And so it begins.

According to LCM’s Asia Tech Tour report of 25 June 2010, both Toshiba and Hynix have customers on allocation. There just isn’t enough NAND to go around.

Apparently Toshiba can only supply its major retail customers with about 70% of what they want.

As LCM’s Dan Amir put it, “We have learned that, in the past few days, Toshiba decreased the amount of NAND allocated to its major retail customers by 10%+ for 3Q. Most players can get only 70% of their total order from Toshiba for 3Q. Toshiba explained that demand is very strong from handset OEMs and Apple. Hynix is also starting to put some of its customers on allocation.”

Excerpts from the LCM report are included at the end of this post.

This has lots of implications for SanDisk. All of them good.

The stars seem aligned for profitability driven by unprecedented strength in pricing underpinned by surging demand and constrained supply.

Demand appears to be driven primarily by smartphones and tablets, where both embedded and removable NAND products are the only game in town. Predictably Apple is leading the way.

Apple’s iPhone 4 and iPAD are setting the pace. Both are NAND-based and huge successes. Amir notes that Apple itself now consumes 25%-30% of the total NAND supply in the world.

As competitors scramble to catch up, NAND is getting scarce. And there are no alternatives.

Back in the Q1 conference call Eli delicately and accurately described the allocation scenario now emerging, when asked about OEMs pre-buying an anticipation of a shortage in the back half of the year (now):

“ Let me answer it without, indirectly. We do feel that you’ve heard me say or us saying that in the next decade we think that flash will be bigger than you think and that’s really based on the mobile and the way it’s going as well as the SSDs.

Two major new markets that are very, very early, so the industry will need to build additional capacity that doesn’t exist. I think we all know that. And you know the key is really the timing of those because the return on investment by itself is tied to the timing.

Now, what that leads me to say is over the next several years there will be periods, I believe, of shortages of flash memory, and there will be times where people will understand, OEMs, particularly large OEMs, that flash is becoming very strategic to their business. You can view it as a commodity, but when you can’t get it at any price because it is just not there because your demand far exceeds the available capacity of the industry, and the cumulative demand of all customers, then of course people will start wanting to secure for themselves some of that capacity, so I do expect some of that may happen.

Of course we don’t know for a fact, but I would not be surprised if the strategic nature in the next few years is going to start becoming effective. It’s no secret flash is a very, very tough technology. There are very few companies that can really deliver leading-edge technology, very high volume, both from the technology and the manufacturing and the balance sheet. So the answer to your question is yes. I think there is going to be some of that but there are some OEMs will start taking steps to assure themselves of supply.”

If supply is as tight as it appears to be, OEM customers really have no choice but to pre-buy in order to assure themselves of NAND supply.

If SanDisk doesn’t discuss pre-buying in the upcoming conference call, I expect analysts to press the issue and, when pressed, for SanDisk to answer yes.

The big deal about pre-buying and allocation is that it is indicative of demand outstripping supply. Such strong demand means strength in pricing.

Stronger pricing translates to both higher revenues and stronger gross margins. As a bonus L&R revenues will get a boost too.

How this translates for the rest of this year and next, should be an interesting story to watch unfold as analysts scramble- playing catchup.

Looking at earnings estimates, I don’t think the street quite gets it yet.

The average SanDisk estimate for Q2 is $0.89 EPS; for 2010 as a whole its $3.82. Most remarkably, the average 2011 estimate is $3.64.

Personally I expect Q2 to be well over $1.00; 2010 to be closer to $5.00 than to $4.00 and 2011 to be over $6.00 EPS.

Significant new NAND supply looks like a second half 2011 story at the earliest. Even then, additions look modest. The latest BoA/ML report notes that Samsung seems to be planning on dedicating its currently available clean room space to DRAM and logic rather than NAND.

To significantly ramp it’s NAND supply, Samsung will need a new fab. New fabs take years and billions of $$$. See BoA/ML report excerpts below.

It appears that NAND flash is starting to swing from commodity to a strategic material. Not a problem for SanDisk.

