Pierr Johnson

Pierr Johnson
Contributor since: 2009
Company: Neoga Capital - Technology Investment Analytics
Great contribution! Your expertise and history with Intel allow you to marshal quite a compelling argument. I would like to think those with counterarguments could see the refutation you provide. However, there is a mania involved that may make this difficult. I have followed Intel for about twenty years and have understood these arguments for a very long time. We are finally at the inflection point in the road map where the laws of physics will give Intel the Mobile market share it has sought for so long. But you have stated it better than I can. Again thanks. I look forward to you future posts.
I would like to invert your argument. Apple, Qualcomm, and Samsung have been major Tech players for one to two decades and each attained its leadership with its own technology(ies). Likewise Google, with Android. The beauty of ARM's IP allowed it to participate, via IDMs and foundries, but hardly made it the singular enabler. Think of the economic value of ARM's contribution in terms of the very low royalty rate it earns, especially relative to what Qualcomm captures with its royalties, which are still a small percentage. I've been a huge fan of ARM and have followed it for over a decade. But it is what it is.
It is Intel's view that its processor designs will achieve parity with ARM in terms of power efficiency in the next iteration or two. In most other respects, ARM is far behind (e.g., 64-bit). Of course, Intel's process advantages are likely permanent. It will always get there a year or two ahead of TSMC and Samsung. The superior position in both design and process, once attained, will be very difficult to reverse. If achieved, this will be a permanent, fundamental change in technology and market position.
Regarding new market penetration by Intel or the ARM ecosystem: years of observation indicate to me that the power of incumbency is very high and Intel is notably persistent. ARM provides processor cores and technologies and can only be as persistent or successful as its customers and their manufacturers and customers. It will be very interesting to see this to unfold
Thanks for the thoughtful article, and for promoting such a vigorous exchange in our community. You provide a good summary of the “ARM Ascendant” thesis. I do think you have neglected to reference or refute key counter-arguments, however. Intel’s technology lead remains and will deliver superior power efficiency with existing and new designs as we move down the roadmap. Samsung dedicates its leading process technologies to NAND and DRAM, and both Samsung and TSM, their achievements notwithstanding, generally lag Intel by a node at any given time. Intel has indicated this will not allow this to change. Moreover, process and design issues at the leading edge have been a recurring problem for TSM, specifically with ARM-based devices. Of course, we know about them because of the resulting finger pointing and can be confident Intel and Samsung experience them as well, with little public notice.
Another thing that occurs to me, as huge a fan of Gulliver’s Travels as I am, is that the challenge to Intel in the App Processor market is from not the Lilliputian of Arm’s ecosystem, but other Tech giants, specifically Apple/Samsung and Qualcomm/TSMC. And as Ashraf notes, both Apple and Qualcomm use custom implementations of ARM technology. This market has become so unattractive, that TI has pulled its OMAP offering. We might wonder why Intel seeks a spot at all. But it is important to note that the design and process approaches Intel now uses with its SOCs differ from those dedicated to volume PCs.
So many here are comparing ARM and Intel to Intel and IBM in computing, but the analogy is just faulty. When IBM chose Intel to displace Motorola in its earliest PCs, it elevated Intel to achieve the dominating position it now has. And IBM was the leading PC manufacturer at the time, intent as it was in fostering the transition from business typewriters (where it led) to Word Processing. Compaq, AST and latterly HP joined the fray, but it was the entrance of Dell that first challenged IBM in this market, as its initial growth phase lost some momentum. It was only with the sale of this business to Lenovo that IBM ceded its leading position in this market. Note, too, that though IBM was part of the PowerPC consortium, it remained committed to the Intel/Microsoft platform for its PC offering.
All of this aside, I really appreciate your article. Two great companies dominating markets that now collide; at the margin is no longer marginal. And I will definitely follow up on some of your references
Thank you for your comments. I am eager to tear into the 8-K and may have some additional financial details soon.
Breakup fees are fairly common and effectively increase the mount of capital a competing bidder needs to finance to win the deal. To the degree the buyout group has borne costs in negotiating the deal, some fees are justified. But the main intent is to increase the threshold for competing bids. (Not a pretty business!)
Yes, to the founder go the perks! It is a good thing you bought when you did. Besides Michael Dell, only those who caught the bottom will do well with this transaction. Congrats!
