I have also noticed the cloud over Qualcomm, in China.

My hypothesis is that the long-term success of Samsung, TSMC and Intel lies with being successful in mainland China. Of these three, I see Intel as being the least favorite, Samsung as the favorite, and TSMC as the dark horse. Changes in the cross-strait relationship could have profound effect on TSMC, but it could go either way. Samsung and Korea have strong trade relations with mainland China. And Intel is simply an outsider in Asia.

 

[hist78]

Hopefully I didn't inadvertently make an offensive statement, that was not my intention. If I did, I apologize.

I have noticed Samsung's investment in mainland China grow in the recent years, while TSMC and Intel's have been dormant, so much so that I forgot about TSMC's fab in Shanghai altogether. I have also noticed the cloud over Qualcomm, in China.

My hypothesis is that the long-term success of Samsung, TSMC and Intel lies with being successful in mainland China. Of these three, I see Intel as being the least favorite, Samsung as the favorite, and TSMC as the dark horse. Changes in the cross-strait relationship could have profound effect on TSMC, but it could go either way. Samsung and Korea have strong trade relations with mainland China. And Intel is simply an outsider in Asia.

No, I don't think there is any unpleasant thing form your post. You have provided an opportunity for me to explore it a little bit more.

Regarding the China market for Intel, Samsung, and TSMC, here is my thoughts:

1. Intel: great potential because they have the technology, patents, and capital that all counties want. The thing will drag them down are: NSA and Intel's own management vision and speed.

2. Samsung: strong technology and fancy product offering in consumer product line, including mobile devices such as phones and tablets. The thing will drag them down are: Samsung's mobile products are directly competing with all their major potential Chinese semiconductor or foundries customers, such as Xiaomi, Huawei, Lenovo/Motorola, Meizu, and Coolpad. They all see Samsung as a serious threat and those unhappy stories about HTC/Samsung or Apple/Samsung only convinced them that Samsung is not someone they can trust.

3. TSMC: So far most major Chinese mobile phone makers either directly (Huawei/HiSilicon) work with TSMC or indirectly (Qualcomm/MediaTek) use TSMC's services. In addition to leading edge manufacturing process, TSMC has something Samsung is lacking in the eyes of Chinese phone makers: the "Trust"!

The huge Chinese national investment effort in semiconductor industry not only bet on the foundries but also heavily invested in the fabless design sector. This meant in the short term TSMC's business can actually benefit from this surge. In the long term, it's hard to say. One interesting aspect is the unification debate across two sides of Taiwan strait. TSMC is such important factor in ROC's (Taiwan) economy that their annual CAPEX is larger than ROC's central government's annual CAPEX. If PROC (mainland China) chooses to use political, legal, and government money to kill TSMC, the negative consequence to PROC will be very severe. I think in many Chinese Communist Party leaders' mind, they don't want to be blamed in the history that they win the Semi but lose the Taiwan.

 

[benb]

Paul just posted a blog on how China has specific goals to develop domestic semiconductor production by 2020. I think it's an intelligent move, given the under-investment in semiconductors in the western world. This creates an opportunity for them. I would interpret it also as real politik--they aren't going to assimilate Taiwan's semiconductor industry anytime soon, so they have to take another approach. It doesn't look like they are targeting TSMC either, judging from the focus on 200mm production rather than 300mm.

 

[user_2013101]

[h=4]INTC correction on schedule [/h]
The previously mentioned short-term trade of INTC seems to remain valid. The correction to a low perhaps below $30 appears to be unfolding.

INTC - SharpCharts Workbench - StockCharts.com

[h=4]Disclaimer [/h]
The posters are not licensed financial advisers. The forum posts are an expression of personal opinions, and does not constitute recommendation of investments whatsoever.

 

[user_2013101]

Intel lowered the 1Q revenue guidance. INTC down 4.73% on Thursday March 12.

Intel Lowers First-Quarter Revenue Outlook - Yahoo Finance

INTC is likely to consolidate around here, or go lower, waiting for the updated whole year outlook in mid April.

Although the PC sale surge from XP termination had faded, Windows 10 is coming in the Fall. Sales of server chips remain solid. SoFIA, or the new name Atom X3, will reduce the mobile losses.

The short-term trade I envisioned is unfolding, looking to average in, on weakness, maybe below $30. No hurry. The summer may be long and unpleasant for INTC.

INTC - SharpCharts Workbench - StockCharts.com

 

[user_2013101]

Before the breakout in June last year, INTC prices had been capped under $27-29 area for 3-4 years. That area now serves as downside support. Besides, there is a gap from last June between $28 and $29. (See chart) That gap is likely to be filled.

$28 should be the reasonable downside target, and can overshoot to $25 if sentiment turns decisively negative; for example, if the perception of “PC is dead” dominates.

Accumulation of INTC should be around $26-28.

INTC - SharpCharts Workbench - StockCharts.com

[h=4]Disclaimer [/h]
The posters are not licensed financial advisers. The forum posts are an expression of personal opinions, and does not constitute recommendation of investments whatsoever.

 

[user_2013101]

[h=4] INTC short-squeeze on last Friday, March 27? [/h]

On Wed March 25, INTC closed below $30, the first time since last June. The next day, it made intraday lower low at $29.31, but closed above at $30.08.

On Friday March 27, INTC struggled at $30 most of the trading day. In the last half hour, the price spiked up from $30 to $33, and closed at $32.

The media contributes the spike to the rumored Intel acquisition of Altera.

INTC prices
INTC Historical Prices | Intel Corporation Stock - Yahoo! Finance

Altera deal merely an excuse

(The deal, if goes through, will be a negative for both companies. I will elaborate on this when I have time.)

Whatever the deal or no deal, it cannot justify INTC’s $3 rise in about 20 minutes (the final 10 minutes giving back $1), right at the final half hour on a Friday.

It carries all the hallmark of a deliberate and powerful short squeeze, as buy-stop’s getting triggered and snowballing upward, timed at the tail end of a Friday afternoon when most traders had left for the weekend.

This is the reverse picture of the March 25 TSM panic sell-off, except that the INTC short-squeeze is much more vicious – it was completed in 20 minutes instead of a whole day and the rise is twice the percentage of TSM’s one-day plunge.

INTC chart
INTC - SharpCharts Workbench - StockCharts.com

Coming back down

When the price spiked up vertically on short-covering rocket fuel, it tends to come back down once the fuel burned out. I think, my original downside target of $28 remains likely.

Coincidence?

Both TSM longs and INTC shorts were severely punished last week.

Readers need to decide for themselves if it is just a coincidence.


[h=4]Disclaimer [/h]
The posters are not licensed financial advisers. The forum posts are an expression of personal opinions, and does not constitute recommendation of investments whatsoever.