**** excerpts from the LCM 25 June 2010 report ****

Lazard Capital Markets
June 25, 2010

Daniel Amir
Lauren Stoller

SNDK: Asia Tech Tour – raising SNDK estimates, price target to $59 from $52 on favorable NAND data points; BUY

On day 4 of our tech tour we have met players in the NAND market and collected data that make us even more bullish on SanDisk.

Toshiba beginning allocation. We have learned that, in the past few days, Toshiba decreased the amount of NAND allocated to its major retail customers by 10%+ for 3Q. Most players can get only 70% of their total order from Toshiba for 3Q. Toshiba explained that demand is very strong from handset OEMs and Apple. Hynix is also starting to put some of its customers on allocation.

Allocation leading to price increases. Our checks suggest that prices for July, which is traditionally a slow month, are expected to increase by 3%-5% on the spot market and contract price is expected to be at least flat. Players in the industry believe that 3Q supply will be tight and are preparing for price increases throughout the quarter. We believe that the sudden success of Apple’s iPad and the launch of the iPhone 4 is resulting in increasing demand for NAND flash. Apple now consumes 25%-30% of the total NAND supply.

Retail market starting to pick up. After a fairly slow 1H in the retail market, demand is beginning to pick up. Module houses in Taiwan expect a more normal demand pattern in 2H, with revenue +20%-25% Q/Q in 3Q. Despite the increasing 3bit supply, we continue to believe that supply will be fairly tight for the retail market in the near term, though stronger than we previously anticipated.

Data points very positive for SNDK. We believe that the increased tightness of NAND in summer and the pick-up of demand at retail is very favorable for SNDK’s vertical business model. We continue to see SNDK as a good derivative play on the success of Apple, leading to very strong 2H GM and sales beating Street ests. We are raising our 2Q forecast to $1.2B/$0.93 from $1.14B/$0.85 and 2010 to $5.1B/$3.92 from $4.7B/$3.48. We are raising our target to $59 from $52, based on 15x our new 2010 EPS estimate. Biggest risk is retail demand.


We like SanDisk for the following reasons: 1) The company is well positioned in terms of cost-downs in 2010 with the transition to 32nm and 3bit and will likely see cost-downs of ~40%, better than ASP declines of ~32%. 2) The stronger- than-expected NAND environment in 2H10 is positive for SNDK’s vertical model. We continue to believe that the Street might be underestimating the strength in the OEM channel, which is driven by tablets/e-books and smartphones. 3) We believe that the company continues to gain share with its product variety and pricing flexibility, especially in the private label segment. 4) The mix-change toward OEM should provide longer-term stability and opportunity for new revenue streams. Our new price target of $59 is based on 15x our new 2010E EPS.

**** excerpts from the BoA/ML 15 June 2010 report ****

Business environment more favorable; PO up 18% to $60

Bank of America/ Merrill Lynch

Simon Dong-je Woo, CFA
Merrill Lynch (Seoul)

Shaina Kim
Merrill Lynch (Seoul)

ASP better, lower Samsung threat; PO up 18% to $60

We raise our PO from $51 to $60 (27% upside potential) for three reasons: 1) EPS revisions (+22/13% for 2010/11E); (2) Samsung’s potential constraints in NAND capacity ramp-up (lack of shell fab); and (3) solid chip-pricing environment. Our ASP assumption changes already indicate 33% higher operating profit in 2Q (ASP: -8% QoQ vs -15% previously; OP: $319mn vs $240mn). This suggests upbeat 2Q results to us (OP consensus: $301mn). Our new 2010-12 EPS estimates are far higher than the 2005 peak ($3.5-4.0 vs $2.0), which we believe is not yet reflected in the share price ($47 currently vs $80 in early 2006). Our PO is derived from 16x P/E, 2.3x P/BV and 7.5x EV/EBITDA based on 2010-12E, which is close to the historical average to be conservative, despite the all-time high earnings. We reiterate our Buy

NAND pricing momentum has been upbeat

Quarter-to-date NAND chip and card prices have been far better than seasonality (mid-single-digit drop QoQ in 2Q vs normal: high-teen decline). Solid demand from smartphone and iPad has well absorbed supply growth. Further, Samsung’s logic-chip capacity expansion at the expense of NAND fab clean room should be a positive for SanDisk. Overall, we project better-than-normal NAND-chip-pricing momentum, at least in 2H10-1H11 (ASP falls less than cost).

