Advertisement

Blog

Micron Rising, but Can It Catch Samsung?

Marketshare matters greatly in the semiconductor memory market. The sector is one of the most volatile in the electronics industry, and wild swings in pricing, demand, and supply have resulted in the demise of companies in Asia, Europe, and North America, including Taiwanese suppliers and Qimonda, a spinoff of {complink 2565|Infineon Technologies AG}.

Scale is also highly important in the sector. The bigger a company is, the easier it is to fund capital expenses, withstand rough cycles, and snag more marketshare. That's why the deal by {complink 3421|Micron Technology Inc.} to purchase a Japanese rival, {complink 1769|Elpida Memory Inc.}, is so significant for the Boise, Idaho, company. For the first time in more than a decade, Micron stands a chance of boosting its marketshare and improving its odds of surviving another market winnowing.

The deal, announced July 2, would lift Micron to the No. 2 global position in the DRAM market, according to IHS. The research firm says Micron would command almost one-quarter of all industry sales, putting it ahead of Hynix Semiconductor Inc. of Korea. {complink 4751|Samsung Electronics Co. Ltd.}, the perennial market leader, would still have a marketshare of almost 41 percent.

Mike Howard, senior principal analyst for DRAM and memory research at IHS, wrote in an emailed research note that buying Elpida would be a huge boost for Micron's status in the DRAM sector.

Micron will see its market share and DRAM manufacturing base nearly double as a result. Furthermore, Micron is gaining access to some excellent mobile DRAM technology, which should greatly improve its product portfolio. The $2.5 billion sale price is reasonable and shouldn't impact Micron's cash position adversely. It will likely take at least six months for the deal to close, but IHS expects a very quick transition and integration once it does.

There are other benefits for Micron and other leading DRAM vendors. Elpida had previously announced bankruptcy proceedings, a move that sharply reduced plant utilization and supply, which in turn stabilized DRAM pricing. IHS said that before the Elpida bankruptcy filing, “the DRAM industry in 2011 had been bedeviled by excess DRAM manufacturing capacity, which drove down prices and caused revenue to decline.” After the transaction closes, Micron is expected to further reduce Elpida's manufacturing footprint. According to IHS:

The consolidation of the DRAM market set off by Elpida's bankruptcy and the subsequent purchase by Micron is bringing new stability to DRAM pricing, helping lead renewed growth to the market.
Global DRAM industry revenue this year is forecast to reach $30.5 billion, up 3.3 percent from $29.6 billion in 2011. Although seemingly small, the revenue expansion for 2012 is a welcome development given the market's stunning 25 percent contraction last year.

Though buying Elpida is seen as a positive move for Micron, it will doubtlessly increase pressure on small players as OEMs consolidate purchases. Companies like Nanya and Windbond, each of which has a marketshare of less than 5 percent, will find it harder to compete against the established players. Even Hynix will not escape unscathed. DRAM production is highly capital intensive, and the higher up the ladder a supplier sits, the easier it is to secure financing for plants and other production activities.

The combination of Micron and Elpida would put Hynix at a slight disadvantage while it is still sorting out problems with investors and financiers. Could Hynix make a move for smaller rivals to boost marketshare? That's unlikely, and it would not dramatically change its position, because its rivals have relatively small marketshare. For now, the wind is behind Micron, and it may pull farther away from Hynix.

Catching up with Samsung is quite another matter. The Korean giant isn't letting up on strategic moves to pile pressure on rivals. It plans to increase capital expenses sharply in the year ahead. That's not a move Micron can afford to match right now, despite its relatively stable cash position. This transaction would increase its cash obligations for the next several years. For now, Micron should just savor being No. 2.

15 comments on “Micron Rising, but Can It Catch Samsung?

  1. dalexander
    July 4, 2012

    Bolaji, Very good write-up. Micron's leadership role in the Hybrid Memory consortium with MIcrosoft now in, almost assures their continuing market share for advanced memory technology. The Cube ” is a major step forward with regards to bandwidth and performance. I went to the UBM sponsored Design Con where speakers talked about the various benefits of the cube architecture including the low energy for transfering and also the reduced latency rates. This translates to lower power and faster parts. When Intel came ont he scene and set the PC-100, PC-133 standards, it brought all the memory players with their technology. Micron is in a key position to help define the standard for the cube such that their memory can be perfectly tailored early on to qualify and catch the market for the cube memory technology.

  2. bolaji ojo
    July 4, 2012

    Thanks Douglas. Micron may have risked it all with this acquisition but it was also playing for market share and the opportunity to advance its technology. The transaction will open up new chances for the company to move to new grounds.

  3. chipmonk
    July 5, 2012

    Regarding HMC ( Micron's Hybrid Memory Cube ). TH eintended first use will be in large Data Centers where its use would reduce power consumption.

    It is still in definition stage as more players ( e,g. MicroSoft ) join the consortium. Samsung was actually one of the earliest recruits ! But Samsung had already by themselves developed and demonstrated ( at IEEE Conferences in 2011 ) an early version called Wide I/O ( and JEDEC pretty much canonized that version in a standard published in Dec 2011 ). W/o going into technical details the difference between the two approaches ( Wide I/O vs HMC ) is in the trade – off between latency ( performance ) and process complexity – with Micron taking the high road in order to have a unique & higher performance product for which they can then ask for a premium. Elpida too had actually developed and presented back in 2009 ( at Conferences I attended in Japan ) a 8 die DRAM stack. Hopefuly their process technology will now help the Micron HMC consortia to move ahead.

