Register for free to join our community of investors and share your ideas. You will also get access to streaming quotes, interactive charts, trades, portfolio, live options flow and more tools.
Register for free to join our community of investors and share your ideas. You will also get access to streaming quotes, interactive charts, trades, portfolio, live options flow and more tools.
Yes it did for a while but not now. ORCL was hit hard and so were some other companies.
Strong $ is impacting all companies who have international business. This is nothing new. It has been happening for a while and was the reasons for poor showing in the last quarters. And it will continue if $ stays strong or gets even more stronger.
Who will buy that graphics division? Nvidea and Intel owns the most of the market shares. AMD needs this for gaming console.
But Nenni thinks that TSMC and or Apple will catch up with Intel.
Even if 10nm is delayed at Intel, TSMC will not be able to spend enough money to keep pace with Intel.
-----------------
Even Apple Can’t Save TSMC
By Shuli Ren
Global smartphone growth is decelerating fast.
China, the largest market for smartphones, is already saturated. As a result, global smartphone shipment is expected to grow 16% this year, down from 2011's 68% high, and taper off quickly to 8% in 2016 and 5% in 2018, data provided by Bernstein Research show.
And this is bad news for Taiwan Semiconductor Manufacturing Corp., or TSMC (2330.Taiwan/TSM), said Bernstein’s Mark Li and David Dai.
Using regression analysis, the analysts showed that up till the second quarter last year, TSMC’s quarterly revenue was heavily correlated with global smartphone shipment. R2 in their simple regression is 0.93. In plain English, global shipment data explains 93% of the variation in TSMC’s revenue.
The most unfortunate thing for TSMC is that Wall Street analysts still expect the company to grow at an annualized 10% in the next five years. How is this possible when the shipment growth is expected to taper off to low single-digit? In other words, Wall Street will have to revise down their estimates at some point, which bodes ill for TSMC’s shares.
To be sure, TSMC saw outsized growth since it became the sole supplier to Apple‘s (AAPL) iPhone and iPad processors in the third-quarter last year. But “Apple only presents downside risk”, because:
TSMC will not be the sole supplier to Apple’s next-gen processors due to the competition from Samsung (SSNLF). Beyond the next generation, it is also unlikely that TSMC would regain the sole-supplier role due to the continuous competition from Samsung & Intel (INTC) and AAPL’s strategy to keep multiple suppliers.
But even if TSMC remained the sole supplier to Apple, the indestructible Apple brand can’t save TSMC, because it is not big enough in the global smartphone foundry industry. Last year, Apple shipped 193 million phones, versus the rest of the industry’s 1.3 billion shipments. Apple can only contribute to 18.7% of the total semi foundry business, estimates Bernstein.
Some may argue trading at 11.7 times forward earnings, TSMC has become cheap. Is it? It depends on the metric you are using: TSMC trades at 4.4 times sales, which is by no means cheap. Investors in this stock are betting on stellar execution and operational efficiency.
Yes. I used it fiber modules in networking products 7-10 years back. It is off shoot of Agilent who was spin off from HP. It has been buying a lot of small and medium size company like LSI logic etc.
This merger with Broadcom is huge.
In the mean time shares holders got shafted big time.
Agreed. But Intel already commands very high share of the market. How much more potential Intel has here?
Very impressive. Chart does not list iphone. I wonder how does it fairs against that.
I am not sure who is more desperate here-Intel or Altera. I am sure Altera shareholders want higher stock price one way or another.
You and a few others like you are headed on the same road. Now you will say I don't read your message.
Agreed. He had no appointments to start with except to argue for sake of argueing. He himself stated that.
It was not Dmcq and it was this other guy who always wants to find reason to argue and win at any cost.
Every one will start questioning how Intel will make money on this.
May even call it another contra revenue opportunity.
Market is negative on Intel-it is not new though. see the comments in bold.
-----------------
Intel, TI, Micron Outpace Market Decline on Mixed March Chip Stats
By Tiernan Ray
Chips stocks are outpacing the broader-market declines today as the Street contemplates sobering numbers from the Semiconductor Industry Association, after the organization yesterday reported sales in the month of March rose 6%, to $27.7 billion, which some analysts find disappointing, though not a disaster.