SanDisk remains one of our global top picks in memory

SanDisk has been one of the best-performing memory stocks YTD (+63% vs Micron -16%, Samsung 0%, Toshiba -8%). Apart from an upbeat NAND price and Samsung’s less aggressive NAND move, we see four potential positive catalysts: 1) technology leadership (better than Samsung in our view); 2) stable royalty revenue (Samsung is the biggest payer); 3) benefit from NAND JV fabs with Toshiba (cost competitiveness for cards); and 4) healthy balance sheet (net cash).

Investment Thesis

SanDisk should be a major beneficiary of the NAND industry recovery. NAND capex spending among chipmakers has been abnormally low in 2009. We expect this to lead to a NAND shortage and new business opportunities for SanDisk, through the leverage of its well established JV fabs and options to use Samsung capacity – the JV fabs were company-specific weaknesses during the NAND downturn. We expect 2010-11 profits to exceed the previous highs of 2006-07, but still remain below the 2005 peak.

Best-performing memory stock

SanDisk has been one of the best-performing memory stocks YTD (+63% vs Micron -16%, Samsung 0%, Toshiba -8%). The key potential positive catalysts as we see them are: 1) technology leadership (better than Samsung in our view); 2) stable royalty revenue (Samsung is the biggest payer); 3) better-than-expected NAND and card-pricing momentum (cost advantage from captive chip production capacity via JV fabs with Toshiba); 4) Samsung’s potential constraints to expand NAND chip production capacity (currently available clean room space among Korea and US fabs will be used for DRAM and logic in our view – see our note on Samsung’s logic capex increase, which will likely hit NAND wafer capacity ramp- ups); and 5) a healthy balance sheet (net cash).

We believe these factors will continue to lead to SanDisk’s share price outperformance as a pure NAND play coupled with solid earnings momentum. We raise our PO from $51 to $60 following +22% and +13% EPS revisions for 2010-11E. Our PO factors in just the historical average of fair multiples based on 2010-12E earnings (16x P/E, 2.3x P/BV and 7.5x EV/EBITDA), despite the all- time high earnings. We reiterate our Buy rating on SanDisk as a global top pick in the memory chip sector.


15 Responses to Allocation

  1. Poofypuppy says:

    Hi Savo,

    Another great post, thanks. There is no question SanDisk is well positioned on a company basis relative to its competitors, although it’s hard not to be concerned about various global economic, market and political risks at this time.

    Saw this article on DRAM prices starting to fall for the first time in a year, which makes me wonder if Samsung will reallocate its capital spending toward NAND instead.


  2. Joncon63 says:

    Great news. I guess we figured it was only a matter of time.

    My question remains: What will Sumit Sadana do to capitalize on the situation? If Apple really uses that high of a percentage of NANd produced, you’d think they’d want to but SNDK.

  3. MorganBucks says:

    JUNE 28, 2010, 5:21 A.M. ET SEMICON REPORT: Apple Helps Sustain Upbeat NAND Flash Market


    HONG KONG (Dow Jones)–The popularity of Apple Inc.’s (AAPL) new iPad and iphone is spurring rival computer and smartphone makers to launch similar products to stay competitive, helping to drive up demand for key components such as memory chips.

    Chip makers and analysts say demand for flash memory chips, used to store data in tablet PCs, mobile handsets and other consumer electronics, remains strong leading up to the busy second half of the year, while supplies remain tight. These favorable market conditions should enable chip prices to remain stable for the next few months, helping to boost earnings at flash memory chip makers including Samsung Electronics Co. (005930.SE), Toshiba Corp. (6502.TO), Micron Technology Inc. (MU) and Hynix Semiconductor Inc. (000660.SE).