    But so long as Samsung remains a part of the HMC Consortia they are going to not just agree to a new standard but would also expect process know how.  So I doubt if by itself HMC is going to give Micron that much of a leg up over Samsung.

  4. elctrnx_lyf
    July 5, 2012

    It may not be easy for micron to catch up with Samsung with this acquisition. But they will definitely have significant market share and can grow faster compared to before.

  5. Barbara Jorgensen
    July 5, 2012

    Although marketshare is a big deal in semiconductors, I'm not sure Micron really wants to catch Samsung. Samsung has slowly and quietly been exiting a number of its older DRAM businesses, and relies on being the volume and price leader to keep remaining businesses profitable. Although Micron chose to compete in DRAM, it seemed to me they had a different model. Analysts don't seem to think the Elpida buy was overly expensive or risky for Micron, so they gained share without over-spending. If marketshare does become the ultimate differentiator in the market, Micron may be making more acquisitions…of more bankrupt vendors.

  6. chipmonk
    July 5, 2012

    The DRAM business is sort of like the Airline business ( ” How do you make a Million ? Start with 10 million and then run the operation for a year, you will be left with a million !! )

    Given the long term reduction in DRAM demand due to ongoing tra, and you will be left with nsition away from traditional PCs to Tablets or even Ultrabooks that use SSDs ( NAND to DRAM ratio is now at 16, soon to be higher ) no business in their right mind would want to get a higher market share of a larger but still loss making DRAM biz. Instead they would put the extra capacity out of circulation to stop the glut and falling prices. So they would go for a higher Market Share BUT ONLY of a SMALLER market overall.

    So the most logcal scenario for this extra DRAM capacity from Elpida Hiroshima Fab would be to mothball it for a while to stabilize DRAM price and then perhaps convert it to higher grade DRAM ( faster, less power ) for Servers ( need to instal 1 new Server for every 70 new SmartPhone / Tablet user ) or even NAND for SSDs. The totally out of the box option would be to turn Elpida's 30 nm DRAM Fab into a 30 nm / 28 nm Foundry to take advantage of the capacity shortage at sub 40 nm for SoCs etc. for SmartPhones / Tablets.

  7. bolaji ojo
    July 5, 2012

    @Chipmonk, The strategy you described has long been used in the DRAM market to mop up excess demand. Unfortunately, these companies also have a tendency for self-destructive behavior. They reduce manufacturing capacity in one cycle and soon hike production simply to make sure the competition is losing money but this eventually backfires.

    The intrigues in this industry could fill a book but the bottom line is simple: Micron didn't buy Elpida simply because it needed the additional facility or technology. The entire industry needed to take Elpida's capacity offline, even if only for a short time.

  8. bolaji ojo
    July 5, 2012

    Barb, Except that the other DRAM vendors in risk of bankruptcy really don't have much in market share. Samsung is the price and volume leader as you rightly said and it achieved this by driving everyone else to either the edge of or right into bankruptcy. Micron is a beneficiary of Samsung's aggressive play. Let's just hope it has another game plan ready when Samsung raises the stake.

  9. Barbara Jorgensen
    July 6, 2012

    Competition is usually a good thing for the market. If Micron emerges as a force in DRAM, what will be the effect? Usuauuly that means driving prices down, which will be good for buyers but not suppliers.Technology advancement? I'm not qualified to comment on that (thanks to Chipmonk for the analysis–very helpful!) but that's usually an upside. But, as Bolaji says, the DRAM market usually just reacts to boom-bust cycles and nothing changes.

  10. chipmonk
    July 6, 2012

    Good, so now you understand that buying Elpida was not just about gaining a larger market share ( as your article had implied ) in a loss – making product ( DRAM ) but to reduce DRAM production capacity so as to stay in sync with shrinking demand and then perhaps even hope for some rise in ASPs. Of course any price rise will help Samsung even more ( 45 % market share ) but it would at least stop the bleeding for Micron. And they would get the breathing space ( and some extra cash ) to cook up their next move.

  11. bolaji ojo
    July 6, 2012

    @Chipmonk, I was never in doubt that Micron needed to mop up excess capacity in the market even if my article didn't indicate that. Micron isn't a selfless enterprise that would say it was acquiring Elpida to do the entire industry a favor. Plus, it would be an anti-trust move if the company said this. You are right in your comments that this was a strategic action to help reduce production capacity.

    What I noted in my blog was essentially that acquiring Elpida will give Micron additional clout in the form of market share gain. Without the additional market share gain, the acquisition of Elpida wouldn't help Micron. It gives it scale and helps it move away from the smaller players. These ones also will benefit from Micron's action in terms of pricing stablility but it will also hurt them since they are no longer the first procurement source for buyers.

  12. Anand
    July 11, 2012

    I think the merger of Micron/Elpida would present Samsung with its most powerful rival yet, but the big question is can this merger help to solve the debt concerns of Elpida ?

  13. Anand
    July 11, 2012

    Micron may be making more acquisitions…of more bankrupt vendors.

    @Barbara,  According to you which other companies is in the buy list of Micron ?

  14. bolaji ojo
    July 11, 2012

    Micron will be assuming the debt obligations. That's part of the transaction and it will be making periodic payments to Elpida's investors and bondholders for the next few years. Elpida is already in bankruptcy proceeding and perhaps some of its debts may be written off to make it solvent again even if the Micron deal falls through.

  15. bolaji ojo
    July 11, 2012

    Micron isn't looking for additional manufacturing capacity. I don't see a pressing reason for it to go for any other memory manufacturer. Elpida would be enough.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.