Tuesday, December 2, 2008

Fujitsu Next

Intel Corporation and Hitachi Global Storage Technologies announced that they will jointly develop Serial Attached SCSI (SAS) and Fibre Channel (FC) enterprise class solid-state drives (SSDs) for servers, workstations and storage systems. This is an exclusive agreement to develop and deliver SAS and FC enterprise SSDs with availability of the first products in early 2010.

This is a complementary partnership marrying Hitachi GST’s drive, channel and system knowledge together with Intel’s NAND flash and flash management expertise. Hitachi selected a strong NAND flash partner that doesn’t compete with it in its core HDD business. Now that two of the three enterprise HDD makers, Seagate and Hitachi, have publicly announced plans for enterprise SSDs, Fujitsu will be compelled to react, especially with its 23% enterprise HDD market share to protect.

Seagate may feel it is big enough to source the NAND flash and develop the SSDs on its own however, Fujitsu may need to develop a partnership with a NAND flash vendor. Two possible candidates: Micron and Hynix - NAND flash vendors with SSD ambitions but no competing HDD businesses. Micron has flash management expertise acquired through its own SSD development and leading cost structure based on 34nm technology. On the other hand, Hynix is lagging on both the SSD and NAND flash technology front.

If rumors of Western Digital acquiring Fujitsu's HDD business are correct, a company that has sat on the sidelines as companies have rushed into the SSD market will have to finally step up to the plate.

Tuesday, November 18, 2008

Spansion Becoming an IP Company

Struggling to achieve profitability in its core business, Spansion has now turned to monetizing its portfolio of 3,000 patent and patent applications. Its first victim: Samsung. Spansion has filed two separate patent infringement claims against the company with the International Trade Commission and in the U.S. District Court in Delaware. Spansion is seeking an injunction in both cases on the sales of electronic devices containing Samsung flash memory and requesting treble damages in the Delaware claim for knowingly violating the patents.

Spansion is claiming Samsung infringed on 10 patents with four of them filed in the ITC complaint and six in the Delaware claim. The patents in question are related to NAND flash memory and include process patents describing methods of forming of the interpoly dielectric, shallow trench isolation as well as programming schemes. The patents, granted between 1998 and 2002, were filed by AMD, the pedigree of Spansion. AMD developed NAND technology and introduced a 64Mb UltraNAND on 0.25um technology in 1998. UltraNAND was supposed to do one better than NAND - 100% good blocks and 100k program/erase cycles without ECC - but it never took off and was subsequently shelved.

Why didn't Spansion instead go after Samsung, the No. 3 NOR vendor, with its much larger portfolio of floating gate NOR flash patents? Any successful claims against Samsung are likely to hurt more on the NAND side because of the size of Samsung's NAND business and the more diversified customer base. In addition, it's harder for Samsung to strike back because Spansion has no products based on floating gate NAND.

Spansion also claims to have the strongest portfolio of patents related to charge-trapping (CTF) technology which may be required to extend NAND flash below 30nm. However, it's not clear how applicable Mirrorbit patents are to SONOS NAND because the programming/erase mechanisms and gate stack are different. Spansion may be trying to plug that hole by commericalizing the first SONOS NAND - ORNAND2. In addition to satisfying its own demand for a NAND device in its MCPs, ORNAND2 is also a learning vehicle for developing CTF NAND IP. It'll be interesting to see what innovations Spansion introduces when ORNAND2 starts ramping in production the 2nd half of 2009 on 43nm.

Tuesday, October 14, 2008

Step 1?: Micron Closes Inotera Deal

Micron and Qimonda finally closed a deal to transfer Qimonda's 35.6% stake in Inotera for $400 million. The move was widely expected given the termination of Nanya and Qimonda's joint DRAM development activities and Nanya's joint DRAM development and manufacturing ventures with Micron.

The acquisition of Inotera will allow Micron to merge its manufacturing JV with Nanya, Meiya, into Inotera - akin to a backdoor listing - and allow Meiya to access the capital markets immediately. The original plan was for Meiya to go IPO three years after the start of operations and obviously, there's no need for that anymore.