    “Given the low level of capital expenditure made by NAND flash makers last year, we foresee some limitation in the increase of overall supply of NAND chips this year,” Keisuke Ohmori, a spokesman at Japan’s Toshiba, the world’s second largest maker of NAND flash memory chips by revenue after Samsung, says.

    Ohmori said supply in the NAND flash market is expected to be tighter in the second half of this year.

    Since its launch in April in the U.S., Apple has sold about 3 million iPads. And Apple’s newest version of its iPhone, the iPhone 4, is also shaping up to be a blockbuster — the company took preorders for more than 600,000 of the devices on its first day of availability, while carrier AT&T Inc. (T) had to stop taking advance orders because of inventory issues and stronger-than-expected demand.

    “The success of the iPhone in the smart-phone category has spurred the launch of a series of competitive mobile phones. Beyond the smart-phone segment, a number of other promising products are driving the NAND market this year including e-books and tablet PCs,” says Michael Yang, an analyst at market research firm iSuppli Corp.

    In recent months, Dell Inc., Samsung, Toshiba and even Chinese battery-maker BYD Co. have announced plans to launch tablet PCs. A slew of mobile handset makers including LG Electronics Inc., Sony Ericsson and Nokia Corp. have also launched new smartphones to compete with Apple’s iPhone 4. NAND flash memory chips are also making headway into electronic readers such as Inc.’s Kindle and Barnes & Noble Inc.’s Nook. South Korean company iRiver Ltd. said recently it is also preparing an e-reader device to sell in the U.S. market later this year.

    Global NAND flash revenue rose to $4.36 billion in the first quarter, up from $4.33 billion in the fourth quarter, according to iSuppli and the researcher predicts that NAND flash revenue will increase 34% to $18.1 billion this year from $13.5 billion last year. Meanwhile, average prices have slid 6.4% in the latest quarter from the fourth, compared with a typical seasonal decline of about 14%. On Monday, the average spot price of a 32-gigabit NAND flash memory chip fetched $7.03 each unchanged from two weeks ago, according to DRAMeXchange, Asia’s biggest spot market for chips.

    Samsung, the world’s biggest supplier of NAND flash memory chips, and an Apple supplier, says it plans to boost production of high-density NAND in the second half of the year to respond to strong customer demand. The firm says the market is currently in shortage and it expects demand to continue to rise as the penetration of NAND chips increases in smartphones and digital camcorders.

    “The iPad, and iPad-like products will consume about 5% to 10% of the global NAND demand in the second half,” says Nam Kim, an analyst at Arete. “We don’t expect the market to be in a shortage as traditional demand drivers such as flash storage card, USB flash drive and MP3 player market have still been weak. However, supply and demand should be in good balance as consumer demand picks up in the holiday season.”

    Table of Average Weekly Spot Prices June 28 June 14 May 20 May 13 64 Gb NAND $14.67 $14.74 $15.09 $15.26 32 Gb NAND $7.03 $7.03 $7.20 $7.33 16 Gb NAND $4.27 $4.20 $3.91 $3.97 8 Gb NAND $3.74 $3.77 $3.79 $3.85 1 Gb DDR3-1333 $2.58 $2.60 $2.84 $2.86 1 Gb DDR2-800 $2.19 $2.29 $2.50 $2.60 1 Gb DDR2-667 $2.20 $2.26 $2.49 $2.56 Source: DRAMeXchange (Yun-Hee Kim, an assistant managing editor at Dow Jones Newswires, has been writing about Asia’s semiconductor industry since 2002. She can be reached at 852-2832-2330 or by e-mail at

  4. Poofypuppy says:

    Saw this interesting article saying the Taiwanese government plans to band several companies together to develop NAND flash technology…

    An interesting (almost funny) quote from the article… “Since Taiwan lacks patents and self-developed NAND flash technology, it is extremely limited in its ability to develop flash products”.