Thank you for the insightful response! It strikes me that, with the departure of Dave Johnson (head of their Strategy effort), the move to go private is an end in itself and not part of a large, fully considered strategy to address the company’s deficiencies. And, assuming they can even assemble the capital to close the deal, hugely excessive debt will indeed hinder the company’s ability to develop and execute a strategy. Even the loosest covenants will lay claim to cash flows for a significant interval. And, yes, Private Equity investors often require up front returns of one sort or another. So I appreciate your points a great deal
I believe Microsoft’s involvement is defensive in many ways, just as in many respects the Surface offering is. The simple fact of the matter is that both HP and Dell wavered initially in confronting the challenges Tablets impose on PCs, especially consumer PCs. HP is back on board, to judge from its Analyst Day presentation. Dell may well be running from its problems when focus is most required. Yes, they are doing something with Tablets. But is not that the proper focus, rather than some massive buyout? Again, thanks for your comments
Thank you for your comments. Yes, with Lenovo and other aggressive players in the market, it is tough indeed. I agree that Microsoft would likely want to avoid the PC hardware business model. And it would certainly muck up their channel with conflict. Still, I do see the potential this year for some interesting changes in PCs with the various hybrids and tablets that are coming in the market. It wasn't clear to me that the present PC makers were up to the task. I am pretty sure this is what prompted MSFT to develop the Surface RT and Pro. Still, there seem to be some fairly interesting products in the pipeline that may or may not find success.
Thank you for bringing this up. This is a thread I have been meaning to trace for a while and the link is a great place to start. While PCs no longer seem to possess differentiation, with the emergence of Tablets, Ultrabooks and hybrids, there is actually room (for a change) for some interesting innovation this year. Of course, this USB-based device is especially interesting! Again thanks
Thank you for your comment, to which others agree. I agree that MSFT feels the need to direct development of the PC, especially in this dynamic market. Still, I do not believe they would want to own the manufacturing assets, given the weak margins associated with the model. It would complicate things with other customers as well. Also, it is so easy to outsource manufacturing. Still, you have a point. Actually, there is one well known commentator who feels that if MSFT unified the Platform with really elegant devices, it could truly compete against Apple.
Thank you for your comment. Consumer PCs have been a issue for some years, in my view. Their Corporate offering seems to be a core asset. And like HP they derive buying power with Intel and others across the Server and PC offering. So it may be hard to disaggregate a business to sell, in many ways. (HP's dilemma as well.)
I agree that the Surface is a new thrust for MSFT. Still, I am not sure they would want the business model that comes with ownership of these assets. It is so easy to outsource this to Asian entities that find the paltry returns (and massive scale) attractive. Still, as you suggest, all cards have to be on the table, don't they?
Thanks again
Thanks for the thought provoking commentary, Dana. Apologies, fellow readers, for not slugging my way through all of the comments. A couple points, as others may well have noticed: I do not believe Microsoft intends Windows 8 to the full release Windows 7 is. In fact, I would argue that Enterprise users are still actively migrating to Windows 7; any Windows 8 adoption will likely be around Windows Pro Tablets and both operating systems will likely interact for years to come. On the consumer side, I believe Microsoft is trying to offset the weakness its leading partners Dell and HP have displayed in this key, easily commoditized market. This surely is the logic behind the Surface RT release, which is a highly differentiated product with appealing engineering, however well (or poorly) it does.
On final comment: While journalists seem to fear we will all be baffled by the touch interface, the fact is that it is very easy to navigate and navigate away from. You do need a touch activated display however!
Thank you for your cogent recommendation. I am working up an analysis of GT and share many of your viewpoints. The stock was very cheap before they reduced the guidance so dramatically. It remains with an apparent attractive value, with the stock down sharply and guidance dramatically reduced.
Management has capitulated, indeed, guiding to a 25% decrease in revenues next year and a 31% to 44% decrease in cash and equivalents. GT has strong leadership in key technologies in each of its business units, with solid new revenue-generating products lined up in each once the markets recover. Management executes well and to date has kept ahead of its massively volatile revenue streams. The issue is that management is now suggesting these recoveries may not generate revenue growth until 2014. So, roughly a year from now, things could be very tenuous in this scenario (suggested by their guidance).
I am not done with my analysis, but sense this will be a strong Buy at some point. Right now, though, it still feels a little early. Downside may seem limited, but uncertainty remains high. I agree that a Buy now will be in the money at some point. The ride may be a bit bumpy, however.
Thank you for prompting such a thoughtful exchange. Courtland's main argument is that the research and capital burdens of getting to 28 nm have reduced the number of semiconductor companies that manufacture at the leading edge from 23 to just four. And of the four, two are foundries that manufacture the designs of other companies and two are IDMs that manufacture their own designs.
