Q4 2007 Wrap

Q4 came in at $0.69 EPS (non-GAAP) on $1.246 billion in revenues, exceeding consensus estimates — once again. Although this part of the earnings’ report was somewhat mixed, it was basically good news.

The same can’t be said for guidance. SanDisk guided to the low end, or below, already lowered expectations, for both Q1 and 2008.

In general, it appears that softness in consumer demand coupled with expected aggressive first quarter NAND price declines will more than offset SanDisk’s strengths in mobile and international markets. SanDisk’s continuing success in fab ramps will help, but only some.

Not reassuringly, SanDisk pointed out the similarities between Q1 2008 and Q1 2007. Then, as now, apparently a least one component manufacturer decided to dump inventory and in the process destroyed pricing.

Guidance for 2008 as a whole is down, but not on the order of Q1’s misery.

One of the interesting twists to the 2008 guidance is an implied strong second half. For example in its 29 January, 2008 report following earnings, Bear Stearns lowered Q1:08 from $0.38 to $0.28; lowered Q2 from $0.35 to $0.21; raised Q3 from $0.50 to $0.55; and raised Q4 from $0.85 to $0.91. Net 2008 is now $1.96 down from $2.08. 2009 is estimated at $2.39. (Non-GAAP).

Although Citi lowered its estimates across the board, the numbers themselves follow a similar pattern (29 January 2008 report.) Citi lowered Q1:08 from $0.40 to $0.29; lowered Q2 from $0.49 to $0.29; lowered Q3 from $0.71 to $0.46; and lowered Q4 from $1.16 to $0.85. Net 2008 is now $1.90 down from $2.76. 2009 is estimated at $2.24. (Non-GAAP).

Although SanDisk refused to elaborate in the call, it appears that SanDisk expects fundamentals to improve in the second half of the year as the market moves to a better balance between NAND supply and demand.

Competitor’s supply cutbacks, including wafer cuts, capex cuts and 200mm shut down will likely play their part. It is not clear if a new catalyst (or two) for SanDisk is also in the pipeline.

In any case, SanDisk isn’t cutting back on the wafer front. No need. In fact SanDisk doesn’t expect to be able to keep up with product demand, without NAND supply from others. This appears to be primarily a second half story. Purchases of non-captive NAND will likely account for 5-10% of supply for the year, as SanDisk growth outpaces industry-wide bit growth in 2008.

One 2008/2009 story that I am watching closely is 43nm. This could prove to be an important milestone for SanDisk. It will mark the arrival of multi-bit technologies beyond 2 bits per cell and it should be the technology node which enables affordable viable MLC SSDs.

Next up is 3GSM on 11 February and this year’s SanDisk analyst day on 25 February.

I have included excerpts from recent analyst reports at the bottom of this post and a Q4 conference call transcript in the “Pages” sidebar.

Q4 and Full Year 2007

As expected, in Q4 international sales were strong and mobile was the number one revenue contributer.

International accounted for 56% of Q4 retail sales. SanDisk expects international growth in 2008 to outpace the U.S. growth with Asia/Pacific getting an extra boost from the Summer Olympics.

On a year-over-year (YoY) basis, Q4 U.S. retail revenue was down 13% and international retail revenue was up 32%. SanDisk’s weakness in the U.S. market appears to be more a case of overall slower Q4 holiday sales than loss of market share.

Mobile was SanDisk’s largest revenue generator by end market both for Q4 and for 2007 as a whole. For 2007, mobile accounted for 35% of total revenue. Mobile product unit sales more than doubled YoY led by continued strength in microSD cards in the retail and OEM channels.

Sales of imaging cards grew nicely in 2007, in part due to resurgence in the digital camera market. Unit sales of high-performance Extreme and Ultra cards targeting the image market doubled YoY. Sales of iNAND, SanDisk’s embedded flash product line, grew significantly in the fourth quarter as SanDisk made substantial gains in mobile and the portable navigation device market.