    With node shrinks approaching the end of the road and large, established competitors having superior technology, this seems a bit foolhardy IMO. But perhaps they will or intend to license some technology from SanDisk?

  5. MorganBucks says:

    Some electronic components in critically short supply
    Scott Bicheno – 5 Jul 10, 10:57am
    Delicate balancing act
    Company Info
    All iSuppli related content on HEXUS

    Discuss in the
    Facebook Digg Slashdot Newsvine Delicious Stumble Market researcher iSuppli has published a report revealing that many key electronic components are now in critically short supply. This is greatly extending lead times on orders and inevitably raising prices.

    Among the types of component feeling the pinch are: analogue, standard logic and memory integrated circuits, as well as low-voltage MOSFETs and tantalum capacitors. The report says these are effectively on allocation status and suppliers don’t have the ability to respond to unforecasted demand.

    Lead times for these sorts of components are normally between ten and 12 weeks, but they are now up to 20 weeks for things like power MOSFETs. The reason for this big jump must be a delay between the increase in consumer demand and the increase in manufacturing capacity to accommodate it.

    “When lead times enter the 20 week range, they indicate a major schism between component supply and demand,” said Rick Pierson, senior analyst at iSuppli. “Supply constraints for electronics and semiconductor components might not come as a big surprise amid the present market rebound. However, specific market and pricing trends are spurring varying degrees of short supply depending on the component market.”

    This imbalance is expected to continue for the rest of the year, resulting in potentially even longer lead times and a consequent increase in average selling prices. While the memory market doesn’t seem to be suffering from the same imbalances, iSuppli reckons the NAND flash market could be headed this way shortly.

  6. MorganBucks says:

    NAND flash prices could rise in Q3 on shortages
    NAND flash companies like Samsung and Toshiba could benefit

    Demand for NAND flash could outstrip supply in the third quarter of this year as consumer electronics companies prepare products for the holiday season, leading to price increases and shortages, research firm iSuppli said on Friday.

    “When there is a constrained market, average selling prices do go up,” said Rick Pierson, senior semiconductor analyst at iSuppli. NAND flash memory is commonly used in electronics like digital cameras, smartphones and video cameras for data storage.

    There was an abundant supply of NAND flash in the second quarter which led to prices trending downward, Pierson said. But devices like smartphones are integrating components like video cameras, creating a need for additional storage.

    Factory output won’t necessarily be constrained because of excess demand, but suppliers will allocate NAND flash shipments to key partners based on contractual obligations and relationships.

    The suppliers will also direct NAND flash shipments to markets like video streaming that represent long-term growth opportunities, Pierson said.

    “It’s also a chance for suppliers to take profit,” Pierson said. Companies such as Samsung, Toshiba, Texas Instruments and STMicroelectronics could benefit from the increased selling prices.

    A shortage of components like capacitors, which helps NAND flash work on PCs and other devices, could push the price increases.

    NAND flash supply and demand will reach an equilibrium in the fourth quarter which could stabilize prices, Pierson said. He couldn’t comment on the effect of the increase in NAND flash prices on the end prices of consumer electronics.

    • savolainen says:

      Hi MorganBucks,

      Thanks for the articles and links.

      There certainly are a lot of shifting dynamics in the NAND market recently. Remarkably there has been little speculation in the press about their interrelated significance.

      It sure looks like NAND supply is going to get real tight in second half of 2010. It will be interesting to watch how OEMs move to secure supply in the face of looming shortages.

      Right now it doesn’t look like Samsung has any intention in ruining the party. Micron and Hynix don’t have the wherewithall. And Intel is lying low and appears to be reevaluating its options.

      Samsung appears to be focusing on SoCs for mobile instead of NAND. Good news as Samsung is probably the biggest threat to supply/demand balance.

      As things stand, Samsung appears space constrained on NAND capacity ramp-up- lack of a shell fab beyond Austin, where extra space appears dedicated to SoCs.

      Samsung and Micron have dialed back on X3, leaving that edge for SanDisk to exploit and hone.

      All-in-all about as good as could be hoped for. The general market is another matter.


