Micron Launches NOR Flash Device
By Zacks Equity Research 3 minutes ago
Semiconductor company Micron Technology Inc. (MU) recently announced the launch of a 45nm MT25Q SPI NOR Flash device. This new MT25Q product offers a more flexible, efficient and cost effective solution, and can be used in industrial applications as well as in consumer, networking and embedded products.
The MT25Q SPI NOR is an exceptionally fast and backward-compatible storage solution. It provides up to 2GB of storage, up to 512MB/sec sequential read and up to 2MB/sec industrial programming speed.
The new product offering is aimed at further innovation in memory technologies which in turn will help the company to strengthen its product portfolio. We believe that the momentum from the launch of MT25Q SPI NOR will help Micron to emerge as a major player in the fast-growing storage market, thus lending greater stability to its revenue stream. Additionally, this will also help Micron to acquire new customers, going forward
Recently, Micron inked a deal with Rambus Inc. (RMBS). Per the terms of the agreement between the two, Micron can use Rambus’ patented technology for its integrated circuit products for an initial period of seven years. The deal also ends all prolonged legal battles between the two technology companies.
Moreover, the other positive factor for Micron is the Elpida acquisition, which has increased wafer manufacturing capacity by approximately 50%, helping it to capitalize on any demand shortage. The acquisition also brought Apple Inc. (AAPL) on to the customer roster, which will be a positive for future growth. The company is also optimistic about supply/demand balance for DRAM and NAND memory chips in 2013 and 2014, which should support prices.
However, it may not be easy for Micron to capture share from SanDisk Corp (SNDK) a key player in the NAND zone. Nevertheless, with support from Apple and Intel Corp., its prime NAND customers, the situation could turnin Micron’s favor going forward
TrendForce: Mobile DRAM Contract Prices Drop during First Half of 1Q14 due to Off Peak Quarter and Sufficient Market Supply
Mobile DRAM contract prices dropped by 5~10% in January due to the channel clients’ inventory congestion and the new smartphone announcements scheduled for March, according to DRAMeXchange, a research division of TrendForce. The prices for the first tier manufacturers’ Discrete memory products experienced an estimated 5~10% decline, whereas those for LPDDR1 suffered a relatively larger reduction given the shift of focus towards LPDDR2 and LPDDR3 and the subsequently weakened LPDDR1 demand.
Looking at MCP&eMCP, as more pricing adjustments were able to be implemented for these products following the weakened NAND Flash price, the overall decline in their contract prices, depending on their specific densities, ranged from 5~20%. Samsung has been noticeably aggressive with the pricing of its TLC eMCP products, and as a result, their price difference with the MLC products is above 5%. The shipment volume of the former product category has begun to show further growth as an increasing number of second and first tier clients seek to lower their manufacturing costs and turn their attention towards mid to low-end smartphone models.
With the eMCP shortage easing in 4Q13 and Micron starting to provide eMCP 4+4 and 4+8 to first tier clients, the market situation for mobile DRAM --in terms of both its supply and demand--has turned out to be relatively healthy. The mobile DRAM market’s supplies, in particular, are anticipated to become increasingly stable as SK Hynix gradually recovers its DRAM production following the September fire accident. Taking the aforementioned developments into account, the Mobile DRAM contract prices are projected to undergo various noticeable downward revisions throughout January.
In 2Q14, the LPDDR3 demand momentum will likely increase rapidly as Qualcomm releases its LPDDR3 compatible chips to tackle the high-end smartphone
December 11, 2013, 4:41 P.M. ET
Micron: 3-D NAND Coming, But Will Take Time, Says Piper; Applied, Lam to Benefit
By Tiernan Ray
Piper Jaffray’s Jagadish Iyer today reiterates Overweight ratings on shares of chip maker Micron Technology (MU) and SanDisk (SNDK), along with Overweight ratings as well on share of equipment makers Applied Materials (AMAT), KLA Tencor (KLAC), and Lam Research (LRCX), after attending discussions yesterday at the International Electron Devices Meeting in Washington, D.C.
