SAN FRANCISCO — Analysts at JP Morgan are calling for Intel to shut down its mobile and communications group to improve corporate profitability in the wake of poor earnings per share in the first quarter of 2014. But the x86 giant and at least one other analyst said the company should stay the course in the hotly competitive smartphone and tablet market.
“We continue to believe mobile is unprofitable for Intel," a JP Morgan statement read:
We continue to believe Intel will lose money and not gain material EPS from tablets or smartphones due to the disadvantages of x86 versus ARM and overall low profitability of the tablet and handset processor market. If Intel were to shut down its mobile business, we estimate it could unlock roughly $0.50 in 2015 EPS.
The mobile and communications group saw a $3.1 billion operating loss in 2013, with 1Q 2014 losses hitting $929 million and revenues at $156 million. While Intel officials acknowledged the loss, several were quick to call recent financial numbers an “investment” in the mobile ecosystem.
“This is definitely more of an investment -- a well understood strategy for Intel’s computing leadership perspective,” Julie Coppernoll, vice president of Intel’s mobile and communications group, told EE Times. “We feel that we have a plan, we understand how we’re going to deliver, and are making investments that improve our particular competitiveness from a platform cost perspective.”
Rather than take JP Morgan’s suggestion to shut down mobile, as Texas Instruments did, and refocus efforts on higher-margin PC and foundry businesses, Intel reaffirmed its commitment to tablets and LTE. Growth in Intel architecture will be one of the biggest in mobile, Coppernoll said, and the developing tablet market will quadruple the company’s volume.
“Our expectations [for the tablet market] haven't changed. There may be [variations] -- the second decimal may have changed, but directionally it's the same,” Intel CFO Stacy Smith said in a quarterly results call. “We’re actually feeling pretty good about the line of sight we have across the customers, the longer-term agreements we have in place, and design wins that we know are coming to market, as well as the Q1 volume.”
Responding to comments about the efficiency of Intel’s x86-based processors, Coppernoll said the company’s chips are competitive at a power level and pointed to the new generation of its Atom chips, code named Broxton, as “hero level” devices. The 64-bit mobile chip for smartphones and tablets is based on new CPU and graphics cores.
“We don’t take any competition lightly, but we're confident that our Intel Atom (Bay Trail) processors will continue to be the performance and performance per watt leader,” said an Intel spokesman. “Keep in mind we are also manufacturing these chips now at 22 nm, and we are in the process of starting up our 14 nm process.”
While JP Morgan’s suggestions might increase profit, Intel has no choice but to stay the course in the mobile market, said Will Strauss, president of market research firm Forward Concepts. Intel should adopt a strategy similar to Nokia's, emphasizing third-world markets rather than the US.
“To compete with Qualcomm, they’re going to have to pump in hundreds of millions to catch up. There’s nobody else out there that has the finances that Intel has to compete... Qualcomm is about two years ahead of anything Intel’s got,” Strauss said.
As for cost, Intel aims to decrease its bill of materials across various products. Smith highlighted the low-tier SoFIA, Intel's first integrated mobile SoC platform, and mid-range to high-range Broxton.
“I think on a like-dollars per unit, it comes down pretty dramatically over the course of 2014. And it should be relatively small... as we get into 2015,” said Smith. He predicted "a better cost structure with our own products and a better cost structure overall with the bill of materials as we enter 2015 and then work through 2015.”
Both Intel and Strauss emphasized the importance of partnerships in Intel’s mobile market success. Intel has signed long-term agreements with Lenovo, Foxconn, and others “to expand tablets and smartphones for those who will utilize Atom on the processor side as well as communications products,” an Intel spokesman said.
Excellent and decent post. I found this much informative, as to what I was exactly searching for. Thanks for such post and please keep it up. friv 2 | friv 10
The 64-bit transition is actually a minor one because 64-bit CPUs can still run all existing ARM code, so there is no reason to switch to a different ISA. Companies are switching on masse to 64-bit ARM. So the market has already chosen.
The fact is, JP Morgan has a point - Intel has been losing many billions per year over the last few years trying to get into mobile without any success at all. So it is not just about this quarter but about the last 24 quarters. There comes a time where they have to admit it was crazy to push x86 into a mobile. Their latest Silvermont has already been beaten on performance and power before it even made it into a phone...
Intel might make some profit in tablet market as they are a brand seen closeness to PCs but mobile its so very difficult. It will be another player in the smartphone market. There is enormous amount of competition. ALmost every company that has presence in semiconductor wants to try out smartphone.