Sales of Sansa Clip MP3 players were strong during the holiday season. Interestingly SanDisk explicitly pointed to its sub-$100 product strategy for audio/video.

SanDisk reiterated the importance of its new Shanghai assembly and test factory for microSD and Memory Stick Micro cards. SanDisk was able to reach industry-leading capacities thanks to its nine die stack assembly technology.

In Q4 SanDisk repurchased $202M in stock completing its $300M buyback program. Citi thinks a new program is unlikely (near-term) given CapEx commitments and the cost of its operating lease in Japan. Personally I like the idea of a share buyback program to offset option dilution. I hope SanDisk puts another program in place.

SanDisk’s Q4 OEM revenue grew 26% sequentially to $391 million, driven by microSD and iNand products sold to mobile handset vendors. SanDisk also saw nice sequential growth in the GPS market, and in the industrial market.

L&R came in at a record $128 million. Up 51% YoY and 8% sequentially. For the full year of 2007, L&R revenue reached $450 million which was 12% of total revenue, increasing from 10% of revenue in 2006 largely due to the addition of Hynix as a new licensee.

Non-GAAP Q4 product gross margin (GM) came in at 29.7% at the high end of SanDisk’s 27-30% guidance.

As expected, non-captive NAND supply was an issue in Q4. SanDisk was supply constrained during the first half of Q4, but was able to secure the additional non-captive supply in the second half of the quarter.

Although SanDisk wouldn’t admit it, there seems to have been a self-conscious decision not to sacrifice GMs over an incremental increase in revenues. Product GMs at 29.7% were one of the bright spots in Q4. Even for Q1, if SanDisk can keep to its guidance of 23 to 26% this is still much better than the teens that some are fearing.


Curiously this Q4 conference call reminded me of Eli’s January CES presentation. Ordinary, vanilla, unexciting, and unexceptional, are all adjectives that come to mind.

Paul Coster from J.P. Morgan had his finger on the pulse of the conference call when he asked whether maybe NAND-based memory had “just kind of run out of steam.”

Clearly this is not the case. Solid state memory is key for the next generation of computing and consumer devices. The story doesn’t lack for color. So why did we get what felt like the same oft-told story on this conference call?

My guess is that there are inter-related explanations. First, SanDisk probably has things to say, but chose not to use this forum. Sometimes its better and necessary to wait on announcements. In addition, Analyst Day is only weeks away. Sometimes a story plays better when told in its entirety.

Another explanation is that quite likely the same old story is the story. In a nutshell, I think 43nm is a huge deal for SanDisk and for the rest of the NAND industry (4xnm). When SanDisk tells us that all is going well with the 56nm to 43nm transition, this simple story has a lot of ramifications. All positive for SanDisk.

The baseline is that the first to make the transition in volume to 43nm are going to have significant cost advantages over the competition. SanDisk/Toshiba look like winners. Immersion lithography seems like the key to 4xnm. As Eli described in the conference call, SanDisk/Toshiba decided to move to immersion lithography at the 56nm technology whereas the competition stuck with dry lithography. SanDisk/Toshiba has the edge in high volume experience.

Probably even more significant is that 43nm is where SanDisk has decided to ramp multi-bit technologies beyond traditional 2 bit MLC (x2). SanDisk has announced that x3 will be introduced in volume at 43nm and that x4 will also be introduced at this node.

Another SanDisk 43nm angle is the arrival of the 32Gb chip, for both regular MLC and x3. 32 Gb x3 and MLC 43nm chips will likely arrive later in 2008 after SanDisk has experience with both on 16Gb chips. If I’m reading the slide below from 2007 analyst day correctly, in the next two years, a 64Gb x4 chip at 43nm is on track to arrive at the end of the year. Even if these are just samples or they slip to early 2009, heads will turn.


It will be interesting to see the competition’s response. Would not be at all surprised if multi-bit technologies beyond x2 end up a key in the SanDisk/Samsung royalty agreement negotiations. In any case, I expect that Samsung will renew its license in 2009, but at lower rates than Samsung’s current IP license calls for.