Shares of Intel (INTC) are down 58 cents, or 1.8%, at $32.61; Texas Instruments (TXN) shares are down $1.36, or 2.5%, at $53.94; Microchip Technology (MCHP) is down $1.12, or 2.3%, at $47.72; and shares of NXP Semiconductors (NXPI) are down $2.45, or 2.5%, at $97.68.
The SIA’s CEO, John Neuffer, was upbeat, noting the rise in sales comes “Despite macroeconomic challenges,” and despite and despite 2014 having been a record year for the industry.
The firm noted the uneven regional performance:
Regionally, sales were up compared to last month in Asia Pacific/All Other (3.1 percent), Europe (2.7 percent), and China (1.0 percent), which is broken out as a separate country in the sales data for the first time. Japan (-0.4 percent) and the Americas (-6.9 percent) both saw sales decrease compared to last month. Compared to March 2014, sales increased in the Americas (14.2 percent), China (13.3 percent), and Asia Pacific/All Other (3.8 percent), but decreased in Europe (-4.0 percent) and Japan (-9.6 percent).
Credit Suisse’s John Pitzer notes that on a month-over-month basis, sales were up 14% in March from February’s level, which was ” ~1,000 bps BELOW seasonal,” though he notes that “February revenue was ~40 bps ABOVE seasonal (-4.4% m/m versus seasonal of -4.8% m/m) and Jan was ~40 bps ABOVE seasonal (-8.7% m/m versus seasonal of -9.1% m/m).”
Pitzer focuses on the actual monthly sales, $29.8 billion, whereas the number SIA focusses on is the three-month average.
Pitzer notes prices were the culprit, with volume actually being better than is typical:
The m/m increase in revenue was driven primarily by units – however, the weakness relative to seasonal was driven by a weaker than seasonal decline in ASPs, which offset a stronger than seasonal increase in units. Specifically, units were +27.4% m/m (ABOVE seasonal of +19.8% m/m), and ASPs were -10.3% m/m (BELOW seasonal of +2.0% m/m). On a 3MMA basis, Mar revenue of $27.7bn, -0.1% m/m, was ~160 bps BELOW seasonal of +1.5% m/m driven by a weaker than seasonal decline in ASPs compounded by a weaker than seasonal increase in units – specifically, units were +1.9% m/m versus seasonal +2.5% m/m, and ASPs were -2.0% m/m versus seasonal of -1.2% m/m.
The outlook for the full year, notes Pitzer, is actually better than the Street is expecting:
For 2015, Semi revenue is tracking +5.4% y/y (+5.4% y/y ex- Memory). This compares to current bottoms-up Street estimates of +1.2% y/y (+1.7% y/y ex-Memory). Excluding INTC, Street estimates revenue of +1.8% y/y in 2015.
Tim Arcuri with Cowen & Co. notes that despite the below-seasonal March change from February, the full quarter was okay: “CQ1 semi sales overall fell 2% Q/Q – spot in-line w/5-yr seasonal and also in-line w/the 51 semi companies reporting thus far during earnings.”
Arcuri sees confirmation TI struggled in analog sales, but he sees the company bouncing back later this year:
During CQ1:15 TXN’s Analog segment slightly underperformed industry Analog growth as it fell 4% Q/Q vs. industry growth -3% Q/Q. Due to slightly less growth than the industry, likely from the negative impact that wireless infrastructure had on Analog, TXN lost 20bps of share in CQ1 (to 18.2%). Share has rolled off somewhat over the past 6 months, but TXN share is still up ~40bps Y/Y. We think TXN should resume market share gains beginning in CQ2 (target is 30-40bps gain/year) as its strong position in automotive/industrial end markets should benefit from a seasonal ramp in Q2.
Notes Arcuri, sales of analog chips were actually a stand-out positive, as were sales of NAND flash memory chips:
Analog outperformed all other subgroups both on an absolute basis and vs. seasonality as it rose 27% M/M (5-year seasonal +22%). This was the result of strong unit shipments (+38% M/M vs. seasonal +29%) and modest declines in ASPs. NAND flash performed in-line with seasonal (+19% M/ M) while DRAM, MPU and global IC ex memory all performed below 5-year seasonal.
Despite NAND’s strong showing, I would note shares of Micron Technology (MU) are down 69 cents, or 2.4%, at $27.99.
