Once and Future King - IDEMA · Once and Future King Richard Kugele Principal ... • The 0.85-inch is DOA ... (factoring increased PMR production)

The Hard Disk Drive:Once and Future King

Richard KugelePrincipal

Needham & Company

Agenda

• History: RAMAC and 50 years of Technological Innovation• The Competitive Landscape: Then & Now• Technology Trends & Battles• Review of 1H06• 2H06 Outlook (Including Media)• Understanding Wall Street-isms• Conclusion

History: First things first… Happy Birthday!

September 13, 1956September 13, 1956RAMACRAMAC

History: Happy Birthday, continued…

The Lima News, September 16, 1956

History: Happy Birthday, continued…

The Lima News, September 16, 1956

History: Major Technological Milestones• 1956: IBM 350 RAMAC

• 5 MB on fifty 24-inch disks (pre-production versions actually used perpendicular first!)• 1961/62: IBM’s 1301-first drive with air bearing heads

• 28 MB on twenty five 24-inch disks• 1963: IBM 1311 initiated era of 14-inch disks and removable disk packs• 1973: IBM 3340 “Winchester” drive arrives• 1980: Shugart (later Seagate) introduced the ST-506, the first 5.25-inch drive (5MB).

• Combined with the ST-412 (10MB) the first hard drives widely used in the PC world• 1982: IBM’s first drive greater than 1GB• 1983: Rodime introduces the first 3.5-inch drive• 1988: First “low-profile” 3.5-inch drives• 1990: MR technology (heads)• 1991: First 2.5-inch (100 MB) Prairietek/First Thin Film Disks (IBM “Pacifica”)• 1995: 2GB Arrives• 1996: IBM introduces the GMR (Giant MR) head ***beginning of extreme AD growth• 1999: IBM releases the Microdrive (170MB/340MB)• 2002: AD growth slows from >100%• 2003: Serial ATA, successor to ATA/IDE introduced• 2005: 500GB drives, SATA 3G, TMR heads, SAS, beginnings of perpendicular• 2006: 750GB drives…

History…continued

• Of course, some ideas haven’t worked out…• Quantum’s Bigfoot• HP’s Kittyhawk• the 0.85-inch

Competitive Landscape: Then and Now

Competitive Landscape: Then and Now, cont..

• According to Disk/Trend, over 230 disk drive manufacturers have exited the space since its inception (from bankruptcy, exits, merger/acquisition, ect..) including:

• Bryant• Telex• Computer Memories • Data Products• Memorex• Century• Marshall• Caelus• Xerox• MiniScribe• Microscience• LaPine• Areal• Priam• PrarieTek• Micropolis• Hewlett Packard• JTS • Rodime• Conner Peripherals• IBM• Quantum• Maxtor

Competitive Landscape, continued…Today

• NHK• Hutchinson• Magnecomp• Innovex• Nidec• Minebea• Agere• ST Micro• Marvell• Ect….

• Seagate• Western Digital• Hitachi (Excelstor)• Fujitsu• Toshiba• Samsung• Cornice• GS MagicStor (?)

• SAE/TDK• Alps• Komag• Showa• Fuji• Hoya• Fujitsu• Kaifa• TK

Technology Trends & Battles

Technology Trends & Battles• Longitudinal running out of gas

• 160GB more challenging than originally expected (perpendicular yields better).

• We estimate 20-30% of all HDDs shipped in 2007 will be based on perpendicular (with a preference towards mobile, enterprise, and high-cap desktop). Perpendicular and Longitudinal likely to coexists only for 1-2 more years.

• Areal density curve returns to the 30-40% growth (from teens), but well below the 60% annual increase in data growth=platter reductions from the 1.5-1.7 range unlikely over the next 2-3 years.

• The 0.85-inch is DOA…not enough volume to lower the cost, not enough capacity to differentiate it from flash, not enough market acceptance to justify new entrants.

• Unclear if the 1-inch ultimately suffers the same fate (Hitachi winding down, GS MagicStor gone, WDC exits/deemphasizes?)

• SAS, after being delayed for so long, needs to broaden adoption or risk ceding the midrange of the enterprise to Fibre Channel, FATA and SATA.