In addition, from what I can tell, 200mm isn’t worth the effort at 4xnm and the DRAM to NAND capacity conversions are impractical for the 4xnm NAND technology generation. These issues aren’t relevant for SanDisk, but certainly are for the rest of the players.

Also it looks like the first significant skirmishes in the SSD wars will be fought at 43nm. In the Q4 cc SanDisk confirmed that its MLC SSDs will be arriving with 43nm MLC. MLC SSDs are expected to sample late in 2008.

In a most revealing aside during the Q&A Eli said that SanDisk wouldn’t do anything to “aggravate the component situation out there.” What wasn’t said is that SanDisk/Toshiba has already taken a big chunk of NAND market share. See the IDC (November 2007) graphic below.


**** Analyst report excerpts below ****

29 January 2008
Craig A Ellis

Buy/High Risk 1H
Price (28 Jan 08) US$25.89
Target price US$46.00
from US$57.00
Expected share price return 77.7%
Expected dividend yield 0.0%
Expected total return 77.7%
Market Cap US$5,931M

SanDisk Corp (SNDK)

Margin Upside Adds Leverage Ahead Of An Eventual Upturn

Stock Strategy — Our bullish call on SNDK share has been the wrong one in the past four months. However, surprisingly robust 2H07 gross margins augur well for EPS revisions ahead if fundamentals firm as seasonal demand strength and supply headwinds converge in 2Q08 as we expect. US demand risks persist, though trough P/E multiples discount 2008 EPS power 10% below our just-lowered estimates, suggesting the stock may now be overshooting on the downside. Target to $46 from $57 for 89% ETR (aftermarket).

Results and 1Q08 Outlook Reveal US Demand Weakness But GM Positives – Clean EPS of $0.72 was $0.07 above the Street helped by pricing (-11% qq vs CIR -20%) and product GM (+330 bps to 29.7%, CIR 28.6%) while 37% bit growth disappointed (CIR 60%, guide 45-60%) on tight supply and US demand weakness. US weakness caps guidance at $775M-$875M in rev (misses Street’s $1.0B) but 23-26% product GM’s are well above high-teens fears.

While 2008’s Rev Growth And Margin Outlook Is At Low End Of Target Model – SNDK expects 15-25% rev growth, 24-28% product gross margins and 13-16% operating margins. The guidance range falls at the low end of the long-term rev growth and GM and OM outlook, unsurprising given US weakness (~33% of sls). Heightening the risk of a downward reset at the 2/25 analyst day, the variance seems significantly explained by visible US recession risks.

Estimates Move Toward Worst Case – CIR 2008E and 2009E EPS decrease 31% and 33% to $1.90/$2.24 (stock -40% and -22% in 4Q07 and 1Q08TD), with the Street likely settling near our estimates at the low to mid point of SNDK’s guidance range.

Sell-Off Looking Overdone, Time To Revisit

SNDK shares trade with NAND flash contract pricing (0.6 r-squared), given the direct relationship to SNDK’s royalty and product revenues. Our industry view has been that 2008’s unusually significant manufacturing supply headwinds (200mm decommissioning (Figure 3) and 40nm node shrinks (all players)) coupled with handset, imaging, video and PC (USB and SSD) demand drivers would bring supply and demand into balance in 1H08, a positive for contract pricing. Against this backdrop we have and continue to view SNDK as well positioned for revenue growth and margin improvement given a strong manufacturing cost position (300mm mix, strong mfg transitions in 2005- 2007), ramping global product market share, and a growing and diversifying royalty revenue stream (albeit with mid-2009 Samsung renewal risks).

On a P/E multiple basis, the 2003-2007 F12M trough is 13.7x (Figure 2), suggesting the $24.5 aftermarket price is discounting 2008 earnings power of $1.78, a level 9.6% below our new estimates. Viewed differently, the shares have pulled back 56% from 4Q07 highs, discounting 81% of the 2001 decline (Figure 1) when F12M EPS fell to -0.67 from $0.63. Such stock action looks extreme in our view, in particular with time increasingly on SNDK’s side as 2Q08’s US seasonal strength approaches and as OEM new product plans for 2H08 invite buyers back to merchant manufacturers at cash cost pricing. We conclude it makes sense to start building positions, adding intensity into 2Q08 as supply chain data points affirm tightening fundamentals.

Financial Outlook

Below we recap positives and negatives from the quarter and outlook and set forth our 1Q08 and 2008 modeling assumptions.

4Q07/1Q08 Result and Outlook Positives

Product gross margins grew 340 bps q/q to 29.7% on mix (high-end cards), the 56nm ramp (~67% of mix), and benign pricing (-11% q/q)

OEM revenues of $391M grew 26% q/q (above 21.7% q/q total product revenue) on strength in microSD, mobile handset, and GPS.

EMEA cards and USB drives market share grew 400 bps y/y in November

Mobile unit sales accounts for 35% of total revenue, doubled y/y.

Inventory DOI to 57 days from 74 days in 3Q07 and the lowest level in days since 4Q04.

Repurchased $202M in stock (~4.5M shares or 2% of shares out) and completed $300M repurchased program, though new program is unlikely given CapEx commitments and cost of operating lease in Japan.

2007 operating margins of 13% exceed company’s 2007E guide of 0% to 10%.

Gross Margin guided to 23-26% (down 375 to 675 bps q/q versus CIR’s -615 bps to 22.5%) on 56nm mix and lower Fab 4 startup cost ($30M in 4Q).

Full-year 2008 operating margin outlook of 13-16% implies a doubling through the year from ~9% to ~18%, helped by expense controls (hiring). 43nm ramp (in 2Q08) is said to be on-track given ‘significantly’ more experience (transitioned to immersion litho in 56nm).

4Q07/1Q08 Result and Outlook Negatives

4Q07 total revenues of $1246M (20.1% q/q) missed CIR/Street’s $1,315M/$1,267M (license revenues of $128M missed our $139M).

Bit growth of 37% was below guidance of 45-60% (CIR 60%) on weak US consumer spending and captive supply constraints.

OpEx grew $39M to 19% of sales, 100 bps above CIR’s 18% estimates on higher SG&A (seasonal ads and merchandising).

1Q08 total revenues guided to $775-875M (-38% to -30% q/q) below CIR/Street’s -27%/20% q/q, reflecting uncertain demand and more aggressive pricing.

1Q08 operating expense of $215 to $220M (25% qq) is above CIR’s $204M.

1Q08 Implied PF EPS of $0.22 to $0.35 below CIR/Street’s $0.40 and $0.42. While the US recession warrants a new degree of management caution, the 2008 margin and revenue guidance is nonetheless at the mid-to-low end of the company’s long term target financial model.

Estimate Revisions

1Q08 Financial Projections

We project total revenues of $827.4M and pro forma EPS of $0.29, estimating $127.1M in royalty revenues and $700.3M in product revenues (Figure 5).

Detailed product modeling includes -13% bit growth (inline with seasonal declines) and 28% price/mb declines. See Figures 6,7 for quarterly product revenue, bit growth, and revenue growth expectations.

While our bit growth estimate (-13%) is near seasonal norms (-10%), the supply-constrained 4Q07 compare is an easy baseline, seemingly embedding a degree of conservatism in our outlook.

On royalties, we model $127M. Every 5% change in contract prices impacts quarterly EPS by $0.02 and annual EPS by $0.075.

We model gross and operating margins down 574 bps q/q to 24.0% and 844 bps to 9.5%, respectively versus 1,400 and 1,500 bps declines in 1Q07 (1Q07 suffered from both a weaker inventory and cost reduction position).

First Call consensus estimates are $1.001B in revenues and $0.42 in EPS, and embed blended gross margins of 33.36% (vs CIR’s new 35.7%). On sensitivities, every 5% change in product revenues impacts quarterly EPS by $0.04, while every 1% change in product gross margins impacts quarterly EPS by $0.02.

2008 Financial Projections

On a full-year basis CIR projects revenues of $4.385B (up 12.5% and below 15-25% guidance) on royalty revenues of $502.7M (up 11.7%) and product revenues of $3.882B (up 12.7%). We model pro forma EPS of $1.90. Product modeling details include 176% bit growth (vs. 207% 3-year averages) and 59% price/mb declines (in line with 3-year averages) (Figure 10).

Our royalty estimates incorporate 125% Samsung bit growth and a 55% price/MB decline with declines each quarter. However, clear precedent exists for increased contract pricing even absent the tough supply challenges facing industry in C2008 For example, from February to September 2007 contract pricing rose 70%+ while in 2006, 4G prices increased 14% over 2 months).

Our royalty estimates exclude revenue from the 25 card/drive companies with which SNDK has brought legal action in the past few months (favorable early settlements with RiTek and PNY Technologies suggest material royalty revenues are possible in 2H08). We estimate annual royalty potential at $40M+ if all are signed.

We do not expect a refresh of the SanDisk/Samsung royalty agreement near- term, which is up for renewal in mid-2009, though we expect SanDisk to address this issue at its February analyst day along with a general discussion on royalties/licensing.

Product-specific revenue expectations are set forth in Figures 8. We expect SSD’s to emerge in 2H08, though primary revenue drivers are handset and imaging cards, as well as USB drives and MP3’s. Take-TV and any related content licensing is not yet incorporated in our model, though could emerge in 2H08.

We model product gross margins of 27% (up 190 bps yoy and 100 bps above the target mid-point) and operating margins of 13.5% (up 49 bps yy and at the low-end of the target range, but with steady improvement through the year (Figure 8)).

Fieldwork points to more gross margin tailwinds than headwinds. Positives are lean retail inventories entering 1Q08, consolidating European market share (from mid 20s to mid 30s), a favorable 300mm mix positions vs Samsung and Hynix, a 43nm node transition on track for mid-2008 and 3 bit per cell launch and commercialization in late 2008 and early 2009, respectively. Partial offsets are Fab 4 ramp up costs and increased second sourcing to 5-10%.

At its 2007 Analyst Day, SanDisk set forth a long term (2008/9) target model
with YoY revenue growth of 25-40%, product gross margins of 24-32%, and
operating margin 15-20% respectively (Figure 8). CIR estimates are below
the mid-point of all SNDK targets excluding operating margin targets.


28 January 2008
Goldman Sachs
James Covello
6 month price target $38.00

NAND S/D and royalty renegotiation will be key for the stock in ‘08

What’s changed

SanDisk (Buy) reported Q4’07 sales of $1,246mn (+20% qoq), slightly below our $1,270mn (+22% qoq) estimate. We estimate that 4Q’07 proforma EPS was $0.59, $0.04 above our $0.55 estimate (including ESOs) and ~$0.05 above the Street (including ESOs). EPS upside vs. our estimate was driven by a higher than expected gross margin and lower taxes. Bit shipments were +37% qoq, while ASPs were -11% qoq. Q4’07 guidance [probably means Q1 2008] is significantly below the Street, driven by weak bit growth and weak ASP expectations. We are reducing our estimates on lower margins and lower interest income: CY08E to $1.55 from $1.85 and CY09E to $2.40 from $2.65.


We maintain our relative Buy rating as SNDK represents a good long term investment opportunity. In our view, shorter term investors need to wait for a better entry point, which could occur when: (1) we have visibility of NAND supply/demand improving, with Samsung and/or Hynix significantly reducing their 2008 NAND capex budgets. And/Or (2) SanDisk renegotiates its royalty contract with Samsung, which expires in 2009. Remember that all of SanDisk’s current profitability comes from its royalty revenues, as its product business is losing money. Samsung accounts for the vast majority of SanDisk’s royalties, and we view SanDisk’s ability to retain Samsung as a licensee as critical for the stock in 2008. While we expect Samsung to remain a royalty payer in order to gain access to SanDisk’s X3 and X4 technology, investors remain skeptical about SanDisk retaining Samsung. As a result, the resolution of the royalty negotiation would remove a significant overhang from the stock and act as a positive catalyst.


We maintain our 6-month price target at $38 (based on a $24 value of the royalty stream, $7 in net cash/share and $7 for the product business).

Key risks

Key risk to our price target is losing Samsung as a licensee. SanDisk Corporation (L) S&P 500 (R)

[GS dropped its 2008E from $1.85 to $1.55 and its 2009E from $2.65 to $2.40. 2010E was added at $2.75]


29 January, 2008
Bear Stearns
Gurinder Kalra

Target Price YE ’08 $45.00
Long-Term Growth 14.5%

SanDisk Corp
Navigating Through Oversupply And Macro Demand Concerns

SanDisk reported 4Q non-GAAP EPS of $0.69, above our estimate and consensus of $0.64. Revenues of $1.25B were slightly below our estimate of $1.32B, but this was more than offset by better than expected PGM (driven by better ASP), and lower than expected operating expenses.

While SanDisk management expects the pricing environment in early 2008 to be similar to what we saw in early 2007 given the current oversupply, they expect the market to move into balance as we move through 2008, as industry supply growth should moderate given a more disciplined stance by manufacturers with regards to capex plans. SanDisk anticipates demand from their customers in 2H08 to outstrip their captive supply, and as a result plan to increase their non-captive purchases. SanDisk’s comments support our view that 2H08 fundamentals should improve relative to 2H07.

We are adjusting our 2008 non-GAAP EPS from $2.08 to $1.96, as a result of our 2008 bit growth estimate being lowered from 194% to 180% YoY as well as lower royalty revenues. While we are modeling for ASP to decline 58% YoY, we expect SanDisk to see another strong year of cost reductions which should offset our estimated price declines. SanDisk should also continue to gain share internationally in 2008, allowing it to outpace industry-wide bit growth. We are introducing our 2009 non-GAAP EPS at $2.39.

We believe SanDisk’s current stock price level presents a good opportunity for investors to accumulate the shares, as upside potential significantly outweighs downside risk. Though we expect supply-demand dynamics to remain challenging in the near-term, we believe at these levels, the stock is currently pricing in an overly pessimistic view with regards to the NAND outlook for 2008. We believe the NAND market should be balanced in 2H08, and that a bottoming in spot pricing in late 1Q08 or early 2Q08 should serve as a catalyst for the stock.

Lowering price target to $45 from $50 for YE ’08.

[BS lowered Q1:08 from $0.38 to $0.28; lowered Q2 from $0.35 to $0.21; raised Q3 from $0.50 to $0.55; and raised Q4 from $0.85 to $0.91. Net 2008 now $1.96 down from $2.08. 2009 is estimated at $2.39]


9 Responses to Q4 2007 Wrap

  1. Goldy says:

    Thank you for a fabulous summary. While 2008 earnings are somewhat unclear at this point and thus I am hesitant to put too much faith in the cheap forward implied PE ratio, I am 100% confident buying SNDK at essentially the lowest p/book and p/sales ratios it has ever traded at… especially when I believe 2008 is trough earnings.

  2. bob77977 says:

    hi savo,
    eli talks about sandisk’s lead in chip production. here is a link showing the possibility of micron getting to lead. 35nm in Q3.
    if going from 56m to 43nm is 50% cut of realestate on chip, the difference between 43nm and 35nm is roughly 40% cut.
    which means micron’s chips will be 40% cheaper to produce – thats a hell of an edge. of course, one have to compare total productions but still it arouses the question – is samsung going to sign with micron a royalty agreement rather then with sandisk? whats your opinion?

  3. savolainen says:

    Greetings bob77977,

    Sorry about not getting back to you sooner. I have been on the road and otherwise distracted.

    Personally I am not that concerned about Micron. This, after all, is a company hemorrhaging cash with a tradition big on talk and little on show. Micron is bumping along in market share at single digits and needs a huge bump to hit the magic 20% target survival share.

    In-Stat has Intel as numero uno, the prime candidate, for dropping out of the NAND flash business. I agree. If this comes to pass, Micron has big problems — No deep pockets, and No IP force field. W/o Intel, Micron’s NAND aspirations will be toast.

    It might take a couple of years, but in my opinion, the writing is on the wall.


    Micron’s NAND situation would make a good subject for a post, but it might take a while for me to get to it.

    This week, I hope to modu. The week after, if all goes according to plan, I will ramble on a bit about a new most promising candidate for SNDK’s partner in SSSS LLC: [R….s]. ONFI’s worst nightmare.

    SanDisk’s analyst day is scheduled for 25 February. This should be a gold mine which likely will distract me for a while.


  4. Dan says:

    “This week, I hope to modu. The week after, if all goes according to plan, I will ramble on a bit about a new most promising candidate for SNDK’s partner in SSSS LLC: [R….s]. ONFI’s worst nightmare.”

    Don’t think it is Rambus, for the following reasons:

    1) The venture was financed with $10.2M from the partners – Nothing on RAMBUS cash flow.
    2) The venture invested $10M for the acquisition of intellectual property (with obligation of another $32.5M should the venture be profitable – again nothing on cash flow, revenue (regarding the 10M) or even in a note and considering that Rambus has quarterly revenue of $40M this deal should have been very significant to disclose.
    3) Eli said: “we have formed a partnership with another major owner of system level flash IP” – major owner of flash IP means the partner has strong patents on system level IP and while RAMBUS has stated once that they are expecting some royalty bearing agreement (that did not materialize), I doubt that they have strong system level that can interest SanDisk. My guess that they are working with Toshiba but no agreement is reached as of today.

    The questions – 1) who is the partner 2) who is the holder of the intellectual IP that is so valuable to be worth $42.5M even now and 3) Is the partner and the holder of IP the same person.

    The lack of disclosure from the other side tells me that this is not a public company. But who has valuable MLC IP that even SanDisk needs?

    Who developed X4 which is system level solution?

    x4 technology – Nand Flash memory
    mSystems x4 technology was developed in collaboration with Tel Aviv University’s (TAU) School of Electrical Engineering and exclusively licensed to mSystems, for use in flash memory products, by Ramot, the technology transfer company and commercialization arm of TAU

    And check out Simon Litsyn patents

    My guess is that RAMOT is the partner and that the SSD venture will also license the X4 – the system level solution in X4 is probably the core of the MLC SSD solution.

  5. Poofypuppy says:

    Dan, my thinking was that Savo meant Renesas. But I trust that he’ll elaborate more soon. 🙂

  6. savolainen says:

    Greetings Dan,

    Nice to hear from you. I know you follow SanDisk very closely, as your comments once again reflect. RAMOT and x4 would be an interesting twist and it would explain the lack of disclosure on the other side.

    As you correctly deduced, R….s was meant to indicate Rambus. You make a good argument against Rambus, but accounting can be slippery.

    In any case, I don’t feel Rambus is a sure thing, and maybe not more likely than Toshiba, but a lot of pieces fit the puzzle for SSSS LLC — if the key is high speed interface IP which can be combined with SNDK MLC IP.

    I don’t think ONFi is going anywhere, but the direction might be relevant and Rambus is working on NAND.

    Rambus started talking about NAND/flash last year about the same time as SSSS LLC surfaced.

    From Q2 Rambus cc:

    “ I’m not going to talk about this [flash memory] too much because we have some very interesting progress in this area. We are engaged with flash companies looking at some future technology areas.”

    “ Michael Cohen – Pacific American Securities: Are you focused in flash on how it — the memory communicates to the controller, or specifically to unique use of the flash memory and high speed connections there?

    Sharon Holt [Rambus] I would say both, yes.”

    From Q3 Rambus cc:

    “And so, as we’re [Rambus] looking at our NAND efforts going on here… taking a look at very difficult system level problems and how we could apply our Flash technologies to those.”

    As no doubt you are aware, Rambus is up to its eyeballs with DRAM lawsuits with most of the other large NAND players. Maybe in parallel Rambus could be working with one of them on flash, but it does seem a bit unlikely.

    Had better stop now or I’ll never finish modu.


  7. savolainen says:

    Greetings Dan,

    While there appears to be reason to believe that Rambus and SNDK/Toshiba might be working together on some interface angle, I think you are probably right about SSSS LLC: partner Toshiba, IP from Ramot and licensing x4. I may use this Sunday’s post as an excuse to go back through my notes.

    A couple of reservations: as far as I know Toshiba and SNDK have never had a meeting of the minds on Toshiba licensing x4. While it is clear that the Litsyn/Ban patent is remarkable, it is not clear that it is directly tied to x4.

    I don’t know what you know of the Amir Ban lawsuit. Way back in 2004, it was what originally pointed me to the older most interesting Litsyn/Ban patent.

    In any case in the lawsuit Ban revealed that FLSH had been working on a technology known at the time as the Stratosphere technology. The key seems to be programming using groups of flash memory cells in addition to the cells themselves– a very clever way to store more info using the same number of flash memory cells, because groups of cells can store info too.

    In any case Ban claimed that FLSH started working on the technology in 2000 and had a working prototype by 2003 and that as far back as 2003 Toshiba was interested in entering into negotiations for the technology.

    I have long suspected that Stratosphere was somehow connected to x4. If there is a connection, no one has ever explained it , to my knowledge.

    Eli has said that current licenses do not cover x4, so it certainly follows that SSSS LLC might be the licensing vehicle for x4. If so, this would also seem to imply that SNDK is confident in the technology.


  8. savolainen says:

    An acquaintance passed along the following machine translation from Hebrew, of a recent interesting article from the site NRG. Ban and Litsyn have agreed to drop their lawsuit against msystems/SanDisk for compensation of $4M.

    This is the first I have seen of mention of a possible direct connection between Stratosphere and x4:

    Compromise between SanDisk to the keys of the disk-on-key Ban and Litsyn, that developed drives of memory of the flash that are small when worked with M-systems, will receive 4 million dollar in the frame of agreement that is secret. In return they will erase the claim against SanDisk


    At the start of this week was signed agreement of compromise between the company SanDisk, that acquired M-systems, and Amir Ban and professor ‘ Shimon Litsyn, that developed drives of memory of flash are small ) disk on key (when worked with M-systems. By right of the use in this technology is sold if M-systems SanDisk in exchange for 1.5 billion dollar. According to agreement of thaw, Ban and Litsyn will receive about 4 million dollar. In return, is erased the claim that served Ban against parent company M-systems and founder of the company Dov Moran in the assertion of question violated their obligation towards him Onicso to themselves you are profitable the invention. The agreement itself will not be served for approval of the court, inasmuch as that will be prepared secret. The agreement was obtained in the frame of procedure

    Ban was project manager ” stratosphere ” of M-systems, that his different name pursuant for 4x and that business in the development of the disk on key. According to him, if M-systems obligated to establish a company to the project, but after became clear her that the project of winner to the great success and that the company Toshiba is impressed from him very – changed her mind and requested to the asset to herself every fruits of the project. Accordingly sued Ban is a third from the rights that produced from the invention if M-systems and has her shares. Also concluded age, that joined the project in the stage more late, served a claim against if M-systems, SanDisk, Amir Ban and Dov Moran.

    [link below in Hebrew]


Leave a Reply

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

WordPress.com Logo

You are commenting using your WordPress.com 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: