You are currently viewing SemiWiki as a guest which gives you limited access to the site. To view blog comments and experience other SemiWiki features you must be a registered member. Registration is fast, simple, and absolutely free so please, join our community today!
Let's face it, the whole Intel buying Altera thing is a rumor even though most reports ignore that fact:
WSJ: Intel Corp. is in advanced talks to buy chip partner Altera Corp., according to people familiar with the matter, a move that would represent the semiconductor giant’s biggest-ever acquisition.
Bloomberg: Intel Corp. is in talks to acquire Altera Corp., people with knowledge of the matter said, as the world’s largest chipmaker searches for growth beyond a moribund personal-computer market.
ALTR went from $34 to $44 on the news so I'm sure the SEC is looking closely at the trades of ALTR stock to make sure those people "familiar with" or "have knowledge of" the matter are not profiting unfairly from this rumor.
Personally I hope the rumor is true. Altera needs a serious management change and an acquisition would take care of that most certainly. Altera also needs help on the manufacturing side and Intel is the leader so that takes care of that. But most importantly it is a very disruptive move and that keeps innovation alive, absolutely.
This leaves me with three questions that I hope you can help me with:
Who started this rumor?
What happens to ALTR stock if there is no acquisition?
What happens to Altera as a company?
Any other questions I should be asking? I have discussed this at great lengths with my Silicon Valley friends but before I blog I would like your input. Help me out, I don't want my Altera bias to show too much here.
1) WSJ/Bloomberg will NOT divulge their sources, but do have rigorous ways of not reproducing lies/rumors. They require at least 3 INDEPENDENT sources to verify the info, when nobody will go "on record". It does not mean that info will produce an actual event, but it means that the info is solid, and is not just BS.
2) Obviously ALTR shares will decline in price, if no deal is announced relatively soon. Did you really need to ask this question?
3) ALTR will benefit from Intel's marketing machine. Intel needs growth, and so ALTR would most likely benefit from Intel using them as a growth vehicle. There will be some redundancy, and that means people losing their jobs..... but, there appears to be strong industry demand for the type of people that work at ALTR.
Your bias is already evident from the questions you asked......
In my experience people do not leak this kind of information on a whim, they do it for a reason. One reason may be to drive up the price. Intel will have to pay billions more for Altera as a result of this leak for example. Sometimes a leak like this is to get a suitor where there was none or additional suitors involved to create a bidding war, right? So again, who leaked this and why?
Obviously the price of ALTR will go down, but by how much? Will Wall Street punish them with lower pricing than before for a failed acquisition attempt?
Intel and Altera have very different cultures, Intel also has a spotty acquisition record, this is far from a slam dunk accretive acquisition.
Things have changed in the last 15 years Ken. Transparency is here, there are very few secrets in Silicon Valley, and like it or not the sources and motivation of this leak will probably come to light. I started this discussion so we can look back and see how good of speculators we all are. Nothing more than that.
Any merge/acquisition has pros and cons. IMHO, it may turn out to be good for Intel if Intel does NOT acquire Altera
The reason:
1. Pay $10 ~ 13 billion to acquire a $2 billion revenue company, Altera, seems too expensive and the benefit is not significant enough to justify the price.
2. Intel's financial strength is not as good as used to be. To put so much capital into Altera might strain Intel's ability to do other big and important acquisition. I have posted a related comment in another thread:
Intel's "current ratio" (measures a company's ability to pay short-term obligations, bigger is better) is as low as 1.73 already and currently it has only about $14 billion cash in hand.
In comparison, even AMD has a better "current ratio" than Intel.
Intel 1.73 SK Hynix 1.79 AMD 1.9 Micron 2.19 Samsung 2.21 MediaTek 2.45 TI 2.92 TSMC 3.12 Broadcom 3.27 Infineon 3.28 QUALCOMM 3.55 Xilinx 3.85 Altera 5.72
hist78, in 2014 Intel "repurchased $10.8 billion of common stock through our common stock repurchase program" If Intel decides that investing in ALTR will produce better returns to shareholders than investing in Intel shares, then the acquisition is financed by cash generation, cash, and borrowings at extremely low rates.
Intel owns $7.1 Billion in Marketable equity securities, that is in addition to its $14 Billion in cash & equivalents.
Intel can EASILY afford to buy ALTR for CASH, if the BOD has determined this is in the best long term interests of its shareholders!
hist78, in 2014 Intel "repurchased $10.8 billion of common stock through our common stock repurchase program" If Intel decides that investing in ALTR will produce better returns to shareholders than investing in Intel shares, then the acquisition is financed by cash generation, cash, and borrowings at extremely low rates.
Intel owns $7.1 Billion in Marketable equity securities, that is in addition to its $14 Billion in cash & equivalents.
Intel can EASILY afford to buy ALTR for CASH, if the BOD has determined this is in the best long term interests of its shareholders!
I'm not questioning if Intel has capital to buy Altera. My concern is the $10 ~ 13 billion price tag for Altera is too expensive to justify, even if Altera is an excellent company. This is NOT Intel CEO or President's money. It's shareholders' money! If $10 ~ 13 billion goes to Altera, can Intel afford to buy another big company, for example: Broadcom or Mediatek, or UMC? Yes, they can always borrow money or issue more shares. But a smart CEO and board will do that extremely careful.