So here are some production facts, Intel economists and industrial scientists don't talk about, or don't want to talk about, today.
On 2013 financial Intel's fully burdened total cost of platform sale; microprocessor and its support components sold in the bundle deal, is $256 average total cost, $147 average fixed cost. Average total cost is determined PE&C < Dep + R&D, + Sales + Marketing + Intel Inside price fix cost / total processors produced. Average fixed cost is determined PE&C < Dep + R&D / total processors produced. Every product priced below these cost rungs, where AMD is forced too compete under conditions or production monopoly and distribution cartel, is subsidized by other Intel products priced in excess of average total and average fixed cost. Every time Intel sells product priced below these cost rungs, the result is a marginal revenue loss. This is a simplified way to look at cost within this general financial framework.
Pursuant to mobile economic assessment of cost price delivers a more precise result:
Haswell Quad Mobile average weighed price bin split exact determined on 1K price is:
Average Weighed Price $504.65
Average Marginal Revenue $353.83
Average Marginal Cost $150.82
Haswell Mobile Desktop and Quad Runs Combined on 1K (they are all the same parts):
Average Weighed Price $398.45
Average Marginal Revenue $346.31
Average Marginal Cost $52.13 (this is the more realistic of the two quad cost answers)
Haswell Dual Mobile SIP i7, i5, i3 average price bin split exact determined on 1K price:
Average Weighed Price $352.76
Average Marginal Revenue $308.93
Average Marginal Cost $43.83
To get a more precise Intel cost : price ratio removing the dealer's cut of Intel margin value:
The general rule to determine Intel first tier dealer price is AWP on 1K / 2.
The general rule to determine Intel marginal revenue is (AWP on 1K / 2) / 2
The general rule to determine Intel marginal cost is [(AWP on 1K / 2) / 2]
The general rule to determine Intel hard dice fabrication cost is [(AWP on 1K / 2 ) / 2] / 2
In these technologic times Intel relies on the production metrics of a brick fabrication plant, or, in the ancient Scot or British vernacular of semiconductor production; a kiln operation.
Haswell Mobile Desktop and Quad Runs Combined (they are the same parts):
First Tier Dealer Price $398.45 / 2 = 199.25
Intel Marginal Revenue = $99.62
Intel Marginal Cost = $49.81
Intel Hard Dice Fabrication Cost = $24.90
Haswell Dual Mobile SIP i7, i5, i3 average price bin split exact determined on 1K price:
Average Weighed Price $352.76 / 2 = $176.35
Intel Marginal Revenue = $88.11
Intel Marginal Cost = $44.08
Intel Hard Dice Fabrication Cost = $22.04
Revenue can also be considered applying gross margin then / 2's
Haswell Mobile Desktop and Quad Runs Combined (they are al the same parts):
Average Weighed Price $398 / 2.* .63 = $125.54
Intel Marginal Revenue = $62.77
Intel Marginal Cost = $31.85
Intel Hard dice Fabrication Cost = $15.69
Haswell Dual Mobile SIP i7, i5, i3 average price bin split exact determined on 1K price:
Average Weighed Price $352.76 / 2.* .63 = $111.94
Intel Marginal Revenue = $55.55
Intel Marginal Cost = $27.14
Intel Hard Dice Fabrication Cost = $13.88
Now let's look at Bay trail SOC
Bay trail full run taking into account 't' for android tablet, 'm' for mobile Chrome and MS 8.x, and 'd' for Celeron Value line that is also 'i' for embedded upping package for better thermals. Noteworthy,there is nothing value about the Pentium branded Bay trail 'm' incorporating $18,313,509 in consumer monopoly overcharge values, in this case a tying dealer value, associated with 2,850,467 of the Pentium branded Bay trail 'm' when sold at one unit price greater than $160.
Bay trail short run 'm', 't', 'd' Average Weighed Price:
Average Weighed Price bin split exact 'm' = $151.54
Average Weighed Price bin split exact 't' = $32.46
Average Weighed Price bin split exact 'd' < 'i' whose
percent break out is currently unavailable = $81.59
When 'm', 't' and 'd' are combined for average across the full run:
On Bay trail $5.50 China white box price Intel looses $9.35 per processor.
$14.85 - $5.50 = ($9.35)
Now consider the 't' for tablet short run alone:
Average Weighed Price = $32.46 (which happens to be the average of 1K price)
Average Marginal Revenue = $22.76
Average Marginal Cost = $ 9.70
Intel Hard Dice Fabrication Cost = $4.85
And consider the 'm' and 'd' short runs combined
Average Weighed Price = $134.58
Average Marginal Revenue = $94.36
Average Marginal Cost = $ 40.22
Intel Hard Dice Fabrication Cost = $20.11
So is Mr. Krzanich's stated Bay trail 't' subsidy the actual dice fabrication cost of the other two Bay trail grades 'm' and 'd'; $20.11?