• There needs to be a big push for Hybrid drives, in conjunction with Vista, and despite Intel’s Robson.

• But Flash vs. Drive argument/overhang will NOT go away. “Flashtops” all over the press.

Technology Trends & Battles, continued…NAND Hype

• “There are so many new applications for NAND, there is an infinite demand for flash at the current price points.” Reza Faramarzi, Hynix,

10/6/05• “The high margin of NAND will push capital spending on NAND.”

IDC report, Sept 2005

• "NAND flash will eventually replace other storage mediums, especially those used in mobile products, creating a "Flash Rush," as NAND continues to register an unprecedented surge in demand as the backbone of the mobile electronics era,“ Dr. Chang Gyu Hwang

CEO of Samsung Semiconductor

• “Hard drives are living on borrowed time and will be replaced with solid-state Flash memory” Dr Chang Gyu Hwang,

9/13/05

Technology Trends & Battles, continued…NAND Hype

• “There are so many new applications for NAND, there is an infinite demand for flash at the current price points.” Reza Faramarzi, Hynix,

10/6/05• “The high margin of NAND will push capital spending on NAND.”

IDC report, Sept 2005

• "NAND flash will eventually replace other storage mediums, especially those used in mobile products, creating a "Flash Rush," as NAND continues to register an unprecedented surge in demand as the backbone of the mobile electronics era,“ Dr. Chang Gyu Hwang

CEO of Samsung Semiconductor

• “Hard drives are living on borrowed time and will be replaced with solid-state Flash memory” Dr Chang Gyu Hwang,

9/13/05

YOU ARE AT WAR!!!!!

Technology Trends & Battles, continued…

• Significantly More Capacity/$

• Faster Write Speeds• Better Bit Stability• Multistreaming Capability

• Robustness (i.e. temperature, shock)

• Smaller Footprint• Lower Power Consumption• Building Block

Architecture/Cost

FLASH ADVANTAGEFLASH ADVANTAGE HDD ADVANTAGEHDD ADVANTAGE

We Believe Each Technology Will Have A Place (with flash <20GB)We Believe Each Technology Will Have A Place (with flash <20GB)


Power Consumption Within Laptops by Component

Graphics Card, 8%

Miscellaneous, 8%

Intel® Memory Controller Hub,

9%

Power Supply, 10%

Display, 33%

CPU, 10%

Clock, 5%

Intel® I/O Controller Hub,

3%

LAN, 2%Fan, 2%

DVD, 2%

Hard Drive, 8%

Source: Intel

Hard Drives are Not the Issue with Laptop Batteries…the screen is.


• Unsolicited advice…• Checks suggest 2.5-inch hybrids with 1-2GB of flash exhibit many

of the same power and reliability metrics as an SSD, with greater capacity and lower cost.

• Half measures won’t cut it (i.e. 256/512 MB may not be enough).• “It is from their foes, not their friends, that cities learn the lessons of building high walls.” Aristophanes

• IDEMA or leading players need to sponsor independent system evaluations to set the record straight on specs.

• Full court press marketing vs. flash.• Major initiative to broaden Hybrid adoption.


• But not all endeavors are clear…• Kittyhawk Revisited: The 1.3-inch drive

project• Launched in 1991/92 by Hewlett

Packard-”small, dumb, cheap disk drive”.

• Capacity was 21.4 MB on two 1.3 inch disks.

• Target market was next generation portable devices (ultra-thin laptops, PDAs, gaming). Sound familiar?

• Goal is half of the capacity of a 1.8-inch drive, and 150% of a 1-inch.

• Conclusion…tread carefully. R&D is expensive, and market is unproven (Apple iPod Mini example)

1H06 Review

• Entered the year with a major consolidation announcement and a slowing PC market• Following the strength of 1H05, year-over-year compares were going to be difficult.

• Ultimately, 1Q proved to be solid, with CE, mobile and enterprise all strong to stable. Desktop down roughly 5% (more for Seagate, less for WDC as they picked up share in the channel post MXO). Channel inventories were roughly 5 weeks, with pricing declines normal. Industry’s 2nd qtr of over 100 million quarterly units.

