by 
{complink 2657|Intel Corp.} is changing alright, but in such incremental and deliberate ways, the process may be near conclusion before its rivals, investors, and many industry analysts realize the extent of the dramatic turnaround the current management is trying to execute.
Over the course of the next five to 10 years, and certainly before the end of this decade, I expect dramatic changes in the operating model of the company, its end-markets, as well as competitors and customers. The changes will determine whether it stays atop the global semiconductor market as the biggest vendor by revenue.
For investors, the emerging Intel is already a puzzle, as evidenced in the weak valuation that followed the announcement of the company's fiscal fourth quarter results last week.
The Intel that is being refashioned out of the enterprise that has dominated the microprocessor market will be a manufacturing giant with unparalleled manufacturing resources and IP portfolio. It will become a leading semiconductor foundry and partner to many of today's OEMs — and future OEMs — in key telecom and data networking equipment markets. That Intel won't walk away from its traditional PC microprocessor business — unlike when it exited the memory semiconductor market. Rather, it will continue as the dominant supplier to whatever remains of its traditional PC microprocessor OEM customer base, but without the current overwhelming dependence on them for a majority of its sales.
That's not all. Within the next five years, Intel will morph into a powerhouse supplier of chips to data-equipment companies, in addition to becoming one of only a few IDMs capable of offering extensive foundry-style services (semiconductor wafers) to leading OEMs in high-margin and highly specialized IT equipment markets.
In other words, in a very short time, Intel has the potential to become more recognized and rewarded for its services as a customized wafer foundry company than simply a vendor of chips to either the traditional PC or mobile device (tablets and smartphones) markets.
The foundation for that future is being laid right now, and (perhaps rightly so) Intel is being punished by investors for the decisions it has taken in pursuit of those goals. Rather than reduce capital expenditure in response to slowing sales, for instance, Intel is pouring more funds into infrastructure and next-generation manufacturing processes.
Though it forecasts “low single-digit percentage increase” in revenue due to challenges in the PC segment, in 2013, Intel plans to spend about $13 billion on capital equipment — in furtherance of 14 nanometer and 10 nanometer plans. Capex for 2012 was about the same level as is now being projected for 2013.
Other numbers support the theory that Intel will be spending more in the future to further some of the goals identified above, rather than scale back its investments. Total R&D and marketing, general, and administrative expenses for 2013 are projected at approximately $18.9 billion, even higher than the $18.2 billion it spent in 2012 and the $16 billion from 2011.
Paul Otellini, president and CEO, defended Intel's spending decisions for 2013, and noted that the capex would help the company maintain its manufacturing prowess, giving it the edge it needs to stay ahead of the competition in the PC segment, as well as in tablets and smartphones. The high capex will probably be maintained even in 2014 and 2015, according to Intel CFO Stacy Smith.
“The world's leading edge fabs are the single greatest asset that we have,” Otellini added during the conference call.
I agree, but I've long been a fan of Intel and its “only the paranoid survive” investment and managerial mentality. However, for many others in Intel's public domain, maintaining a strong capex and operating expenses budget in the face of weakening PC sales may not make sense, especially since the company is getting clobbered by ARM in the mobile communications equipment market.
Intel certainly has major challenges, but those who focus on its failure to make inroads into the ARM-dominated mobile segment need to take a closer look. If Intel successfully makes the transition I've outlined above, neither ARM's current success, nor the weakness (today or in the future) of the PC market will matter a great deal. The majority of Intel's sales would be independent of each segment, and it would have fashioned a new business model for what some of today's embattled IDMs could become — had they Intel's deep pockets.
First, though, it must successfully make that transition. Failure would turn the “greatest asset” that Otellini talked about last week into the greatest millstone around its neck.
I'm also confident that INTEL has what it takes to make it in this market (as it has over the years). My question is what happened? Why did it stay behind in the smartphone and tablet chip market – where ARM is definitely king.
Mr. Roques, Intel has tried over the last decade to break into the communications semiconductor market and it wasn't very successful. The smartphone market came out of nowhere really and tablet PCs followed. Intel is playing catch up here too.
One simpler way for Intel to make transition is make lot many acquisitions. Some them may not be good, but it will give them big break. Is Intel interested in RIM hardware business?
In my opinion , BIG DATA is where Intel can focus its attention.