Chris Danely with Citigroup reviews the earnings season now drawing to a close, and reiterates a Buy rating on Texas Instruments, and on Microchip, though he’s somewhat concerned about a dip in GDP:
Now that 1Q15 earnings season is mostly complete, currency headwinds are negatively impacting the semiconductor sector a little worse than we expected, and we believe semis could experience an additional selloff. However, we remain constructive on the group as we believe the negative EPS revisions are mostly over […] Consensus 2015 EPS estimates for the group have declined 5% driven by currency headwinds, weak PC demand, and lower spending from the communications end market. […] We believe most of the EPS cuts are over, as exchange rates have stabilized and we believe demand should stabilize from the PC and communications end markets over the summer […] We expect demand back to normal in 3Q15. As long as exchange rates remain relatively stable, we believe FX adjustments should be completed in 2Q15 and demand will be back to normal by 3Q15 […] Our biggest fear is another cut to global GDP which we believe could cause a correction in the semiconductor sector similar to 2011 and 2012 when the SOX index declined 15%-20% from the peak.
One chip maker that still has further estimate cuts ahead of it is Intel, he thinks:
Intel requires roughly 17% revenue growth in 2H15 to achieve its 2015 revenue growth target of flat YoY. Since 2010, Intel has grown 2H revenue just 5% on average and has never had more than 8% revenue growth. We believe Intel is too bullish on its 2H outlook and guidance has to come down again.
I saw that and its comparison to A8 of apple was equally outstanding.
TSMC is in business to make money and it does make money with Sofia. In any case Sofia will come home in 6 months.
Back out mobile and every one is dead. The real world is what it is. Server business is there to stay and so is mobile.
Unfair comparison is not the way to look at this.
I agree that TSMC and Samsung has that advantage over Intel. Qualcom and Apple will give you big amount of money upfront to lock in the access to new process. Intel does not for what ever reasons.
The only reason in my mind is that Intel is the main competitors where as TSMC and Samsung are not in that sense.
This is a known fact based on what is written in the process. This process is not available for any one else except Apple. One can always come out with something to please one customer and that also of Apple size. To make money or get good yields is a real tesst of the process.
Yes, it is true that Intel stock price has not gone up with technology lead it has.
Let us separate the two issues.
There may be better investment than Intel insemi business but there is no denying that Intel has a big lead and it will become bigger with time. Details of 10nm(leaked so far) if, are true the others will have hard time duplicating Intel technology success.
Absolutely. TSMC is skipping 20nm and Samsung 14nm is not equivalent to Intel 14nm. Yes Samsung dominates mobile semi business along with mobile product.
But Qualcom results after the market will shake up things in mobile. It is planning to do big cuts in the coming month as it has lost some business. Apple is gaining and Samsung has abandoned Qualcom. That will lead to lower revenue at TSMC.
Things will get very interesting in near future.
No one is ignoring changing facts. Keep in mind 14nm yields have improved over time. With improving yield you will need less capacity of at 22nm and adjust your capacity on 14nm. Yes demand has effect as well. No one denies that.
Putting all these factors, the reduced spending on capex is insignificant.
10nm was planned a long time ago. Yes it probably changed along the way. Intel knew what it will cost to do this-reuse of equipment; new equipment etc. In my opinion reduced capx will not affect this.
What is that got to do with capx spending? 10nm was in development long ago. If this was about TSMC, you will say that is the end of Intel. You will say see what I told you. TSMC is increasing its capx.
Your twisted logic.
TSMC provides services and Intel provides products. Intel owns the server market and PC(though declining) markets and provides tons of products. That is why the difference in R&D and other expenses.
In your book that does not matter.
Parts are parts for some of you.
Don't forget that if Intel is looking for a partner, these companies are also looking to be partnered with. These companies also know that as stand alone it will be tough for them to compete.
That is how I feel. If you have any serious comments only then post.
Qualcomm Break-up? Good Idea, Says Berenberg; Intel Might Buy
By Tiernan Ray
Following speculation a report this morning by The Wall Street Journal that activist shop Jana Partners is urging chip maker Qualcomm (QCOM) to break itself in two, one of the first Street responses comes from Berenberg’s Adnaan Ahmed, who thinks it’s a good idea.
Ahmed, referring to his own open letter to Qualcomm two years ago, argues Qualcomm’s chipset business, QCT, has had volatile margins, and that its royalty business, QTL, has seen its royalty rate decline.
In a break-up, he writes, Intel (INTC) could buy QCT to boost its place in mobile chips, and QTL could buy ARM Holdings (ARMH) to boost its patent position.