The part that interested Iyer was discussion of the forthcoming technology called “3-D NAND,” a new form of flash memory chips .
Both Micron and SanDisk, along with Koreas SK Hynix (000660KS) and Taiwan’s Macronix (2337TW) gave presentations at the show that made clear that 3-D NAND will be the “most logical step/roadmap for non-volatile memory,” but that it will also take a back-seat while vendors continue to extend the viability of their current products.
Costs are much higher to make 3-D versus 2-D NAND, for the moment limiting what vendors will be willing to make:
Given the ongoing challenges with the architecture change from 2D NAND (floating gate) to 3D NAND (charge trap flash) it is clear SNDK wants to push 2D NAND beyond 1Z (ramp in 2H14) leveraging its fab capability for lowest cost /GB while Micron will complete 16nm next year before migrating to 256 Gb 3D NAND. Hynix indicated it will move to 3D NAND after its ramp of 15nm next year suggesting that a majority of 3D NAND ramp besides Samsung could happen in 2H15 in some volumes while 2014 could be a learning year. With capex/wafer ~4X compared to 16nm planar NAND, memory makers will likely deploy in high volume only on attractive cost/GB metric. It was also evident that not all end markets will need 3D NAND and it appears that both 2D and 3D NAND will coexist for the next few years.
Hence, efforts are limited at this point, with Hynix “doing some pilot level activity on its first generation 3D NAND device in its factory in Korea,” he writes, and SanDisk perhaps allocating a “portion” of its manufacturing capacity at Toshiba for the technology, before a full ramp some time in 2015.
Micron will ramp its last run of ordinary NAND in a 16-nanometer process technology in 2014, and will sample 3-D NAND “early next year but sees large scale deployment well beyond next year.”
But early manufacturing trials by Samsung Electronics (005930KS) could be of benefit to Applied and Lam in terms of equipment sales next year:
Given that Samsung will ramp its 3D NAND next year and other NAND makers are accelerating their learning on 3D NAND, we see AMAT and LRCX benefiting in deposition and etch tools. With thicker films and slower productivity, particularly in etch, we see rising capital intensity there as Litho spending eases. With abundant opportunities available for defect monitoring in these high aspect ratio structures, we see KLAC benefiting as the need for brightfield tools will continue to rise. Additionally, the increasing deployment of Optical CD for measuring these stair case structures benefits KLA. Micron’s illustration of ~4x capex per wafer for 3D NAND versus planar suggests that there is significant equipment opportunity ahead for AMAT, LRCX and KLAC as this NAND industry inflection unfolds over the next 12-24 months.
Iyer offers up a couple slides from Applied about the challenges to semiconductor “etch” and “deposition” phases of lithography when moving to 3-D NAND:
Jimmy Carter was so very happy when the Obum came along because that boosted Carter's status to only the second worst president in history.
Sentiment: Strong Buy
Just sold half position net profits a little over a quarter million in just 3 weeks, thanks to flash market research (google them). They were just mentioned on CNBC. Good luck all!
Cramer is a buffoon. Remember his recommendation to buy Bear Stearns in 2008 a few days before it collapsed and the stock lost 92% of it's value?!
The Micron story is about Elpida and industry consolidation (not to mention secular demand trends such as SSD growth).
Sentiment: Strong Buy
The unconfirmed construction of the new Hynix fab hasn't really had the effect that could reasonably be expected from the reported event. It would appear that such news should have a broader or less focussed implications; however, MU experiencing 100 million in daily volume certainly appeared to be the chief recipient of reactionary trading.
Why would Hynix trade down today? The implications woudl seem to be that Bloomberg authors have more insight into the memory industry and future pricing than firms actually working in it, otherwise why build a fab who's production revenue would be offset by a decline memory pricing anyway?
In any case, whether the report was genuine and unmotivated or not, as many have mentioned, how a potential 2015 (at best) production increase can have such an influence over the market of today is beyond my comprehension.
The substance appears to be thin, the reaction appears to be excessive, and without any confirmed or substantial change in MU's fundamentals, this should reveal itself as a superficial and short term dip.
Hopefully, this will prove to be the major shake out before MU continues its progression towards its true value.
Maybe some sellers should have read beyond the headline. First they are only "considering building a new fab because its current facility is aging." So it isn't new capacity at all.
Now this as yet unbuilt plant, that hasn't even been decided to be built, will not be finished until 2H2015 at the earliest (you can read that as Dec 31, 2015!). I guess completed means the walls and roof are up since no one can get a new brick and mortar plant expansion up-- and producing wafers for volume production-- in 2 years.
So some time in 2016/2017 a new Fab to replace an old fab will come on line..maybe. This is a story? I think not.
The "News" was: "Potential new DRAM fab modestly negative for Micron, says Susquehanna
After news reports from Korea suggested that Micron competitor Hynix plans to build a new DRAM fab, Susquehanna reports that Hynix is considering building a new fab because its current facility is aging. The firm doesn't expect the fab to be completed until 2H15 at the earliest, and it keeps a Neutral rating on Micron."
Sentiment: Strong Buy
Insye website shows DRAM Spot Price rising through now in early January. (Bears were wrong since October as price is rising not even stabilizing). Few in October foresaw price still rising in early January. The momentum is vendors are accustomed to paying a higher rising price (not a flat same price or typical learning curve changing price).
As Moore's Law learning curve production costs improve with every bit shipped, in a climate of prices trending above the 30% cost reduction curve results in UPSIDE to marginal profit.
Question: Has anyone seen any forecasts on MU die size shrink plan, wafer additions, or yield improvements? If there is any step-function production cost improvement immediately it will fall to the bottom line.
Too bad Wall Street analysts haven't worked as a semiconductor production line planner or financial cost accountant for an SC front end fab. The cost side of the equation is outside of most generalist stock commentators DD scope, as they lack semiconductor knowledge or information to model it.
Print Twitter Facebook Google+ LinkedIn
January 22, 2014 2:08 am JST
Japan's Elpida resuming capital spending in bid to catch Samsung
Elpida's Hiroshima foundry will produce 20nm DRAM chips.
TOKYO -- Elpida Memory will invest roughly 80 billion yen ($758 million) to expand chip production, adding cutting-edge fabrication equipment with an eye toward catching up to DRAM market leader Samsung Electronics.
This would mark the chipmaker's first capital investment in three years. The company is rebuilding its business with help from U.S. firm Micron Technology.
With the fresh investment, its Hiroshima plant will begin mass-producing DRAM chips with 20nm circuit widths in the latter half of this year. Currently, the most advanced mass-produced chips measure 25nm, while Samsung has the technology to make circuit widths in the lower end of the 20nm range. Reducing the size to 20nm will enable 20% more chips to be cut out from each wafer.
Designed for use in smartphones, Elpida's new chips are expected to consume 10% less electricity, a boon to device makers concerned about extending battery life.
Elpida spent 70 billion yen on equipment in the year that ended in March 2012, but an economic slump and poor sales of personal computer products hurt its business. It filed for bankruptcy protection under the Corporate Rehabilitation Law and became a wholly owned subsidiary of Micron last July. The move doubled Micron's share of the global DRAM market to 28%, putting it on a par with second-place SK Hynix.
To challenge Samsung, which holds 37% of the market, Micron is boosting its capital spending for the year ending in August to as high as $3.1 billion, double last fiscal year's figure. About half of this sum will be invested in Elpida plants in Japan and Taiwan. The Hiroshima site, which boasts advanced microprocessing technology, will be positioned as a supply base for cutting-edge products.
Smartphones and tablets are bolstering the DRAM market. Prices doubled over the course of 2013 and have remained high. The handsets are expected to become widespread in Asia, further tightening the DRAM supply-demand balance and fueling competition among manufacturers.