BIG DATA requires a radically different approach in processors, servers, software and related analytics.
Intel is the right company to junp into this arena ( I don't know if they are already in) and become the leader in offering state of art hardware and software solutions with a well planned product road map
Intel is already in the big data business. Its only business unit that increased revenue in the fourth quarter was the data center group, which reported sales rose 7 percent sequentially and 6 percent year-over-year. It's an area its focusing on.
@Mr. Roques,
We can ask the same question about Nokkia, RIM, Microsoft … w.r.t to them lagging behind in the handheld devices market. Why isn't INTEL winning the smartphone chip war? The anwser is simple – They didn't see the war coming – that fast.
Intel co founder and ex CEO Andy Grove claims he escaped Soviet invasion of his homeland Hungary. So its ironic that after getting to the USA he created a totally STALINIST organization INSIDE INTEL !! Apple's 1984 Superbowl ad lampooning the lemmings of Big Blue ( IBM ) applies even more to Intel.
At present Intel “inside” ( Mgmt., Sr. engineers ) is totally geared towards dominaton of a dying monopoly ( WinTel processors ). Within Intel R&D has been geared towards creating incremental improvements ( < 20 % ) every couple years and then beat the drums of Moore's Law loudly.
Without going into the technical details, let me just say that all the hoopla over FinFET ( 3-d transistors ) over the last 2 years have exposed the mediocre performance from much vaunted Intel process / mfg. R&D. Yet they have kept on marching in lock step to bigger and more expensive Fabs when its not clear that there is a growing market for PC processors. They have not delivered on SoCs because the design expertise is still wasteful X86 CISC oriented.
At this time Intel needs to get off its traditional emphasis on ever smaller transistors ( sub 20 nm ) or larger wafers ( 450 mm ) or $ 10 billion new Fabs to build them and instead put more emphasis on various chip architectures w/o undue bias rowards X86 CISC.
@HH, the difference here is Intel realised this and the company is turning itself around. Bolaji noted in the article that Intel “will become a leading semiconductor foundry and partner to many of today's OEMs …and future OEMs” in telecom and data networking equipment markets. I truly see them doing well. Good on them!
Intel surely has some of the best semicon technology these days but selling foundy services might not have the same gross margin as owning the whole shebang so Intel will need to get its strategy bang on to succeed.
@HH, have you read/heard of tough xTablet Windows Rugged, Tablet PC running by Intel processor?
Flyingscot, Correct but Intel may still go ahead to offer foundry services to secure a stronger presence in some market segments and also because fabs are so expensive that the handful of companies that can still finance their own fabs my have difficulties feeling them.
The assumption that an IDM with proprietary design and a near monopoly to sell it at a high margin seems to be no longer true. In 2012 Intel managed a margin of 20 % but TSMC roared ahead with 36 % even though their 28 nm was just a small part of the total $ 16 billion.
Intel needs to open its doors to more collaboration. As far as big data is concerned, it think the biggest customers are HP and Dell. Given Intel is basically hardware provider, i do not think that they can have a good profit margin unless they own the products as they have done in traditional PC microprocessor business. The path forward for Intel is to try to capture as much hardward market in cloud computing and big data as possible.
Bolaji,
I agree with you that the complete transparency across all the supply chain partners is unattainable because a lot of the internal information is too sensitive for the businesses to make it visible to the outside world.
But yes, the pertinent information or indirect ways of giving information to partners can be explored. For example a distributor may not reveal the exact inventory in his warehouse but he can indicate lead-time pointers based upon his inventory level -which may be based upon the actual inventory plus the expected material in-flow. Similarly the manufacturer may not disclose his daily production volumes but indicate lead time for each manufactured item based upon some formula which takes into account the existing inventory, the expected production and the expected orders.
So by giving the pertinent information without disclosing the sensitive internal data , the supply chain partners can bring about the transparency in their operations
@p_d: Very interesting point p_d, although what happened until now in reality, has depicted how strong is the right trade-off for providing info but preserving sensitive (and in a such case privacy) data or industry secrets. And the matter is increasing its complexity while the market is trasforming in a global market, by leaving local geography.