Also notice the marginal revenue loss between the combined 'm' + 'd' runs verse tablet alone; $94.36 - $22.76 = $71.60. That is Intel's actual revenue loss over the full run for every unit of bay trail allocated to tablet. Opportunity cost of selling 10,000,000 bay trails into Android or MS 8.x tablet over Chrome book and Value mobile delivers a $716,000,000 marginal revenue loss. Allocating into embedded 'i' adds to that marginal revenue loss considering 'i' 1K price range $31 to $52.
If you haven't noticed lately a lot of those Bay trail 't' grades have just been supplied into the industrial embedded market as Bay trail 'i' grades that is a Tunnel Creek replacement.
Bay trail 'i' is aimed to knock ARM derivatives out of the embedded compute modules product categories, where Freescale I.MX6, is just hanging on given price for the applications feature set. I.MX7 and 8 have been annoucned.
There is another way to look at this, please see Mr. Groff's analysis in Seeking Alpha, Intel Tablet Contra Revenue Will Cost About 3.5 Cents Per Share in Q2, April 17, 2014.
See http://seekingalpha.com/article/2146613-intel-tablet-contra-revenue-will-cost-about-3_5-cents-per-share-in-q2
Mr. Groff calculates Bay trail contra revenue (per platform device production subsidy), on Mr. Kranich' 5,000,000 q1 2014 volume statement from the financial conference call at $178,000,000 in q2 2014. Mr. Groff's assessment in Seeking Alpha explains why.
Now this example calculates for Bay trail unit volume against Mr. Groff's $178,000,000 quarterly subsidy estimate.
$178,000,000 / Bay Trail Full Run Marginal Revenue $68.66 = 2,592,428, is roughly one half of Mr. Kranich's q1 2014 volume statement of 5 million units supplied. Taking into account Mr. Kranich's guidance of 5 million 't' units supplied into tablets delivers an interesting result; $178,000,000 / 5,000,000 = $35, the median of Bay trail tablet average price. Meaning Intel is giving Bay trail 't' for tablet away in the platform subsidy deal at 1K list price.
On the simple AWP / 2 method to find marginal cost = $17.5, or using gross revenue method $35 * .63 = $22.05; that is marginal revenue on economics, delivers these results:
$178,000,000 / $17.5 = 10,222,222 units that include a potential Intel cartel signal.
Or, on the latter, $178,000,000 / $22.5 = 7,950,617 units.
Now what if Intel provides Bay trail 't' approximately at marginal cost; $8.75 and $11.37 per processor respectively.
$178,000,000 / $8.75 = 20,444,444 units. Again a signal?
Or, on the latter, $178,000,000 / $11.37 = 15,733,411 units
Now what if Intel provides Bay trail 't' approximately at hard dice fabrication cost,
$4.38 and $5.68 per processor respectively.
$178,000,000 / $4.38 = 40,842,211
Or on the latter, $178,000,000 / $5.68 = 31,494,522 units.
In both examples hard cost is less than marginal cost. Price $4.38 to $5.68 < Marginal Cost on Economics $9.70 = Areeda Turner Price < Cost Trade Commission Review.
So what is Intel's contra revenue subsidy strategy?
Intel strategy is to price Bay trail 't' progressively down tablet short run marginal cost curve to run end hard cost, to reach their 40,000,000 unit supply objective. And appears too be doing so already, addressing the very low end of China, Android Tablet market.
But there's more. And that is called the 22 nm to 14 nm transition. Lacking a shrink, or bumping into production obstacles, q3 2014, aiming for 60,000,000 units, Intel marginal cost can go through the roof. And that will really add to the Bay trail tablet platform subsidy.
@Rick, I agree. This is why we can't make good things. The bean counters want immediate returns. I remember when the US abandonned LCD manufacturing to the Japanese because they didn't project positive earnings for 10 years.
It seemed the EPS gains JPM suggested by leaving the mobile market could be miniscule compared to the possibility of revenue from a major design win in the mobile sphere.
To save this item to your list of favorite EE Times content so you can find it later in your Profile page, click the "Save It" button next to the item.
If you found this interesting or useful, please use the links to the services below to share it with other readers. You will need a free account with each service to share an item via that service.