Samsung's second front

Most generals would try to avoid warfare on a second front, the diversion of resources and focus are often lethal. But sometimes when the opportunity presents itself you need to take it. Could Samsung - whilst they are currently battling Apple - be about to open up a second front; this time on Intel? This is the argument for doing so.

Suffering from memory loss

The memory business is maturing fast. PC growth is over & the industry must adjust to a different set of dynamics. Capex has been reduced this year and will be cut further in 2013. The supply base is starting to consolidate. Samsung will still be the king of memory when we have moved into the next phase but once there they will be forced to look at other sectors for continued growth.

Samsung have developed burgeoning logic and foundry businesses in recent years. They are already a force in this arena and one of the real threats to TSMC. That said they will likely lose part or all of their Apple A6 business next year and although will have lined up some new customers in 2013, this must still be a concern.

So memory is maturing and will need less capital going forward. The foundry and logic business is doing well but competition is going to increase. So where does Samsung go next? Where else, but to the money! To the last remaining large profit pool in the industry. To Intel.

This is not a new idea, there has been plenty industry chatter that Samsung could make this move, initially into server CPU and then potentially the PC. It would however require a monumental effort not just to develop the products but also to bring the server industry along with them in adopting the ARM architecture in the mainstream. That said, toppling Intel is never going to be easy.

We never saw that one coming

When Andy Grove and Intel made the decision to leave the DRAM business and focus on CPU's for the then nascent PC business, they gave birth to the industry's original 800lb gorilla. Through their masterful control of the PC platform (with Microsoft) they rode the phenomenal growth for over 25years, became a virtual monopoly and delivered staggering returns.

But the PC growth story is over. And like most monopolies before them, Intel became too inward focused, with the result that they totally missed the Smart Phone and Tablet revolution. 

It is debatable whether or not Intel will make inroads against ARM. But what is not debatable is that they will never dominate Smart Phone and Tablet like they did the PC and therefore cannot deliver the 60%+ gross margin's the company and it shareholders have grown to depend on. So in effect the Intel we all know is finished. The huge margins, the control, even the swagger; it's gone.

Intel warned investors earlier this month that Q3 revenue was going to miss expectations. These things happen you say, so not necessarily a big deal? But what surprises me most about this miss is that they've managed to put if off for so long. And now that it has happened you would suspect it is going to happen with more regularity as long as PC sales continue to fall.

The cure is worse than the disease

To have any hope of re-igniting PC sales again the industry needs to deliver on both price reductions and innovation. But the Ultrabook debacle this year would have to worry you on both counts.

It should have been one of the hot products this holiday season, a combination of sleek design and an attractive price-point would have taken share of mind (and wallet) of many consumers. Instead the PC makers effectively dropped it in Q2 because they couldn't make the numbers work. Many have played the blame game since but you can trust that the key reason the UItrabook has failed to deliver this year is because Intel was looking to keep too much profit for itself. So if Intel cannot be trusted to do the 'right thing' ahead of protecting their historical margins, then I am struggling to understand how they expect to get PC numbers out of reverse in 2013.

One other thing to consider is that despite the huge profits Intel have garnered over the past 25 years, they are not exactly in a bulletproof financial position. Look at this simple analysis of their balance sheet. By necessity the Intel model has required gigantic capex every year to keep building those fabs and maintaining their technology leadership. A couple of years of over-investment here and you could really do some damage.

Kick him when he's down

Intel are vulnerable, more than they have ever been since they exited the DRAM business in 1985. Once they finally accept that their world has changed forever they will then be consumed with the massive re-structuring and re-focusing that implies.

Samsung needs to seize the moment, by mounting a concerted assault on the Server CPU market, promoting and helping to enable the ARM architecture across the industry. The strategy should not be just to take a small slice of the market but to go after Intel and to do some real damage. The PC CPU may follow in the second phase but Server must be first up and kick-off by 2014.

Samsung Semiconductor's achievements in the past 20 years have been close to those of Intel's. When you consider where they have come from and the fact they have essentially cornered the entire memory market's profits for over a decade. They have pre-empted, adapted & excelled at all the changes the industry has thrown at them in the past 10 years. Intel on the other hand has pretty much missed them all but has been protected by it's virtual monopoly position in PC & Server.

Many years ago Samsung management distilled the semiconductor business down to two key ingredients; capital & timing. If they want to take on Intel and become the #1 semiconductor company by the end of the decade, they need to be bold and act now. Even if they have to open up a second front to do so.