Talking about smart CEO and smart corporation board. A company can have good and bad reasons for buying back stocks. One good reason is the "company' revenue, profit, and cash growing rapidly that the excess amount of cash can't be consumed quickly just by expansion and acquisition. A good example is Mr. Buffett's Berkshire Hathaway and Apple's stock buyback program.
On the other hand, a company can still use their hard-earned cash to buy back shares in order to increase their stock price because they run out of good and big idea to grow their business. I hope Intel is not falling into this trap.
From my research the move by Intel seems a move to defend their data center and cloud business. It seems FPGAs can be used to make data centers handle data far faster and with substantially less power. There are several articles online about this from reputable sources. Xilinx is still the winner, which calls into question the fundamentals of why the deal should be done.
I take a cynical view of this potential acquisition of Altera by Intel. From my experiences of these acquisitions in the past, none of them of have gone that well. If Intel is not careful, it will end up being played for a sucker, again...
None of these companies are open to acquisition unless they really need help and are raising the white flag. Either the target company is in more trouble than it is letting on, and needs someone with deeper pockets to help boost them out of their hole. Such issues can be in the form of new competition or changes in the market environment, missteps of their existing product line(s) that have not yet become publically visible, or some other deteriorating business condition(s).
Intel has a history of not performing enough diligence to flush out such issues ahead of time, and if not careful here could end up with another white elephant. As an example here from the past, Intel main interest in the acquisition of Level One Communications was for its gigabit ethernet controller products, based on its leadership in its previous ethernet controllers. But after the acquisition it became apparent that Level One had some serious product stability issues with the first gigabit ethernet controllers supposedly ready for market. It took additional unplanned and massive engineering resources from the Intel side and another couple of years before they got a stable product to market. Meanwhile many of the original more talented Level One employees cashed on the merger deal, took their money and left at the first opportunity. By then, Marvell and Broadcom passed Intel by in this product space. In the end, Intel owned a communication chip company with a lot of second-run products, gutted of its technical leadership, and no roadmap to regain or hold product leadership in this space.
As Arthur H. has indicated, I would also be reticent about getting involved with a second-run FPGA vendor like Altera, when you have the likes of Xilinx who have a clearer leadership in this space.
The second motivation I have witnessed for such acquisition rumors is for the purpose of creating a short term spike in the price of the target company stock. Just having the rumor of a possible acquistion by Intel will be enough for some serious profit making in the short term by some unscrupulous insiders. At this point, the deed has been done. If you own Altera stock now, and the Intel acquisition rumor falls through, grab your ankles and get ready for a rapid descent.
From my research the move by Intel seems a move to defend their data center and cloud business. It seems FPGAs can be used to make data centers handle data far faster and with substantially less power. There are several articles online about this from reputable sources. Xilinx is still the winner, which calls into question the fundamentals of why the deal should be done.
Currently a lot of computation acceleration is done by using the GPU through OpenCL; Intel's GPUs are not up to the same performance as AMD's or NVidia's. A next step is to use FPGA's for OpenCL acceleration and I think Altera was a pioneer here. To me replacing the Xeon core + GPU with Xeon core + FPGA for server chips makes indeed a lot of sense for Intel.
Tweet on Altera-Intel talks came after options trades
Not surprised at all:
Tweet on Altera-Intel talks came after options trades (Reuters) - A March 27 tweet sent the same minute as news broke that chipmaker Intel Corp was in talks to buy Altera Corp appeared to come after very timely trades in Altera's options by several seconds, according to Thomson Reuters data. The tweet by Wall Street Journal reporter Dana Mattioli, sent the same minute as the headlines, led to speculation that it may have spurred the timely options trades. The tweet, however, was sent 19 seconds after the initial options trades.On March 27, short-term call options conveying the right to buy Altera shares at $36 and $37 traded in unusually heavy volume at 3:32:39 p.m. EDT, according to Thomson Reuters data.
Currently a lot of computation acceleration is done by using the GPU through OpenCL; Intel's GPUs are not up to the same performance as AMD's or NVidia's. A next step is to use FPGA's for OpenCL acceleration and I think Altera was a pioneer here. To me replacing the Xeon core + GPU with Xeon core + FPGA for server chips makes indeed a lot of sense for Intel.
I would agree with the advantages of the GPU in this space given they tend to be sufficiently programmable with a higher degree computational datapaths that can be helpful for certain compute server workloads.
The use of FPGAs for customized acceleration has been explored for a couple of decades now. But it still seems to come up short as a solution for broader deployment due to inefficiencies in the nature of the reprogrammable aspect of FPGAs with respect to power, max clock rate, and area/cost. They seem to be a better fit for lower volume solutions or as a prototype ahead of a cost reduction to a full fledged ASIC. When used in a server context (or any other compute solution under general purpose OS), the computational state also needs to be quickly swappable for process context changes or for virtualized environment support. This is another deficiency that tends to get overlooked in such acceleration hardware.
While I have no clue about 1 and 2 (for 2, the logical answer would be that the stocks goes back to were it was before - but what is logical on stock markets?), my opinion for 3 is: Nothing. Both companies (Xilinx and Altera) will continue the battle they have since 20(?) years on a quite even level, both with good margins.
(The much more interesting question is, what happens if Intel buys Altera: If done right, it could be a big win for both Altera and Intel. If done wrong, it could be the opposite.)
What could do harm to Altera would be a long period of uncertainty, as this opens the doors for FUD.
This is not a surprise, and frankly I am glad to see that Intel did not get suckered in on this one. Look at the insider trading records on Altera reported in the next quarter to see who benefited.