  7. bob77977 says:

    hi savo,
    lots of attention is given lately to future finance.
    some even speak about a buy out.
    if 3D is what I hope it is, I can imagine intel, sumsung and others fighting over who is going to lend us the money to construct a fub or even more than one fub in return to a 50:50 partnership. loan will be returned by setting off royalties due to us with or without some extra from future profits. that way, we eliminate the risk of oversupply. investing firm takes all the risk. intel could grab such an offer if we promise them sumsung will be out.
    it all depends on what 3D is.

    • savolainen says:

      Hey bob77977,

      I agree that 3D R/W is potentially a huge deal. But it’s not here yet.

      I am very encouraged, especially by SanDisk’s more confident tone recently. But realistically R/W is at least a couple of years off and if viable, production will be part of the fab 5 story. 3D R/W will require its own manufacturing space and it makes no sense to take highly profitable NAND production offline.

      It’s interesting that SanDisk is once again pushing it’s 3D one time programmable (OTP) technology. It would be interesting to know if this is being produced at Yokkaichi (Fab 3 or 4) or TSMC.

      At one time it was coming from TSMC. SNDK had expressed an interest in producing it themselves.

      If it is being produced at Yokkaichi (probably unlikely), this would mean that fabrication equipment would be in place for 3D R/W as well, as the technology is basically the same – as I understand it.

      At investor day, Eli said that 3D R/W was in intensive R&D at Yokkaichi, which I interpret as at least some of the equipment in in place.

      Eli said that 3D R/W is not in pilot production (yet). Pilot production of 3D R/W seems like the next important milestone.

      In a best case scenario, SanDisk/Toshiba would include pilot production of 3D R/W in the upcoming build-out of fab 4- and thus be ahead of the game when fab 5 is up and running. No hints of that though.

      It would be interesting to know what the technology node is for the new 3D OTP chips.

      45 nm 3D OTP chips were expected to arrive about 2008/2009. These would have been fifth generation chips. My guess is that the current 3D OTP product is 45nm with 32nm possibly to follow.

      In any case as you no doubt know, one of the big deals about SanDisk’s 3D is that this cross-point diode array technology is, in theory at least, much more scalable than NAND. Also if it lives up to its promise, endurance issues will not be an issue.


  8. bob77977 says:

    thanks savo,
    no doubt 3D isn’t going to be produced meaningfully in less then 2 years or even more. but isn’t it the proper time to start some strategic planning?
    I think the old way of letting everybody produce & pay royalties is not suitable anymore for sandisk. huge investments are needed that we don’t have, and in no time we will find ourselves a minor player in the market. that’s why I think that on top of royalties, any interested party will have to lend us the money to become 50:50 partner in his fub.
    I am glad you mentioned the endurance aspect of 3D. that means X4 will be working perfectly making it X10 or even X11 and not just X8. I hope 3D will also be quick enough to eliminate need of DRAM.
    indeed, exciting time ahead of us.

  9. Poofypuppy says:

    Hi Savo,

    One question that I have… In terms of managing SanDisk’s own inventory for sale in the next few years, do you have any sense for how much wafer capacity and what type of NAND (SLC? 2-bit? 3-bit?) supply SanDisk can procure from Samsung as part of their L&R renewal? How would much would SanDisk margins be affected by drawing on their reserved supply at Samsung? Obviously, SanDisk gets the lowest cost (and highest margins) by using its own captive supply from its JV with Toshiba. But is it possible that SanDisk can get enough supply (at “semi low” cost) from Samsung that SanDisk may want to take a “go slow” capex funding approach to Fab 5 with Toshiba (assuming that SanDisk doesn’t sell itself to, or enter some kind of major strategic partnership with, Intel first)?


    • savolainen says:

      Hi Poofy,

      As far as I know, no specifics have ever been released on the details of SNDK’s non-captive supply agreements.

      We know that SanDisk’s plan is to have 15 to 20% non-captive and that SNDK has legal agreements that guarantee SNDK access to capacity from Samsung and Hynix. There is also a forecasting period required. In other words lead times. My guess is a few months at least.