Writes Ahmed,
As you are aware, there is news this AM that a large investor is agitating for a spin out of this biz. With the Intel-Altera deal potentially off the table for now, Intel for QCT could be an option (I would buy ARM on the back of that). More importantly, if this spin out does materialise, QTL, needs to diversify its revenue base and once again ARM could be a target here.
I agree with you. He needs to argue for one reason or another. It was so peaceful on this board and far the 2-3 days it is argument over cameras. There are different needs and that is why there are different products.
There are folks who don't have smart phones and have very limited need for camera.
If you read the details, Altera was asking fora ridiculous price. It rejected Intel offer of more than $50.00
It appears that Intel and Google are forcing hands of various OEMs.
Big guys like Samsung will not allow this happen unless it changes designs towards Intel. That will not happen as it(Samsung) wants to design products based on its own chips.
All small OEMs all over the world would welcome this change.
Another take from Wired.
--------------------
To Stay Relevant, Intel Explores Its Biggest Takeover Ever
INTEL, THE WORLD’S largest chip maker, is in talks to acquire Silicon Valley chip builder Altera, according to a report from the The Wall Street Journal. And if such a deal goes through, it would be Intel’s largest takeover ever.
The Journal‘s story is thin on specifics. It doesn’t cite a specific source, and it says the talks are nowhere near complete. “Terms of the potential deal and its timing couldn’t be learned, and it is possible there ultimately won’t be one,” the paper reports. But the acquisition makes a lot of sense for Intel. Altera isn’t a big name, but it builds chips that represent the future of the massive data centers that power the internet.
“I am absolutely not surprised,” Jason Mars, a professor of computer science at the University of Michigan whose research focuses on hardware used in the modern data center, says of the rumored deal.
Altera builds what are called FPGAs, or field programmable gate arrays. These are chips you can program to perform very specific tasks, and they don’t require as much power or as much space as the chips that traditionally have powered our internet services. They’s why Microsoft used them to help power Bing. People like Mars believe these chips could help power a wide range of services, including the new breed of artificially intelligent services, such as Apple Siri and Microsoft Cortana.
“This is where the things are moving. And Intel is responding,” says Mars, who is part of a research team that recently published a paper showing FGPAs are far more efficient than traditional chips when running things like Siri and Cortana. Intel makes most of the chips that power our internet services today. But it doesn’t make FPGAs (though it has said it is exploring this type of silicon).
Intel declined to say anything about the The Journal story, and Altera did not respond to a request for comment. But the two companies are longtime partners. About two years ago, they signed a deal for Altera to build future chips in Intel foundries, the enormously expensive plants where Intel manufactures chips. And, most notably, the two are collaborating on an experimental computer motherboard that includes traditional Intel server chips and Altera FPGAs.
The project is called HARP, short for Heterogeneous Architecture Research Platform. Mars and his team are applying to receive and test one of these boards. Although the world is moving toward thing like FPGAs, traditional server chips will continue to be used as well, and these boards explore ways of marrying the two. “It’s a very good move for Intel,” Mars says.
Many companies are already using smaller, simpler, more efficient chips to power certain services. As Microsoft uses FPGAs to help power Bing, Google and Facebook are using GPUs, or graphics processing units, to power various “machine learning” services, services that can, say, identify faces in photos or identify spoken words.
For Mars and others, these chips will play an even larger role in the data center as time goes on, and we may also see many data center machines powered by low-power chips based on those that run your cell phone today. All this creates a bit of a quandary for Intel. But the company is exploring all sorts of alternative chips. It may even buy a company that makes them.
I am in agreement. Intel stated this will be disruptive technology.
I was responding to Michael why Samsung won Apple business.
altera is up 22% or more.
Yes Intel does but I don't think these products contain 3D flash.
Intel is up big time on rumors that it will buy Altera
Thanks Chipguy. I was hoping that you will respond.
Because Samsung has products today. These Intel-MU parts are samplings now.
Here is something that could potentially disrupt Intel lead in servers.
--------------
http://www.wired.com/2015/03/10-ways-worlds-first-open-server-architecture-disruptive/#!/
What are you doing here? Call MS and become an advisor and tell them that is what it should do going forward. Reinvent RT with your price and features in mind.
RT is dead regardless of what happened.
If you noticed, Nvidia wants to get royality (free) money from Intel and this may be the only way to get it.
But TSMC is having problem with yield with high 16nm 3D process, so who knows.