• 2Q was more seasonal than industry had seen in 2-3 years+. Enterprise and CE once again were solid. But mobile saw a decline, which shocked some. Desktop units were intact (down 3% seq.) but pricing was aggressive (particularly at OEMs) ahead of May 19th acquisition close. 3rd quarter of more than 100 million quarterly units.

• By this point, it was becoming painfully clear that the space was fighting over Maxtor share.• Aggression broad-based, but Asian players (Hitachi, Samsung-both of whom lost money in 2Q)

stepped up efforts…WDC in the channel. • Result was meaningful price declines on 80/160GB in particular and billions in lost revenue for

the industry without spurring additional demand. • Guidance for 3Q reflected seasonal units, but a continuation of the price aggression (and

the inevitable lower GM).

2H06 Outlook• Thus far, 3Q exhibiting normal seasonality…pickup mid-late August continuing at decent

pace here in September. Processor price cuts towards the end of August helped somewhat. But pricing remains aggressive (especially on high-cap and mobile).

• Channel inventories have hovered in the 5.5 week range, up slightly from the 5 week range of the 1H (but should decline as denominator changes and seasonality picks up).

• Recent negative preannouncement by Komag interpreted as a negative for the entire space...but this is erroneous. Raw material build (of roughly 1-1.5 million units) represents “buy-ahead” to cover for 2H06 tightness and facility shut downs. Represents, less than 1 million drives (or fewer than 1% of what industry will ship in the qtr.)

• Overall, 3Q should see unit growth of 5-10%, with 5-7% ASP declines. Mobile units could be strong, but ASPs abysmal. Enterprise looks strong, again, as does CE despite Sony.

• Looking into 4Q, we expect unit growth of 10%+, with strong CE, mobile, enterprise and desktop. ASP pressure likely to continue on both high-cap desktop and mobile. But WDC and STX should remain solidly profitable, with STX benefiting from integration synergies later in the year.

• We expect Hitachi to lose money again in 3Q, with its best hope in 4Q. Samsung? Fujitsu has been reactionary, and not an aggressor.

…and we don’t expect Vista to influence anything in 2006…

Media Situation: 2006Assuming 14% Y/Y HDD Unit Growth

Q1 06E Q2 06E Q3 06E Q4 06E C2006E103 100 111 118 432

1.70 194 189 209 224 816Most Likely 1.67 191 186 206 220 802

1.65 188 184 203 217 792

Estimated Industry Supply: 187 194 199 209 789

Shortfall/Surplus @1.70 (7) 5 (11) (15) (28)Shortfall/Surplus @1.67 (4) 8 (7) (11) (13)Shortfall/Surplus @1.65 (1) 10 (4) (8) (4)

Source: IDC, Needham & Company Estimates

Table 1. 2006 Supply vs. Demand at 14% yoy Growth

HDD Demand (millions of units)

Average Platter Count Total Platter Demand (millions of units)

• Extensive capacity additions in 2005 and early 2006 have helped to close the supply/demand gap, but increases remain rational.

• Seasonality in 2Q, enabled some raw material build that affected media production in 3Q, but not demand.

• Tight conditions should return in 2H06.

• Do not anticipate perpendicular recording impacting throughput in 2006 or platter counts.

• Overall, very healthy conditions.

Source: Needham & Company Estimates

Figure 1. 2006 Supply vs. Demand at 14% y/y Growth

194

187

199

209

180

190

200

210

220

230

Q1 Q2 Q3 Q4

Disks in Millions Media Supply

Platter Demand

1.70

1.671.65

Note: Demand lines factor 90% blended utilization and yields.