There is a lot she has left out to focus us on this basic issue. The adoption of 450 mm wafers and DUV lithography present additional challenges, but three of the four have allied with ASML to move forward in some sort of Intel-led tandem. No doubt, Applied and Lam will progress down the road map as well. And as a major manufacturers of NAND and DRAM memory, Toshiba and Micron have to be planning for a seat at the table as well. And then their are the numerous fabless companies that sell advanced logic at the "Intel-minus-one" and "-two" nodes, such as AMD, nVidia, Apple, Qualcomm, Altera and Xilinx.
I think Otellini is correct in suggesting the industry will experience significant change. And he is no doubt correct is stating that Intel, while engaging in foundry activity, will not manufacture devices for its competitors (i.e., AMD, nVidia, and likely Qualcomm as well). So while it is bound to get ugly for some, there will likely be spots for Samsung and TSMC, as well as Intel. Even as the capital burden rises, the economics for high volume advanced memory (DRAM and NAND) and logic (Processors, FPGAs and select ASICs) will continue to work, at least for a node or two more.
Actually, IBM operates a sizable Semiconductor business and conducts research on leading edge technologies, as last week's exciting news suggests. They manufacture much of their non-Intel processors which are used across their server product line. Since Apple abandoned the PowerPC, the business has operated "under the radar". Back when Global Foundries was on AMD's books, their were process technology agreements that survive to this day, as others have noted in this exchange
Thank you for your paired recommendations on HP shares.
Some observers itch for Whitman's departure, though continued management turmoil is hardly a good thing at this juncture. She has articulated a detailed turnaround plan and shared it with investors. I am guessing she will be given the chance to implement that plan.
The goodwill from most recent acquisitions has been written down in a series of non-cash events that nevertheless reduce book value and the metrics that are embedded in its debt covenants. There is something of a clean slate, though.
The stock is a nice dividend payer and on the fiscal 2013 EPS target (which corresponds to the following year consensus), the stock has a PE of 4.1. This is incredibly cheap. The stock is up 20% or so from its recent trough. As I argue elsewhere, I believe the shares will continue to find value buyers, limiting downside and enhancing upside near term. Key things for me to track in the year ahead are HP's traction with Commercial Tablets, hybrids and hyperscale computing platforms.
Thank you for the link and comment. That was Rich Gardner back in 2010--the good old days when Hurd was under-investing in R&D and dining with pretty ex-actresses!
Ahh, to think of the mess that has come since!
Yes, the Citi downgrade got me to thinking how out of line the valuation is, even with the fundamentals being what they are. Even with his negative metrics, it is very cheap. He did reiterate the Sell rating more recently. I can track it down if you need it.
Again, thanks.
Thank you for reading this and contributing a comment. Yes, beauty is in the eye of the beholder. I read Suva's sell rating in a report dated 11-20-12, when the stock was about $12/sh and his target $10.50/sh (down 12.5%). It actually prompted me to consider a contrarian call because the valuation was so out of line with the revenue and EBITDA HP will generate even in his negative scenario. Of course the stock is up 20% at present.
That the company would induce a breakup seems unlikely to me, given the reception of Apotheker's initiatives. That a corporate raider would try seems unlikely because the leverage is so high. Who would commit a lot of equity to such an undertaking? I do feel that an attempt to equitize its Enterprise businesses in some way would be very interesting. But I do believe a period of stasis, after such turmoil, is the most likely course
Again, thanks for your comment
Thank you for your input. Love your comment too.
To turn this rig around, she'll need the ocean blue!
The time frame is lengthy and seems open to accommodating ongoing change in the company's markets. So we will see.
Again, thanks for taking the time to read this
Kudos to Ashraf, Dana and the community for sponsoring such an informative exchange of view points on what is a very large and hard to comprehend topic. Indeed, the exchange in many ways exemplifies the complexity.
Dana wrote the piece to which Ashraf responds upon reading a CNET report on Linley Gwennap’s recent post on the Apple A6. (Dana provides the trail above.) Long the author of The Microprocessor Report, Gwennap is a leading authority on MPU and SOC design. Notably, he reports that the A6 is in fact based on a customized ARM core by a design team that originated with its 2008 acquisition of PA Semi. (Note: processor cores, MPUs and SOCs are all customizable devices; given designs are fixed for specific high- and low-volume applications.)
Typically, ARM licenses a trademarked Processor Core with defined design parameters. However, Gwennap reports that Apple negotiated the right to develop its own ARM compatible CPUs, one of the few companies able to do so. This no doubt reflected Steve Job’s goal to have Processor that was not only “insanely great,” but optimized to its unique Apple-specific purpose. Gwennap notes that Apple as a result has beaten Samsung to market with Cortex-15-class CPU by a three month period that encompasses the holiday selling season. (Note: the Cortex 15 is a recent licensable ARM core.) It is ironic that Samsung is the foundry that manufactures the A6 for Apple.
I also want to offer a couple of related points. So much of the dialectic seems driven by the notion that Intel is toast, stranded with yesterday’s moribund offering, selling into yesterday’s moribund platform. And, indeed, the proliferation of Tablets and Smart Phones deprives the PC market of that growth that results from activities better performed by these new devices. That said, computing on PC’s and Servers will continue to generate demand for millions and millions of advanced processors manufactured by in large by Intel.
More to the point, Intel’s effort to secure sockets in tablets and smart phones betrays a greater nimbleness than most investors understand. First, as the company discussed at length at last week’s IDF, it has adopted an SOC approach in developing processors for these new markets. It has even optimized its process development around alternative CPU and SOC offerings. It appears they have moved their strategy far beyond porting Atom processors down the roadmap until they are more power/performance efficient than ARM. It is clear that not enough of us are keyed into this. In other words, Intel management has been planning around this for some time.
Good points. The neat thing about Tablets is they displayed no traction until Apple launched the iPad as a Smart-Phone processor-based device. The beauty is that it is not a compute device and the resource demands of a Intel Microprocessor are entirely unwarranted. So Intel's absence is not that surprising to me, especially since iPad-type tablets are still relatively a new platform. I agree that Windows 8 offers one route into this market for Intel. But I also feel that Intel's true traction in the Handset and Smart Phone offers another one.
I did not comment here on valuation and have both in my model portfolio as top tier Technology Plays. As an IP provider, ARM captures a sliver of the end market value of its products, hence its very high valuation metrics and low revenue model. Even as an IP provider, it commands a valuation that is extended to even a greater degree by its remarkable traction. So if ARM falters in any way, a sharp contraction in these metrics will result. Any sign of Intel's success in the Smart Phone market could provide this trigger. Moreover, the likelihood of such success, even to just a modest degree, has never been higher. So, yes, things are fundamentally changing and there is a classic pair trade in play here.
We have two great companies here with amazing technology poised at the frontier of each others' markets. The slightest tip of Intel towards ARM's core markets will reveal a profound mis-allocation of capital, and a sharp contraction of ARMH stock will result. The economic impact of any Intel success may be trivial, indeed. But ARMH shares could be crushed, as they say.
Dana, thanks for the reply. I agree that strong tablet sales are muting PC demand. But I don't think that is what has produced the revenue miss, given its magnitude. Intel identified channel de-stocking, weak Enterprise PC demand and weakening emerging market demand as factors behind the miss. Tablet demand is as they expected, I'm guessing.
Dana, thank you for inspiring this spirited and interesting exchange.
A couple other points worth note: Intel identified the factors that produced the large revenue miss; ARM Holdings is not mentioned and the impact of strong iPad sales was likely fully factored into prior guidance.
I have a comment on the design flexibility of the ARM-IDM-foundry model: this has always been a strong selling point for ARM in the handset market, as has its low power design. That said, Apple with its iPhone has sparked an emphasis on systems and platform integrity that could ultimately require more uniform processing solutions and this could well foster Intel’s entry into the market. It remains to be seen.
Sufficient power efficiency is the byproduct of Intel’s Technology Roadmap, as it executes one node ahead of the rest of the industry.
The competitive dynamic between Intel and ARM holdings is as interesting as it is complicated. I have written on it in the past and expect to do so again soon.
The ASML press release, which differs markedly from the Intel release, makes it very clear that other customers can participate in this "Co-Investment" program and purchase as much as an additional 10% of the company. Note that Intel is consistently well ahead of any other semiconductor manufacturer and its lead will likely remain intact.
Thank you for the comment. I would say Itanium is what it is and has been for the better part of the decade. High Performance computing has always been a niche market. And as you note, the performance capability and growth has come at the x86 end of the market, where the road-map is indeed much faster. Clustering, Blades, Virtualization, etc. are indeed facilitating a move to utility Cloud computing, both private and public. That said, specific high performance computing applications will continue to have the requirements that Itanium-based and competing systems satisfy. It is (on a unit basis) relatively small. But it offers (and integrates) some of the most advanced technologies achieved, albeit to relatively narrow ends. Economically less significant? Yes. A bit quaint? OK. But these applications are of critical importance to those who require them.