Micron paid only $400 million for 50% of the capacity of Inotera or 60k wpm. A 60k wpm fab would've cost them $2.7 billion to build - quite a bargain. However, Inotera will have to install copper tooling for the 68nm process plus convert from trench to stack process. This will require about $800 million although this won't necessarily come out of Micron's pocket.

As for Qimonda, it's a new lease on life. They would've burned through their cash by the end of the year wthout the cash injection from the deal. They're closing Richmond 200 and getting out of consumer and mobile DRAMs to focus on areas where they're strong - infrastructure and graphics. Whether that is enough for them to survive as a standalone company is questionable given the fact that Inotera accounted for roughly 1/3 of QI's internal wafer capacity and the importance of economies of scale in memory manufacturing. They will at least need to re-engage with Elpida in the joint development activities to reduce the R&D burden. Or maybe the Inotera deal was just step 1 in a phased acquisition of Qimonda and step 2 and 3 are in the offing.

Friday, September 5, 2008

Why does Samsung need SanDisk?

There’s been a lot of speculation and analysis on why Samsung would acquire SanDisk. Here’s my take.

1. Samsung needs the capacity. – Samsung has enough capacity, especially in this market.

2. Samsung is attracted by SanDisk’s retail reach. – Samsung has a strong retail brand and could have leveraged its own retail channels to push flash cards through its distribution network if it wanted to. Why do it now?

3. Samsung saves on royalty fees and the savings justify an acquisition. – Pay $3 billion (SNDK’s market cap) to save on $400 million a year? Sounds pretty good. >10% ROI. But the level of that royalty stream is not guaranteed and I don’t think Samsung would do a deal just to save on royalty payments.

4. Samsung wants the IP. – Sometime in the next few years, FG NAND flash is likely to hit a brick wall and an alternative technology will be required. CTF has its problems and one of the potential options is 3D memory (see http://forward-insights.blogspot.com/2008/06/sandisk-and-toshiba-to-jointly-develop.html). SanDisk appears to have some fundamental IP in that area. Sure, Samsung could negotiate an extension to their existing cross licensing agreement with SanDisk to include 3D memory but they don’t get the knowhow. The knowhow is what is going to allow the leaders to scale the brick wall.

Samsung is not known for big acquisitions and its last big acquisition of AST Research 20 years ago is not considered a success. And it’s hard to imagine SanDisk’s manufacturing and joint development partner, Toshiba standing idly by as Samsung takes ownership of SanDisk’s most prized assets.

Saturday, August 16, 2008

Seagate to Purchase SanDisk? Definitely, Maybe Not

Rumors are flying about Seagate buying SanDisk or Intel's stake in Intel-Micron Flash Technologies (http://www.eetimes.com/news/latest/showArticle.jhtml?articleID=210004140). Does it make sense? Seagate does need access to low cost NAND flash to fuel its SSD ambitions. Here are some ways to think about it.

1. If Seagate were to purchase SanDisk or Intel's portion of IMFT, what are they going to do with all that capacity? SSDs are not going to be able to eat up all that capacity, at least not for the next couple of years. Does Seagate want to manufacture flash memory cards or supply to MP3 players? SanDisk's operating margin in 2007 was 8% and Seagate's 6.8% in FY08. However, if you strip out income from royalties and licensing, SanDisk's operating margins were -5%. Combining a primarily OEM HDD business with a low margin retail business could be challenging.

2. Seagate is primarily interested in enterprise SSDs which uses exclusively SLC NAND flash technology. Any SanDisk acquisition provides Segate with 150k 300mm wafers per month of MLC NAND technology. Spending $4 billion (market cap of SanDisk) to obtain MLC NAND flash technology for enterprise SSDs? (SanDisk and partner Toshiba currently do not manufacture any high density SLC NAND parts, although that could change in the near future.)

3. Seagate spends almost $1 billion in capex ($930m in FY08) p.a. in their HDD business. SanDisk in 2007 spent roughly $1.6 billion and Intel $1 billion in NAND flash. Looks like Seagate's capex would have to at least double if it wants to become a NAND flash manufacturer and it's balance sheet would be strained supporting such a high level of investment.

