During the holidays I asked the leading movers and shakers of the industries we track to share with us their visions and apprehensions for the coming year or two. Twenty-one of them responded, all of them busy with family and business, which I took to be the most flattering thing I can think of. I've tried to summarize their thoughts, and I've shared them with you unabridged.

The most significant thing in our industry in 2004: The most common thing mentioned was PCI Express; over 25% of the respondents picked that. Convergence in home entertainment and the explosion of media centers was the second most popular vision, and credited by some as leading the tech revival. In addition, networked devices (e.g., DVDs DVRs) were seen as pushing the growth of convergence, and digital TV pushed by huge increases in the supply of TFT LCDs were cited. SoCs in embedded and handheld devices are expected to continue to proliferate, and some of the respondents feel we are on the cusp of a real revolution in media capabilities on handheld devices, and enhanced graphics in handhelds (cell phones, etc.) will further drive the economics of the game market, as part of the continuous movement toward mobile communications. At the high end there will be new and better games fostered by the growth in 3D production.

Some of the respondents expect more industry consolidation and at the same time continued fragmentation of customer requirements, and yet, some thought it will be just more of the same, faster, higher image quality, cheaper, etc., while one respondent thought we'd see more investment in 2004.

Technologically the respondents are expecting to see immersive visualization environments, active wallpaper—on-the-wall wallpaper, large use of 90nm, and the emergence of Longhorn. Some expected Intel to do some type of backpedaling on IA-32 vs. Itanium.

Industry growth: The forecast for growth varied from as low as 3% for the big workstations to as high as 25% for notebooks. PCs in general ranged from 5% to 15%, with most of the opinions nearer to 5% (which is our forecast also). Consumer electronics will be "huge," and although no one forecasted a number for it, they all thought it was going to grow drastically.

The most troubling or scariest thing the industry is facing in 2004: Privacy and rights protection and the potential for over reaction to those issues plus spam could force regulatory responses that would stall innovation and growth—death by communication / spam on your phone.

Global political instability and terrorism, along with Industrial terrorism, were on the minds of a few of the respondents.

Failure of innovation by the vendors to deliver usable products combined with lack of killer apps spurring purchase of new apps and upgrade cycles—to continue to grow the industry needs a next big thing. There is no economically viable model to create graphics cards for high-end professionals because the volumes are not sufficient to justify the development cost.

Some of the respondents think the home entertainment industry needs a new OS. Microsoft's monopoly in OSs can and will stifle innovation. There was a concern about the lack of competition... Microsoft, Intel, NVIDIA, ATI.... no one new to stimulate things. That combined with the government buying the same old ways from the same old guys, despite the rhetoric.

As for margins, the lessening gross margins due to the continuous price wars have some worried, while others fear the continue growth in demand but decline in ASP profits in industry are coming more from expense cuts than from real growth or technology improvement. But then there was also concern that capacity shortages (and price increases) could stifle the growth rates that are returning.

The increasingly complex PC buying experience had some worried along with the problems of getting all of these devices to actually work. One respondent was worried about whether there would be a significant resistance to the adoption of LCDs in TV.

And after all this time—we still have no compelling use for 3D outside games or design applications, which could result in the number of jobs in 3D and may not keep pace with the number of students being trained.

---

What do you think will be the most significant thing in our industry in 2004?

Paul Allen: I guess I see the melding of mobile communications technology (CDMA/GPRS/EDGE), etc., into mobile devices (smart phones, miniaturized PCs) with mobile computing platforms as the probably the highest impact trend ... Our own "mini-PC" has come a long way, you should see it ...

The evolution of the set-ops to be more capable (the Digeo MC1 is what I'm involved with) as well as game platforms (PSX, etc.), PC as a media platform, combine to make this area an interesting one as well.

Bob Bennett: Exciting growth of 3D production capability in non-traditional geographies/locations.

Rick Bergman: PCI Express will have significant transition and market share implications with PC, chipset, and graphics vendors (but will not really technically have a major impact). It is not the most exciting or sexy thing, but it will create change and start to open doors to new form factors and different thinking around the PC.

A 3D graphics benchmark that is accepted by the industry would be really significant.

Bob Bishop: Accelerating industry consolidation.

Richard Brown: I think the most significant thing will be the continuous movement toward mobility, either through 802.11-enabled notebooks or through smarter phones with built-in photographic imaging, video, and audio capabilities. We saw significant growth in both these segments in 2003, and I think growth rates will be even faster in 2004.

Louis Burns: The most significant thing in 2004 will be the continued convergence of computing and consumer electronics. Moving forward, we believe you'll see the emergence of a new converged industry with new products, services, and business models. Intel will be leading the way with silicon building blocks, reference designs, and form factor/design innovations for this new converged industry, as well as facilitating the development of standards and availability of premium content.

Pat Gelsinger: The widespread proliferation of wireless communication capabilities in a myriad of devices. As we work toward "Radio Free Intel" we expect to see radios embedded into more diverse set of devices in more and more industries from consumer electronics, information technology devices, health care, transportation, and manufacturing. While this vision will require several more years to occur, we expect 2004 to show tremendous progress in this area.

Alex Herrera: I don't have a great answer, but here's a couple random thoughts:

• Recordable DVD breaks through to mainstream. Also, the "make or break" year for consumer video editing. There are no cost or infrastructure constraints anymore. DVD-R is cheap and a perfect target medium, big disks are cheap, USB 2.0 and/or FireWire are in place. So if consumer video editing is ever going to take off, it should start to show an uptick in 2004.

• Intel does some type of backpedaling on IA-32 vs Itanium. Itanium will not be their only 64-bit option and they will come clean and re-spin.

• PCI Express.

Jen Hsun Huang: I think this will be the year of media centers—lots of them. Many devices will evolve into media centers:

• Consumer desktop PCs will evolve into media centers—already on home networks, huge disk drive, and can serve media nodes throughout the home. High-res display and keyboard makes it the best platform for managing lots of content. Media center PCs will drive resurgence of consumer desktops.

• DVRs will be networked add media center functionality, and can back-up to PC or stream from PC.

• DVD players will be networked, add media center functionality, and can stream from PC.

• Media center is where the PC and CE industry will converge/collide.

• Linux will find its 2nd major market (after servers).

Michael James: I believe our industry will see a long overdue recovery. It will not be back to bubble levels but tech will start being purchased and deployed. Business purchasing will start to pick up. The incredible 9% productivity increases in the USA will be correctly credited to tech deployment. However, I think that home entertainment will lead the tech revival. Media PCs will evolve into viable products. Displays will be the high-ticket item driving the home market. The average amount spent on home entertainment systems will be 400% of that spent just 3 years ago. For the cash the user will get a vastly more entertaining experience. With the installed base of HD displays in DVR-equipped, Internet-connected family rooms the content will change both in quality and distribution. This will lead to a new round of Silicon Valley start-ups and the VC cash will flow again.

Chris Lin: High-performance PCI Express GPU supporting DirectX 10.0 and Longhorn by 90nm process technology.

Gerry Liu: One of the most prevalent trends already showing itself is the conversion of PC hardware and software companies to consumer electronics. Witness the reduction in importance of Comdex with the rise of CES as the dominant show in the year's calendar for PC-related companies. Microsoft, who is finally getting traction with their Media Center PC, just announced the Portable Media Center, which will no doubt be aggressively marketed as the next thing. The Apple iPod, which is their most successful product in years, is making waves at consumer electronics giants such as Sony. Unlucky for Sony, the iPod mini just announced promises to be an even bigger hit. Even graphics chip makers like S3 Graphics will join the "urge to converge" with game console hardware/software content in the likes of other high-profile graphics chip companies that have a piece of the game console pie. All things considered this is a great trend for a lot of exciting technology that needed to get out of the beige PC box!

Further, the S3 Graphics's GPU-powered Via Glory PGC (Personal Gaming Console) platform, which took the CES 2004 Innovation Honors, is a perfect example of synergistic co-marketing by a global chip company in the consumer space. Look for more innovations like this to come in 2004.

Paul McNamara: Jon, over the past few years much of the growth in computer graphics has come in the entertainment space, with some really spectacular advances being made in the gaming space. SGI has, for several years, been the leader in large, immersive visualization environments, which typically consist of a wide field-of-view display (often 10 to 20 feet wide, and sometimes wrapping around to the side walls, ceiling, and floor), and a high-performance graphics computer capable of driving 4 to 8 million pixels (and sometimes more) in real time.

Wide arrays of large customers use these centers to facilitate collaboration among their scientist, engineers, and technicians. Today, these environments are only affordable to the most elite users—national labs or multi-national corporations. We see this kind of immersive environment in 2004 being used in conference rooms and other small team environments. Product teams can get together to review the latest design concept. Production teams can visualize log data daily instead of monthly or quarterly.

The vast majority of users today visualize data on a single-user workstation. While these individuals get significant benefits from this kind of visual environment, the really powerful benefits can only come from real-time interactive collaboration. By making these immersive environments affordable to a small team we will see the benefits of collaboration and interaction realized at a department level.

John Metcalf: General upswing in the electronics industry as a whole.

• more investment

• firmer pricing (but still falling)

• increased utilization of fabs and consolidation in 0.13um with some moves toward 90nm

• More devices with programmability due to increased cost and complexity of SOC

Dave Orton: I continue to be intrigued with the advances in three areas:

Digital TVs, digital full-featured phones taking on the camera and the PDA, and the Digital Camera, partly because I am now more active in these areas, and part of it is the impact that each will have on the older infrastructure in each area.

In the PC industry, we have PCI Express coming but we will need more than this to create a new step function. I see more coming to align with this: Longhorn, new gfx, new multimedia, etc.

Bob Raikes: Taking the displays industry, the big change will be the huge increase in the supply of TFT LCDs in the market over the next year. Desktop computers will be sold as standard with LCDs in most Western countries and a substantial part of the TV market will be LCD-based. A potential cloud on the horizon will be flagged up by the entry of Intel alongside Sony and Philips as recent developers of LCOS technology, but this won't prove much of a threat in the short term.

Dirk Schunk: The introduction of PCI Express as well as new games driving the new need for better and greater hardware.

Ross Smith:  I don't see anything of huge significance happening in 2004.... just more of the same, faster, higher image quality, cheaper, etc. The one thing that we will see incrementally is more embedded training starting to happen. Also, next generation sensors will start to happen as well, which will drive our market a bit also. The other thing that will start to creep in is some more digital cinema products like beyond-HDTV projectors, etc.

In the consumer space, enhanced graphics in hand helds (cell phones, etc.) will further drive the economics of the game market. Whereas the consoles are top dogs today, I think handhelds may overtake them one day... although not 2004. In the console space, we'll see the next-generation Xbox and PSx start to materialize.

Neil Trevett: Programmability and Shaders. The evolution of graphics processors from hardware state machines to general purpose compute engines will make significant steps forward this year—sowing the seeds of a far-reaching revolution in how and why graphics hardware is used. Open standard shading languages will be ratified and refined, graphics silicon will shake off more of the vestiges of their fixed function heritage and become ever more CPU-like in generality and programmability. Most importantly, the industry infrastructure and ISV awareness will reach critical mass so this new technology will begin to benefit end-users—making interactive graphics applications more realistic, and simulation and compute applications more interactive.

Embedded and Handheld Graphics. 2004 will be seen as the year that handheld 3D graphics began to ship for real—with an explosive growth reminiscent of PC 3D graphics in the mid-nineties—as the market suddenly becomes energized by the convergence of multiple enabling threads: a standard API, a killer application (games), and enabling silicon technology—except this time the key metric is polygons per watt not polygons per second. OpenGL ES will be the cross platform API that triggers this market growth—with software implementations shipping in the first 6 months of 2004 and hardware accelerated versions beginning to take over in the second half of the year. There will be lots of debate, just as there was in the PC industry, as to whether CPUs by themselves will be fast enough to drive graphics. But the graphics hardware will win out again, not just by providing more polygons and pixels, but by providing them for less microwatts than general purpose CPUs.