Media Outlook, continued…2007

Assuming 15% Y/Y HDD Unit GrowthQ1 07E Q2 07E Q3 07E Q4 07E C2007E

118 115 127 136 497

1.70 230 224 247 265 965Most Likely 1.67 226 220 243 260 948

1.65 223 217 240 257 937


Shortfall/Surplus @1.70 (13) 2 (12) (17) (39)Shortfall/Surplus @1.67 (9) 6 (8) (12) (22)Shortfall/Surplus @1.65 (6) 9 (5) (9) (11)


Table 3. 2007 Supply vs. Demand at 15% y/y Growth




Figure 3. 2007 Estimated Supply vs. Demand at 15% y/y Growth

217

226

248

235

210

220

230

240

250

260

270

Q1 Q2 Q3 Q4


Platter Demand1.701.671.65

Note: Demand lines factor 87.5% blended utilization and yields (factoring increased PMR production)

Assuming 12% Y/Y HDD Unit GrowthQ1 07E Q2 07E Q3 07E Q4 07E C2007E

115 112 124 133 484

1.70 224 218 241 258 940Most Likely 1.67 220 214 236 253 924

1.65 217 212 234 250 913


Shortfall/Surplus @1.70 (7) 8 (6) (10) (14)Shortfall/Surplus @1.67 (3) 12 (1) (5) 2Shortfall/Surplus @1.65 (0) 14 1 (2) 13


Table 2. 2007 Supply vs. Demand at 12% y/y Growth




Figure 2. 2007 Supply vs. Demand at 12% y/y Growth

235

248

226

217210

220

230

240

250

260

270

Q1 Q2 Q3 Q4


Platter Demand1.701.671.65

Note: Demand lines factor 87.5% blended utilization and yields (factoring increased PMR production)

Media Outlook 2007, continued…• After years of underinvestment, 2005 and 2006 expansions have helped to

begin to bridge the production gap in media. • But the scope of the industry (more than 900 million platters in 2007) signals

that continued spending will be required…virtually any growth rate will require meaningful increases in capacity.

• We believe 2007 will continue to demonstrate this tight/in balance situation that is healthy for the HDD industry. Perpendicular is likely to have only a minor impact in 2007, but could grow in future years. Component and outright replacements need to continue to be prepared.

• Despite investor fears, a rapid decrease in platter count due to perpendicular is unlikely (due to the gap in data growth and AD, and high-capacity markets such as CE, near-line). We believe 1.5-1.7 represents the likely outcomes over the next 5 years+.

• Overall, the media producers have expanded rationally and prevented becoming a gate to industry growth, or an accomplice to overproduction.

Understanding Wall Street-isms

“If you can’t convince them, confuse them.” Harry Truman

Wall Street-isms…• Conference presentations have become less relevant due to Reg FD, with

marketing trips, one-on-ones, etc… the best way to interact with management teams in a productive way.

• Analysts are now measured and compensated by stock recommendation performance. Translation? Tendency for analysts to act like they are traders, whipping ratings around based on arbitrary price targets.

• Following criticism on ratings distributions, some firms have rules on the number of buys (forcing downgrades to get upgrades).

• Research was deemed too superficial during dot-com era. Today, proprietary data (such as channel checks, market forecasts, etc…) has increased in value to clients…of course making everyone suddenly an expert on “the channel”. Read carefully and consider the source.

• Bottomline, analysts can be a wonderful resource to the HDD industry when used correctly. When done right, analysts represent an impartial aggregator of data that can reflect what you are actually doing (not just the marketing spin) back to you. But we are only as good as the information coming in.

Conclusion• First, relish in your success…you have earned it. So massive is the impact of

this technology on the world, that it cannot be overstated.• Since its introduction, the HDD has been the king of primary storage, and it is

likely to remain so for the foreseeable future. • But this is no birth right, but rather a hard earned place in the IT world,

achieved on the hard work of pioneers such as Johnson, Shugart, Connor, and countless others.

• To stay on top, you need to recognize both threats and opportunities, and pursue both with every asset you have at your disposal.

• The industry remains healthy today, but is also sitting at a crossroads. In our view, the greatest risks besides the economy over the next 1-2 years are: 1) alternative technologies seeking inroads; and 2) irrational expansion plans/share goals.

• If we are to celebrate another 50 years, it will be due to the successful navigation of these two threats.

Thank You

Once and Future King - IDEMA · Once and Future King Richard Kugele Principal ... • The 0.85-inch is DOA ... (factoring increased PMR production)

Documents