      Right now SanDisk isn’t using any of this non-captive supply, but has hinted that some might be necessary towards the end of the year (Q4). Of course non-captive will be much lower margin.

      Also non-captive doesn’t work for OEM or specialized/optimized uses like SSDs.

      SNDK has said that it can grow its bit supply 75% this year, perhaps a little more if inventory turns can be increased. Beyond that non-captive will be required.

      To my mind non-captive has the potential to be a Q4 buffer, and maybe suitable for picking up a few bucks. Maybe with the low capacity private card business if supply is really tight.

      There should be some way SanDisk can profit from non-captive in these tight times.

      I don’t think non-captive really factors into the fab 5 equation.

      I expect we’ll hear a lot more about fab 5 in next week’s cc. It sounds good to me. The best info is in yesterday’s 8k. Lots of flexibility for SNDK.


  10. MorganBucks says:

    July 14, 2010, 6:36 AM ET

    Apple Drives Chip Sales, Too

    The popularity of Apple’s new iPad and iPhone has driven up sales for the Cupertino, Calif.-based company, but it’s not only Apple that’s celebrating.

    Asian component makers — those that make chips and touch-screens used in Apple’s devices and other similiar gadgets — say they are having a tough time meeting orders thanks in part to stronger-than-expected demand for smartphones and tablet PCs.

    In recent months, Dell, Samsung, Toshiba and even Chinese battery-maker BYD have announced plans to launch tablet PCs. A slew of mobile handset makers including LG Electronics, Sony Ericsson and Nokia have also launched new smartphones to compete with Apple’s iPhone 4. These components are also making headway into electronic readers such as’s Kindle and Barnes & Noble’s Nook. The new devices are creating more demand for chips, touch-screens and other types of displays.

    “Due to the recent emergence of new, promising applications, such as tablet PCs, smartphones and iPad, we expect the supply condition in the NAND flash market will be tighter in the second half of this year,” said Toshiba spokesman Keisuke Ohmori.

    NAND flash memory chips are the main components used to store data in music players, cellphones and digital cameras even when power is switched off.

    Samsung and Toshiba, the world’s two biggest makers of NAND flash memory chips by revenue, say they are boosting production now due to strong demand. But it may take a while before supply can keep up with rising demand, analysts say. Chip prices as a result have been on the rise. According to DRAMeXchange, Asia’s biggest spot market for memory chips, the price of a 32-gigabit NAND flash chip fetched $7.04 each late Tuesday, up 0.5% from Monday.

    For Apple and other smartphone and PC makers, rising chip prices means they will have to look at ways to cut other component costs. But that won’t be easy because display prices are also rising due to strong demand. If the component shortages are prolonged into the crucial back-to-school and holiday shopping season, Apple and others could be forced to transfer the burden to consumers by raising prices of their smartphones and PCs.

  11. bob77977 says:

    hi savo,
    a puzzle. 2x production starts about now. fub 5 will start production a year from now and will also use 2x. why not 1x ? anything to do with 3D ?

    • savolainen says:

      Hi bob77977,

      Fab 5 will be starting off with 2x because that will be the workhorse technology for the second half of 2011- and the first half of 2012.

      3D won’t be ready for mass production. As far as we know it isn’t even in pilot production yet.

      Fab 5 won’t be starting with 1x NAND, because it won’t be ready yet. 1x (20nm or a little bit under) is apparently going to take a lot more effort than 2x and that effort takes time and $$.

      NAND shrink is slowing down.

      Slide 54 from last investor day tells the story. 24nm starts in the second half of 2010. 1x nm doesn’t even start until the second half of 2012. I will probably throw this slide into my next post as it is very interesting.

      My guess is that 3D will start to be introduced around the same time as 1x nm (2H 2012) and hopefully will be able to ramp up and start displacing NAND for the next NAND generation- 1y nm (closer to 15nm). 1y will likely be the end of the line for NAND, if it is even feasible or practical.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: