I am a big fan of SeaMicro’s technology even if I sometimes come down hard on AMD (NYSE: AMD). I think AMD has a lot of problems that do not bode well for it, and hard decisions will have to be made. That might lead the share price down before it starts back up, but without it the company could become irrelevant and then fold. The processor business seems to be the biggest drain on the company, and it is known as a processor business. It is falling behind in the GPU business as well. AMD needs to establish something new that can lead AMD into the future, as well as catch up in GPUs.
Didn’t AMD sell this building? AMD’s Austin campus (Photo credit: rutlo)
The SeaMicro acquisition is all about density, energy-efficiency, and space saving. It is about packing as much capability into as efficient and small a package as possible. That allows these massive data centers of the future to be done more effectively. The love affair the market has with the cloud will require massive and numerous server farms to house the cloud. The cloud is not incorporeal, it is just over there instead of right in front of you.
Intel’s Perpetual Competition
Intel (Nasdaq: INTC) is entering into the microserver space as well with its Atom platform, and it acquired its own version of SeaMicro’s Freedom Fabric. Intel bought Infiniband. Good news for AMD is that the microserver market is new, and they have a chance to carve out a place for themselves. With good leadership and research it might become a dominant player. The bad news is that Intel is a far stronger company than AMD, and it is the leader in the processor market when it comes to PCs.
Intel has more cash and less debt. Anything it can’t develop internally it can acquire. AMD is forced to develop internally. I do not think it has the balance sheet to go out acquiring companies unless it is for relative pennies.
Intel may seem like the favorite, but it has no reason to move out of PC-chips. It does need to explore new areas, but the company is not the turnaround story that AMD is. I think that means lower potential gains. Also, I think continued news about how amazing tablets are and how sales are increasing will weigh on Intel’s share price. Traditional PCs are Intel’s bread and butter and seeing a profitable business decay will do more harm. I would avoid Intel since it is not doing poorly enough to have the potential of a turnaround, but it is not doing well enough to suggest a rip-roarin’ future.
Microservers and Fabric
Microservers are a small portion of the giant server market, but can grow substantially if the technology takes root. AMD has the potential to become a dominant player in this market and could lead the way into making it a significant part of the overall server market, which is very large and profitable.
The keyword is density.
Everything I have read thus far about the SeaMicro acquisition focused on the hardware. It took some digging to read about Freedom Fabric, which basically allows clustered processors to work effectively as a group. The whole thing would be an inefficient mess without the fabric. I have not found anything that compares different fabrics so I assume that AMD’s belief is that they bought something that is more efficient with more units in the cluster.
The Indirect Beneficiary
A hidden part of this story is ARM Holdings (Nasdaq: ARMH), which is the like the facilitator of many successful products, such as Nvidia‘s Tegra and Apple’s A5. The company has an insane PE in the 70s, but it could be argued it is justified. Tablets, smart phones, and these new high density microservers will be using ARM-based chips. That means more licensing fees for the company. The revenue growth for the company has been phenomenal in the 20% range for each quarter, and earnings growth is just as good.
A small position in ARM does not seem like a bad idea considering tablets are only getting more popular. I am concerned with the valuation though. There are a lot of misunderstandings around the company. Just because a chip is ARM-based does not meant that ARMH is the one making it. Also, a business that relies almost completely on royalties and licensing will see revenue that is a fraction of the entire industry it is supplying to. Licensing will not be as profitable as selling an iPhone and I think the market will realize that. It is a good company but a bad valuation. I think a correction is due for ARM and I would write puts to take a position, or you could sit on a limit order waiting for the price to come to you.
For the microserver market, ARM has an opening, but the real risk is being taken on by AMD. Microservers are not 100% ARM -based. Intel’s x86 instruction set shares the market, and is the incumbent. However, if ARM-based chips with AMD’s fabric prove superior it could capture most of the market. AMD will continue dealing with competition from Intel and any newcomer that tries to break into the market once it is validated, but ARM will be licensing to most of the players if it comes out on top. That will expand revenue like it did when smart phones started using ARM-based chips.
If you believe in AMD’s turnaround it is worth a speculative investment at best. The restructuring will not be complete till at least late 2013 and I would not be surprised if the cuts are deepened and the time period is extended. It will be minimizing the importance of its traditional chip business, which is a good move. The ARM-based Opteron servers will not be till 2014, so any revenue hope from that segment is far off. Until then I hope that AMD can sell some SeaMicro boxes since it bought the hardware in its acquisition too. I remain on the fence, until I see something regarding 2013 that is not the restructuring. I am getting frustrated looking for openings though, because I keep waiting for something around the corner and I only see things down the hall.
Related articles
Emerging Technology