Intel dominates the data center and will for some time. On the advantages of leading process monopoly, secured and sustained through structured market rigging, the enterprise is almost unstoppable in any of its component category pursuits. While vertical integration to capture downstream producer values is a noteworthy strategy at technology transitions where costs increase and so must margins, this analyst does not believe Intel will compete with its traditional channels; because their organization, product and structural change through a computing paradigm transition are just as burdensome in defining the new competitive realities.
Analyst suspects Intel's next big push is not in mobile, where Intel can dump at and below cost all day causing all sorts of competitive havoc that have nothing to do with a superior product offering, but in processor network switch integration.
This is the next big market Intel is capable of consuming within boundary of a 10% wafer cost increase for the exponential margin of NIC in processor and perhaps some hardwired security functions that compliment. This move surely presents a political issue among some of Intel's component compliments, and customer's with the kind of mass even Intel has a tough time pushing around. Subsequently reveals Intel's entry path into foundry on the type of high margin specific volumes Intel has already suggested and is engaged in on an experimental basis.
What is astonishing about this sort of move, and this is the wild thought here, communications switch and route design manufacturers once procured Intel microprocessors that were good enough for their product applications procured from the unwanted overage of Intel PC IDM sales bundles. In this way Intel could sell into communications gear at price levels acceptable for the procurement requirement. In other words at prices hidden below Intel measures of cost delivering some margin upside to the PC IDM broker of these unwanted product grades for commercial IT & consumer computing products. Bottom of the bin that in all cases provided more than good enough performance for control plane processing.
So now will the reverse occur? Will Intel communication foundry customers, possibly CISCO, benefit from NIC in processor supply from Intel; specific to their captive requirement, while becoming a broker of their own overage to others?And, subsequently, would create a new form of Intel sales channel.
In the secret sauce IP world of communications switch gear not likely off the bat but please consider the potential of this procurement channel reversal.
Mike Bruzzone, Camp Marketing
Interesting post, Bolaji. While visibility in the supply chain has it's own benefits, do we really need visibility in all cases? Can there be a situation where too much visibility be harmful and it's better if there's limited visibility?
@prabhakar: Good point. I agree that in certain cases giving away too much information from your supply chain can be harmful for the business. Companies should have control over the system so they can restrict the sharing of information at any time if they feel the need.
Matteo, It's difficult getting the right balance between the need to protect sensitive information and the need to disclose some information to business partners. The company first must decide which partners it could trust and then the level of trust and what type of information to entrust to the partner.
Often the information you described as “sensitive” data is really what key partners need to have in order to provide the best support to an enterprise. Yet, that information may be the most critical to a company's success.
Total visibility, even if achieved, can have its own problems. You can end up with the too-much-information syndrome where a company has more data than it can really manage, analyze and use. Plus, it's easy to get hung up on “visibility.” The data must be used. Right? It's not simply visibility, it's what you do with the info.
I agree, definitely; after all, there is another point to consider; if we focus only for a moment on privacy – for instance – we find that a general rule worldwide valid, is not completely in place. Then, what it works within a region, doesn't work for a different region. As consequence, partners need to agree through contracts on top of that, but this process requires analysis and deep studies from legals; it takes time and it could be a cause for speeding-down business relationships.
Yes there are so many software solutions available in market that it becomes so very difficult to choose from. But then at the the organization level each company in supply chain business would have invested in few different kinds of software packets.
Partners as long as things are verbal and can be managed at informal relational level they would be happy to oblige but if you ask them to sign papers no one would.
SP, I trust you, professionally speaking, I've met with that situation several times.
@Bolaji,
“Total visibility, even if achieved, can have its own problems.”
That's correct. What do you think is the best way to determine what kind of visibility is good for your company? I agree that is it not a good thing to just follow the trend.
@SP,
“there are so many software solutions available in market that it becomes so very difficult to choose from”
There are many software packages you can choose from, but still you need to do an accessment in order to deternime which one is good for the company. When this is done, it becomes easier to make the right choice.
SP,
I agree entirely.
There is so much sunk into maintaining the Existing Software Infrastructure already that you try very hard to maintain the Status Quo.
But eventually an Enterprise(any smart one for that matter) has no choice but to move towards a more Integrative/Comprehensive Solution for the Entire Enterprise;Else the inefficencies that creep in are massive.
“The data must be used. Right? It's not simply visibility, it's what you do with the info.”
@Bolaji: I agree. This is what leads us to the importance of having a good analytics system to represent the information as well besides storing tons of it.