Thank you for adding this data. I didn't have the IDC analysis and data handy, but hope to dig them out soon. Unix system ASPs are much higher, so market share on a Unit basis is lower. It is very instructive to view analyses that provide market share by units and revenue with ASPs. Margins are higher, but so is the engineering and service burden. And service support provision for the systems vendor is valuable add-on business as well.
Thank you for the comments and questions. I reported management's view that a market with four surviving participants would be sufficiently competitive. I am very interested to see if this plays out. Regulators have minds of their own, after all. I did not have time for a more thorough market analysis, though I am eager to perform one. The market dynamics vary considerably by form factor, as does the concentration to some degree. I agree, too, that the market power of Dell, HP etc., will likely be a key factor, however the deal proceeds. Regulatory guidelines regarding competition and market concentration ostensibly are systematic (something else I hope to return too), though the process can be politicized as well. Let’s see what happens. I am eager to write more on this topic.
Hats off to the author and commentators for the interesting discussion. For me the key issue seems to be the degree to which MIPS, with its heritage in robust processing applications, can deployed in systems with the sparest of power requirements. So the GreenDroid merits a close look. In addition, it seems that proliferation of tablets and other devices, which so clearly is benefiting Intel, offers MIPS opportunities as well. Still, we could be in for a bumpy ride, given the stock appreciation over the past six months. That said, from either side of the trade, it is worth a closer look.
Kudos for the insights. Still, it is important to point out the Lumizyme approval and related manufacturing issues are separate from the viral infection that halted production of Cerezyme and Fabrazyme at the Allston plant. The company has moved beyond remediation and is well through its first run of production on both therapeutics. It has transferred production of Myozyme to the Belgian plant as is seeking approval there, which should remove pressure for the approval of the related Lumizyme. (An aside: because Myozyme displayed slight variations when manufactured in 2000L bioreactors, the FDA required a new appellation and approval track.)
The impact on the marketability of competing products to Cerezyme is unclear to me. Anecdotal evidence suggests many patients could adapt to the temporary shortage of both Cerezyme and Fabrazyme. But consensus estimates encompass a return to 20% annual growth. A stock that serially stumbles deserves its detractors. That said, if the company meets its established mileposts (at long last, indeed), the smarter money will be long at this point in time
In my view, the logic of this deal has always been suspect. Yes, the value of Sun’s technologies is understated by the market cap that its existing financial structure allows. Unlocking this value through some sort of dis-amalgamation of Sun’s legacy and forward-looking businesses should be Mr. Schwarz and Co.’s primary goal. But IBM could hardly be the key to this value creation, because it directly competes against Sun in most of this latter company’s core businesses. Would IBM support Solaris against its own AIX? Though most of us hardly think about Unix computing these days, it is critical to most of Sun’s customers. And why acquire Sun if you don’t need (and shouldn’t acquire) Sun’s Unix-server and StorgeTek businesses? Indeed, if IBM’s goal is to acquire and retire these legacy businesses, anti-trust enforcement will indeed preclude its completion.
My advice to Sun management: It is time to slice and dice. In a normalized economy, technologies and businesses must not be supported with the expense structures of bygone eras. With regard to your products, show us (as Michael Capellas once said) what’s under the Kimono. Once these businesses are more profitably and clearly reflected in financial results, higher valuations will follow. Once non-core middling’s are gone (or at least disaggregated), investors will accord shares the appropriate valuation.
I agree that both companies should be left to stick to their knitting. I have covered both Vishay and International Rectifier for over a decade and knew previous managements of both companies pretty well. And I eagerly awaited IRF's restated results in plotting the purchase of its depressed stock. No sooner were they here than Vishay pounced with a paltry premium.
Surely IRF was brought to its knees striving to seek the highest value in its franchise through serious accounting and business malfeasance. That said, the reporting mechanisms are back in place and a new management presides over a higher value Power IC and Discrete franchise that is intact. Even in its weakened state, this business model is inherently more profitable than that Vishay has so capably pursued over the years. It requires higher levels of investment and different channels that target different sorts of customers.
It is not clear to me why Vishay is expending such effort to pursue what seems clearly to be an ill-conceived business combination. Granted, IRF shares are severely depressed, and the accounting scandal has strained its business. Yet a cheap price alone hardly justifies this intervention in IRF's turn-around. And given its size and difficulty, it could very well gum up Vishay's execution as well. For now, I am on the side lines (and unable to vote).