4. Intel is not ready to exit NAND flash, at least not in the near term. IMFT has typically been behind the technology leaders by 1-2 process generations. With the 34nm announcement, they are about to leap ahead. With the technology lead, IMFT will presumably have the lowest costs. In addition, Intel will be releasing a slew of competitive SSD offerings in the next months. If, with the lowest costs in the industry and a very strong product lineup, Intel still can't make money, then it will consider whether it makes sense to be in this business, but they're not going to quit before giving it their best shot.

Seagate needs access to low cost NAND flash and they don't need to obtain that access by getting into the retail flash card business or doubling their annual capex. They could achieve the same means by either investing a small equity stake in a NAND flash vendor or placing an upfront payment to secure NAND flash capacity at preferential pricing. Apple did something similar when it introduced its first flash-based iPOD. Such an arrangement also affords Seagate sourcing flexibility especially if those wild hockey stick projections of SSD shipments don't pan out.

Wednesday, July 23, 2008

Finally! A Simple Metric for Solid State Drive Endurance

We've all seen the impressive MTBF (mean time between failures) specifications of SSDs versus HDDs of one to two million hours versus 600 thousand hours in a notebook pc class application. However, unlike HDDs where a mechanical failure can make the entire HDD unusable, the main failure mechanism in a SSD relates to the memory cells becoming unusable. MTBF is a statistical calculation that unfortunately, does not capture the effect of write endurance.

The write endurance of a SSD is a function of the type of flash memory used (SLC, MLC), storage capacity, frequency of writes, data sizes of the writes and the amount of static data which relates to the wear leveling algorithm. System level endurance of SSDs in the industrial, enterprise and miliary space have ranged anywhere from one million to five million program/erase cycles. There are three main problems with this.

1. The endurance will depend on how the SSD is used in the application. As a result, SSD vendors can tweak the parameters for their own marketing purposes.

2. Not all SSD vendors use the same formula for calculating the endurance and the formulas can be overly complex.

3. A straight endurance metric is nebulous and difficult to grasp its implications, especially for OEMs who've never had to grapple with endurance issues in HDDs. For example: What's the effect on the lifetime of the drive?

SanDisk aims to solve this with the Longterm Data Endurance (LDE) metric. LDE is simply defined as the total amount of data writes allowed in the lifespan of the SSD. The metric is based on the Bapco write usage pattern for a typical business user and assumes the data is written equally over the lifetime of the drive and that data is retained for one year once the LDE specification is reached.

LDE allows OEMs a simple way to compare SSDs and determine, based on the applications usage patterns which drives are suitable for a particular application. For example, a drive with an 80TBW (teraByte write) LDE can support 20GB writes per day for 10 years (equivalent to 73TBW). For an application requiring support for only half the number of writes per day (10GB), a 40TBW rated drive would be sufficient.

The beauty of LDE is that it captures endurance in one single, understandable figure. A common metric is necessary to facilitate SSD adoption moving forward. Now comes the hard part: garnering support from other SSD vendors and OEMs.

I've uploaded a copy of Don Barnetson's presentation on LDE, "Solid State Drives: The MLC Challenge" on my website at http://www.forward-insights.com/.

Thursday, July 17, 2008

Samsung Introduces Server-Grade NAND flash for SSDs

Samsung Electronics announced today that it collaborated with Sun Microsystems to develop a SLC NAND flash memory with five times the endurance of conventional NAND flash memory. This would put the "ultra-endurance server-grade" flash memory at 500k program/erase cycles.

To extend the endurance, Samsung likely tweaked the underlying cell process and/or modified the programming algorithm. The net result is that the improved endurance comes at the expense of reduced performance or retention. However, by targeting high transactional enterprise applications with this device, the degraded retention should not be an issue.

Tuesday, June 17, 2008

SanDisk and Toshiba to Jointly Develop 3D Memory

In a SEC filing today, SanDisk disclosed that it had signed a collaboration agreement with Toshiba for re-writeable 3D memory. Both companies will cross-license related IP to each other and SanDisk will receive licensing payments from Toshiba.

Investment for SanDisk-owned equipment which includes tooling for 3D R/W and 3D OTP memory is forecast to amount to $400 million in 2009 and $200 million in 2010.

The 3D R/W memory which consists of stacked vertical diode arrays shares most of the process modules and design architecture concepts with the 3D OTP memory. In the case of 3D OTP memory, four layers of stacked memory cells are in volume production at the 80nm node with 45nm currently under development.

A four level 3D R/W memory will have to at least catch up with NAND flash on process technology to be considered competitive with x4 NAND flash. This would put 3D R/W memory at least 3-4 years out. The other issue is whether an eight level stack is manufacturable at high yields. An eight-level stack was demonstrated by Matrix in 2003 on a much less advanced 0.25um geometry. Producing eight level memory stacks at leading edge technology is another matter.

Tuesday, June 3, 2008

Hynix Develops x3

Hynix announced it has developed a 32Gb 3-bit per cell NAND flash slated for production in October. Hynix stated that the 3-bit per cell technology would enable cost reductions of 30% over 2-bit per cell technology. The 32Gb chip is based on 48nm process technology.

Based on my estimates, the die size of this device should be over 200mm2, much larger than the 172mm2 of the recently announced 34nm 32Gb MLC NAND flash from Intel/Micron. It is unlikely the chip is using the All-Bitline architecture developed by SanDisk and Toshiba implying a fairly low program performance.

The 32Gb chip will probably be employed as a learning vehicle for applications enablement until a more competitive 41nm offering comes out next year.

Thursday, May 29, 2008

Intel and Micron Leapfrog the Competition

Intel and Micron announced today that it will be sampling a 34nm 32Gb MLC NAND flash to customers in June with production slated for the second half of 2008. Owing to the aggressive gate half-pitch, immersion lithography with self-aligned double patterning employing spacers is most likely being used. Also expect changes in the bitline and wordline materials as well as a higher k interpoly dielectric in comparison to the 50nm generation.

At 172mm2, Intel-Micron's 32Gb product will be the only 32Gb monolithic MLC device capable of fitting in a TSOP package. If the ramp of IMFT's 50nm 16Gb device is any guide, we should expect to see volume in December or in early Q1/09. It's quite remarkable that Intel-Micron have managed to catch up and surpass the other NAND flash vendors on process technology in the short span of three years. IMFT achieved this milestone by skipping the 6xnm and 4xnm nodes. However, any cost advantage could be short-lived if IMFT fails to ramp up the technology smoothly and SanDisk/Toshiba ramps its 43nm 32Gb x3 in Q1/09 as planned.

Monday, April 21, 2008

Macronix: After the Qimonda Divorce

Just 3.5 months after signing the agreement, Macronix International and Qimonda AG announced that Qimonda terminated the Flash technology license and joint development agreement. Having lost Euro 482 million in the most recent quarter as well as its technology partner, Nanya Technologies to Micron Technology, Qimonda is in survival mode. By cutting back its flash technology development, Qimonda intends to re-deploy resources to focus on ensuring the success of its new buried wordline DRAM technology.

Macronix, on the other hand, will need to either develop flash technology internally or with another partner. With only $487 million cash on its balance sheet at the end of 2007, Macronix will not be able to fund technology development as well as invest in a leading edge 300mm fab. It will clearly require partners.

However, the number of potential partners is limited as all NAND flash vendors with the exception of Samsung are already aligned with others. An earlier agreement to collaborate with Powerchip Semiconductor on flash memory development and foundry services at fab 12M ended due to a fight between the two companies over control of the Macronix board at last year's shareholders' meeting.

One long-shot possibility is Spansion which is rumored to be developing a SONOS-based memory in a NAND architecture dubbed "ORNAND2". However, the ORNAND2 cell is based on Mirrorbit technology and is therefore different than the BE-SONOS technology championed by Macronix which is based on electron tunneling for programming. The fact that ORNAND2 is based on Mirrorbit technology means it is not really a true NAND. Other than the diverging technology strategies, lingering negative sentiment over Spansion's 2006 trademark infringement lawsuit against Macronix makes any collaboration unlikely.

Despite its collaboration with Numonyx, Hynix may be the most attractive partner. Hynix and Numonyx jointly develop product designs, however, technology development rests mainly with Hynix. Normally a technology follower, Hynix will no longer to be able to just copy the technology of others if it is to become a technology leader. BE-SONOS offers a viable scaling path for sub-40nm NAND and a combination of Macronix's strong development team and Hynix's low cost 300mm wafer manufacturing could make a potent team.

Monday, April 14, 2008

Seagate Fires First Volley Against SSD maker

Hard drive maker Seagate a lawsuit against solid state drive manufacturer STEC in the Northern District of California claiming STEC infringed four of Seagate’s patents related to how a SSD interfaces with computers. This is the first time a HDD manufacturer has sued a SSD maker.

It’s hard to see the financial motivation behind such a move. STEC’s enterprise SSD revenues were only $11 million in 2007, although it could reach 4-10x more this year depending on the ramp of the ZeusIOPS and Mach8 MLC product lines. With a market share of over 50% in enterprise HDDs, Seagate clearly sees a longer-term threat from the leading maker of enterprise SSDs.

However, a bigger motivation would be to send a signal to flash memory makers about the value of its intellectual property. It’s no secret that Seagate has been courting NAND flash vendors to secure NAND flash for a SSD it’s planning for the end of the year. A rumored JV with Micron fell apart last year and it is apparently in discussions with two of the three largest vendors. One of the main stumbling blocks is what Seagate could bring to any cooperation as all NAND flash vendors have ambitions to develop and market SSDs. This litigation could be a validation of the IP which Seagate can offer. Western Digital, which along with Seagate, was an early investor in SanDisk (then SunDisk) apparently holds some critical controller and wear leveling IP and could be next to enforce its IP.

SanDisk is taking no chances. In Q2/07, it set up Solid State Storage Solutions LLC with unknown partners that will license IP, presumably SSD-related IP. In July 2007, Solid State Storage Solutions LLC invested $42.5 million for the acquisition of intellectual property. It has apparently purchased relevant SSD and controller IP from Renesas Technology. This IP could prove effective in extracting royalties from flash memory card manufacturers and controller makers entering the SSD space, but it is questionable whether it is enough to counter any future claims by Seagate and WD.

Tuesday, April 1, 2008

Hynix Signs Another Deal

Hynix signed another deal, this time with Grandis to jointly develop spin-torque RAM. Unlike the current MRAM products marketed by Freescale, STT-RAM uses the angular momentum derived from the spin of the electrical current to alter the magnetic orientation of a free layer. By having the current directly pass through the magnetic tunnel unction, the fields required to switch a bit can be scaled down as process geometries shrink.

This is the third in a string of technology deals starting last October with Ovonyx on phase change memory and with Nanosys on nanocrystal memory. Hynix has traditionally been a technology follower, mimicking Samsung’s advances on the technology front. However, as conventional DRAM and flash memories encounter scaling challenges alternative approaches may be required. STT-RAM is one of them.

Monday, March 31, 2008

Numonyx Born #1

Intel and STMicroelectronics today announced the official launch of their JV, Numonyx. Numonyx combines the NOR flash assets of Intel and NAND and NOR assets of ST under one roof.

Numonyx starts out its first day as the number one NOR flash vendor, overtaking Spansion. Now, like Spansion and other NOR flash vendors, it must find a way to make money. It’s not going to be easy with the economy in bad shape and NOR flash pricing affected by falling NAND flash prices. The new company, does however, have a few things going for it.

1. By merging two similar businesses, there’s a lot of low hanging fruit and
opportunities to cut costs.

2. Numonyx will have dedicated facilities for production unlike in the past when it had to compete for resources with chipset or logic products.

3. Numonyx is 1.5 years and 2 years ahead of Spansion and Samsung on the technology roadmap.

4. At Intel and ST, flash memory was a side business. At Numonyx, flash memory is a core business.

This last point is what makes Numonyx a formidable competitor. It must succeed just to survive.

Saturday, March 22, 2008

Hynix Connects the Dots

On March 20, Nanosys, Inc. announced that Hynix Semiconductor Inc. will collaborate with Nanosys to employ Nanosys' quantum dot flash memory technologies (QDM) for NAND based flash memory. Hynix is the third major semiconductor manufacturer to sign up for access to Nanosys’ technology. Intel invested $38 million in Nanosys’ second round financing that closed in May 2003 and subsequently, announced a technical collaboration in 2004 to explore the use of nanocrystals in memory devices. In 2006, the collaboration was expanded to include Intel’s NAND flash memory partner, Micron Technology.

NAND-based nanocrystal memories can potentially provide a scaling path for current floating gate (FG) technologies by reducing the capacitive coupling between cells and eliminating SILC. The main challenge has been the controllability and uniformity of the nanocrystal size and distribution with technology scaling.
Nanosys claims to have solved this issue. Nanosys’ QDM with proprietary ligands are applied with a traditional spin-on process and self-assemble into a consistent monolayer on the wafer. The integration involves fewer process steps than either FG or nitride charge trap flash (CTF) alternatives while offering a very large voltage threshold window for enhanced MLC capability.

If QDM can deliver on its promise, we could see QDM enter into production at the 3xnm node and beyond.

Friday, March 7, 2008

Micron-Nanya Deal Re-shapes DRAM Landscape

The signing of a memorandum of understanding between Micron Technology and Nanya Technology to jointly develop and manufacture sub-50nm DRAM technology re-shapes the DRAM landscape and places Nanya partner, Qimonda AG in a precarious situation. Although specifics of the final agreement are sketchy, one of the options is to form a 50/50 manufacturing JV in Nanya’s Fab3 which is only equipped to half of its full capacity. Nanya would license Micron’s 68nm technology platform and possibly participate in the joint development of 50nm and below technologies.

The following is our assessment of the impact on the various players.

Micron Technology

The big winner in this is Micron which reduces its capital intensity and increases its access to low cost Asian manufacturing capacity. In the last few years, Micron’s DRAM market share has dropped as it diversified its product portfolio into NAND flash and CMOS image sensors. As the 5th largest DRAM vendor in 2007 with roughly 10% market share, Micron is well poised to vie with Qimonda and Elpida for the No. 3 position.

Nanya Technology

It appears that Nanya, having completed joint development of 58nm trench technology with Qimonda harbored doubts about the scalability of the technology. Its concerns were validated last week when Qimonda announced the development of its “Buried Wordline” stack technology and a roadmap to continue scaling its DRAM technology based on this technology.
By cooperating with Micron, Nanya gains a clear technology roadmap for its business and more importantly, the relevant intellectual property and trade secrets required to manufacture stack technology.

The downside is that it’ll have to continue supporting two technology platforms in the next 2-3 years and absorb any manufacturing inefficiencies and additional fab investment required to make the conversion.

Qimonda AG

Qimonda loses a strong partner in Nanya and indirectly, the Formosa Plastics Group, and except for Samsung, is now the odd man out in relation to the partnerships of Hynix/ProMOS, Elpida/Powerchip and Micron/Nanya. In three years time, Qimonda may find itself in the position that Micron is in today: No. 5 and unprofitable – unless it is able to execute on its Buried Wordline technology.

A bigger issue is the impact on Inotera and the capacity implications for Qimonda.


Inotera will have to eventually transition to stack technology whether it be from Micron or Qimonda, however, it is unlikely that Nanya will continue to work with Qimonda on the future technologies once the Micron deal is finalized. Qimonda has neither the cash nor resources to buyout and manage Inotera.

The likely outcome is for Qimonda to exit Inotera either by selling its shares to Nanya or institutional investors or to sell the shares on the open market. If it sells the shares via the open market, it may take a year to unwind its stake as Taiwan Stock Exchange regulations limit the amount of shares which can be disposed daily.

Qimonda which obtains 50% of the 120k wpm capacity from Inotera will lose out on 60k wpm which will not be easily replaced.

Perhaps a harbinger of the future of Inotera and the intense negotiations going on between Nanya and Qimonda, the domain name
www.inotera.com has expired.


Infineon exited ProMOS after a nasty spat with partner Mosel Vitelic. ProMOS has been having difficulties in obtaining 66nm technology from Hynix due to Korean government restrictions on the export of advanced sub-80nm technologies. Is re-marriage in the cards for the former Infineon/Qimonda partner? Difficult to say, but don’t expect Hynix to give up on ProMOS without a fight.

Winbond, SMIC

Expect Qimonda to shore up its alliances with Winbond and SMIC. It needs the capacity. Winbond and SMIC have the upper hand and should be able to negotiate more favorable terms for producing the Buried Wordline technology. A deeper relationship with Winbond could be in the offing.


It is unlikely to impact the supply-demand situation in the short term, but could negatively impact the supply-demand in the mid-term as competitors react by increasing output to maintain market share.
The Micron-Nanya cooperation creates a more formidable competitor in both the DRAM and NAND space and sets up Micron to become one of the top three memory vendors along with Samsung and Hynix. Elpida and Qimonda, pure DRAM vendors are looking for a NAND play. Stay tuned.

Thursday, March 6, 2008

NAND Flash Capex - Full Steam Ahead!

Despite the current oversupply environment and the heavy price drops this quarter (Intel reported ASP declines of 53%), NAND flash manufacturers are maintaining aggressive capex plans. Samsung's ramp of its Austin facility and Hynix's ramp of M11 will offset some of the retirement of 200mm NAND capacity from both companies.

IMFT has maximized its capacity at Manassas and Lehi and will ramp its Singapore fab with maximum capacity of 60k wpm starting in Q3/08. Toshiba/SanDisk's Fab3 reached its full capacity of 150k wpm in September 2007 with Fab4 volume ramp started in December. Fab4 is a massive fab with maximum capacity of 210k wpm.
Powerchip Semiconductor is currently building two fabs P4 and P5 dedicated to the manufacture of NAND flash. These fabs are not expected to ramp in volume until the 50nm generation in 2009.

Spansion as well as foundry partners TSMC and SMIC capex figures includes fungible capacity that may be used to manufacture Mirrorbit NOR.

Wednesday, March 5, 2008

Qimonda Digs Itself Out of Trench

Qimonda created a stir last week when it unveiled a new DRAM technology called “Buried Wordline Technology” which purportedly consumes less power than conventional stacked capacitor DRAMs. This is achieved by reducing the capacitive coupling between the bitline and wordline by burying the wordline in the substrate of the device. An image of the buried wordline in Qimonda’s presentation appears to show tungsten as the wordline material. The middle of line process in the memory array is greatly simplified compared to standard stacked capacitor devices resulting in reduced process complexity.

Another surprising feature was the absence of a trench capacitor. Qimonda has responded to skepticism regarding the scalability of trench technology by switching to a cylindrical stacked capacitor for its DRAM roadmap down to the 3xnm generation. Qimonda plans to ramp its 65nm buried wordline technology with cell size of 6F2 in parallel with its 58nm 8F2 trench capacitor technology in the second half of 2008. The 65nm 1Gb device sports a die size of 55mm2 comparable to Micron’s 68nm 1Gb DDR2 chip size of 56mm2.

Subsequent 46nm and 3xnm 6F2 generations will be introduced in 2009 and 2010 respectively with a 3xnm 4F2 technology targeted for 2011. According to Qimonda, the 46nm technology is expected to effectively double the number of die per wafer versus the 58nm trench technology allowing the company to take the lead in productivity vis-à-vis the competition. This productivity boost is particularly important for driving down costs, especially in light of the massive €598 million net loss the company posted in the fourth quarter of calendar year 2007 on net sales of €513 million.

Qimonda stated that an incremental investment of €100 million over its existing investment for trench technology would be required to make the transition to the new technology. This would include deposition and etch tooling for hi-k capacitor formation.

The company hinted that the new technology opens up partnering opportunities which is expected to include Qimonda’s manufacturing and foundry partners, Winbond and SMIC although Inotera's participation is questionable given the recent tie-up between Nanya and Micron Technology. If licensing deals with these parties are reached, they would be expected to make similar investments for the conversion from trench to stack technology.

Currently, Qimonda, along with partner Nanya Technology, are the sole remaining DRAM players engaged in the development and production of trench capacitor DRAM technology accounting for a combined 18% of the DRAM market in 2007. The remaining 82% consists of Samsung, Hynix/ProMOS, Elpida/Powerchip and Micron Technology, which are all part of the stacked capacitor camp. With Qimonda’s announcement, trench technology is expected to be phased out of the market in the next three years.

It is also difficult to ignore the timing of the announcement. It appears that Qimonda was pressured to reveal details of its roadmap plans after rumors of a joint development and production agreement between its partner, Nanya Technology and Micron Technology surfaced. The rumors were subsequently confirmed when a memorandum of understanding was announced between the two parties on March 3.

Monday, February 11, 2008

Flash Memory Replacement Achieves Major Milestones

At last week’s ISSCC, Intel and STMicroelectronics presented a paper demonstrating the first multi-level cell phase change memory [PCM]. The chip is a 256Mb device based on 90-nanometer process technology. PCM has been touted as a potential replacement for flash memory due to its fast reads and writes and superior endurance.

On the same day of the ISSCC paper, Intel and ST began shipments of prototype samples to customers of a 128Mb chip also based on the same 90nm technology. Careful reading of the ISSCC paper implies that the 256Mb MLC device and the 128Mb device are one and the same chip. The MLC technology is still in the research phase and it appears the prototype samples have the MLC functionality disabled.

The 128Mb product is mainly a learning vehicle to improve the technology for volume manufacturing, to learn about which applications may be suitable for PCM and to develop the firmware to support the devices. Due to the high initial cost of PCM, the 90nm devices will have limited production volume, however, this is expected to change as PCM migrates to more advanced process technologies and MLC technology is deployed.

PCM’s time will come as flash memory encounters scaling limitations within the next five years.

SanDisk's 3-Bit/Cell NAND Flash Technology Not a Short-Term Play

The development of a 16-gigabit 3-bit per cell (x3) NAND flash on 56-nanometer process technology by SanDisk and Toshiba offers the promise of further NAND flash memory cost reductions.

Due to higher design complexity, lower number of usable blocks and longer test times, x3 technology is expected to offer approximately 20% cost savings over comparable MLC devices at the same process generation.The key innovation of the x3 16Gb chip is the write performance of 8-megabytes per second (MB/s).

This outstanding performance is comparable to 5xnm MLC devices currently on the market and was enabled mainly through the development of an All Bit Line [ABL] architecture and advanced programming algorithms. ABL will also be deployed on all of SanDisk/Toshiba’s MLC devices boosting write performance to 34MB/s – comparable to SLC devices on the market.Alas, the full cost benefits of x3 won’t be realized for another two years, the reason being x3 is currently about one year behind MLC on the technology roadmap.

Nevertheless, Samsung and Intel - Micron Flash Technologies are expected to join the fray in 2008/9 and by 2012, 3-bit per cell NAND is forecast to account for 52.8% of NAND flash memory bits, followed by MLC NAND at 25.4%, 4-bit/cell NAND 16.6% and SLC NAND 4.4%.

Friday, January 4, 2008

Qimonda and Macronix Join the Flash Memory Fray

On January 3, Qimonda AG and Macronix International announced they had signed an agreement to jointly development a variety of non-volatile memory technologies over a five-year period. Qimonda will contribute its 300mm wafer-based technology and non-volatile development expertise while Macronix will contribute its extensive know-how in flash memory technologies. Both partners will share development costs and contribute engineering resources and know-how. The joint development will be conducted at Qimonda’s 300mm R&D and manufacturing facility in Dresden, Germany.

Our Take

Macronix’s original Taiwan Stock Exchange filing reveals that this agreement is a “Flash Technology License and Joint Development Agreement”. This coupled with Qimonda CEO, Kin Wah Loh’s remarks concerning Qimonda’s intent to resume production of NAND flash memory chips in 2009 implies that both parties intend to participate in the NAND flash memory space.

Macronix has made some notable progress with its charge trap flash technology, BE-SONOS and it is quite possible the joint development will direct its efforts on this technology, among others. Macronix’s plan to develop 45nm BE-SONOS in 2010 does not totally square with Qimonda’s 2009 timeframe. BE-SONOS is unproven and even if both parties were able to introduce product in 2009 or 2010, they would still be one to two process generations behind the current NAND players. On the other hand, if they are successful, they could have a technology which is more scalable than the current floating gate technology.