There were a number of posters on this board who did not believe those of us who spent a lot of time in fabs telling them back in Sept this was going to be a big problem. I don't envy the team that has to solve this yield issue. It could be a fair amount of time before quality and reliability are back to customers' AQL specs. The other problem with poor die yields is it overworks the back end of the line as you have to test many more parts to get good ones, so it just gums up the entire flow, particularly with DRAMs where if your laser redundancy fixes go up high-enough, customers may demand burn-in before taking parts, which is a killer for through put. If customers are actually rejecting shipments in any quantities, then other customers will push their screening levels up, kind of a vicious circle. They may have to dump the parts into the China white box market at a hefty discount to move them. The worst kind of problem is the one which has the DRAM age very quickly and go out of spec after they've been in the product for a while. Killer for the OEMs. If that happens, expect Wuxi to have real issues for a long-time.
Seeking Alpha is just a collection of individuals, and the content is only as good as the person who wrote it. That particular article is junk. The writer claims Micron is shifting NAND capactity to DRAM, when in fact the EXACT OPPOSITE is true. MIcron has communicated this very clearly, most recently a few days ago in London.
The fact that Micron is reducing DRAM capacity just as Hynix is coming back in line is just a happy coincidence hjwever. The real story regarding Hynix is that none of the other suppliers (Micron, Samsung, SanDisk, Toshiba) increased capacity to take advantage of the fire (as might have been expected a few years ago), thus there is no glut as they come back on line. The supply/demand situation for 2014 remains very good,
Sentiment: Strong Buy
First, any of the very few analysts negatively commenting about Dram supply and pricing should be fired for faulty analysis. Dram spot prices are currently rising and almost at an all time yearly high, just look at any chart yourself. Dram supply will minimally if at all be impacted by any Hynix upgrade or construction, read Hynix's recent response yourself where they say that their supply will not be increased, only transferred to an upgraded plant.
Second, there are a few wild cards that may surprise analysts to the upside in the upcoming earnings releases next month and beyond. First, there will be gains recorded regarding YEN conversion rates and they could add some serious pennies to the EPS, not factored in by analysts.
Third, the CGS and OPEX will continue declining due to ELPIDA cost write downs when acquired for both plant and equipment, as well as MU's continued focus on cost reduction and economies of scale due to the consolidation of Elpida. Bottom line improves more than anticipated and free cash flow trend is on an uptick. EPS increases more than currently anticipated.
Forth, cash flow improvement due to all the above plus MU having to pay less for Elpida purchase due to favorable yen conversion rates that may continue into the future due to Abenomics. This will reduce bank debt faster and thus interest expense, improving Balance sheet metrics and EPS due to less interest expense.
Fifth, revenue increase due to Apple/ China deal just announced, plus new technologies that MU recently announced to enter new revenue streams. Hynix customers switching to MU due to Hynix fire, may be either short term uptick or longer term if customers like MU's products and switch for good. PC revenue decline may be stabilizing according to Intel. Can only help.
All the above point to tailwinds for MU, and as such any talk of structural declines in MU's stock price are totally misleading and not supportive. Stock price higher from here, not lower.
I just read an article that truly applies to all on this board who feel they are smart investors, and those who seem to comment on every short term blip in MU's stock movement. I wish to cut and paste and hope this has value to those on the board .
" An investor should think like a business owner, not a renter. Most business people don't get up in the morning and ask whether they should sell their business that day. If they own a pizza shop, they don't think about whether what they really should own is a shoe store instead. They show patience and persistence and try to understand their underlying business ( i.e. MU ) better so they can earn the greatest return for the longest period of time.
So investors are in many ways misled by stock volatility ( i.e. MU, especially this last week or so ). The values of the underlying businesses just don't change as quickly as stock prices do. You really don't have to watch those changes hawklike day after day."
I value these words of wisdom. It doesn't mean to be married to MU, but be patient if the underlying aspects of MU are compelling to hold it.
Hope this clarifies or adds conviction to those who own MU, long, short or in between.
May 2014 be better for all, no matter what your investment choice.
Things don't go in straight lines and we're less than 10% off the highest point it's been in many years. Look at the chart in the past year, it dropped from around$10 to $9, then $14.50 to $12.50, then from $15 to $13, then from $18.50 to $16.50, then from $23 to $20.50...this is just a normal tiny dip, so far hardly even worth noticing.
Sentiment: Strong Buy