New Format Displays. New display technologies, such as OLEDs, will begin to liberate graphics from being encased behind a flat piece of glass. Devices with new organic forms will have displays embedded in them in innovative ways that enhance both form and function. We are also one step closer to the real killer application for graphics, active wallpaper—on-the-wall wallpaper, which will enable you to interactively drag and drop HDTV displays, hi-resolution pictures as well as wallpaper patterns anywhere in your home—instantaneously.

Dan Vivoli: We are on the cusp of a real revolution in media capabilities on handheld devices. Similar to the 3D revolution in PCs of the late 90s, 2004 will see a step function increase on handheld graphics and video capability.

We will see a growing distance between the capabilities of "free" integrated graphics and stand-alone GPUs. GPUs are increasing capability and performance at a much faster rate. This year, "cinematic computing" (a visually compelling, immersive computing experience) will achieve a whole new level. This is especially important because consumer interest in gaming as well as digital devices such as MP3 music players, digital cameras, and DVD players is skyrocketing. This increasingly places PCs at the center of home digital entertainment systems. And places the GPU in a more central role.

Media Center type devices will be huge this year. We are finally starting to see the "convergence" of media computing and consumer appliances that people have predicted for so many years. I'm particularly excited about this because it opens up new vistas of mainstream consumer media experiences.

And, wireless and broadband access will continue to explode. One year ago, finding a wireless connection was rare. A few hotels offered broadband. Today, you can find wireless connections surprisingly often. And most mid to upscale hotels offer broadband access. I expect the acceleration to continue.

Dan Wood: Within the sphere of Matrox's primary market focus: the productivity PC desktop.

2004 will see the continued fragmentation of customer requirements, reflecting the end of revenue growth solely from general purpose performance improvements. In particular, productivity customers with unique needs will seek flexible, custom solution providers. We also see high growth in multi-monitor capable desktops driven by the strong multi-display advantages offered by LCDs, which will ship to market in high volumes.

---

What do you think will be the industry's growth (%) in 2004?

Bob Bennett: About 5%.

Rick Bergman: PCs: 9% overall, 13% notebooks, 7% DT.

Discrete Graphics: Track with above numbers with slightly higher growth in the 2nd half of the year due to blockbuster titles hitting (finally, finally) and consoles starting to get stale.

Workstation: Breakout year due to new tools and productivity spending fuel by strengthening economy worldwide.

Bob Bishop: 3% to 4%.

Richard Brown: I am quite optimistic for 2004. I think that the desktop market will grow by at least 10% in volume, while growth in the notebook market should exceed 25%.

Louis Burns: "The industry" is pretty loosely bounded, so we can't give a percentage prediction on growth. What we can say is that we see some strong indicators that we are poised for the strongest market growth in the last 3 to 4 years. 2003 holiday strength in digital devices, digital content, and wireless home networking is an indication of a substantial growth opportunity. Intel is poised to take advantage of this opportunity given our breadth silicon technology—CPUs, chipsets, communications, as well as our manufacturing capabilities and scale.

Pat Gelsinger: What Louis said (above).

Alex Herrera: Tough question, since the industry over-contracted and now will probably over-expand to compensate, such that the 2004 growth numbers are probably not indicative of anything long term. I haven't spent a lot of thought here on the industry in general, but my guess says near 10%.

Jen Hsun Huang: No idea how much the industry will grow this year. You will tell me when it's over :-)

Michael James: I don't think in percentages. However, I believe we will all be pleased at this time next year.

Chris Lin: No solid idea.

Gerry Liu: This kind of question will always come back to haunt you in 6 months. Never put a number on things when a vague qualitative answer will do. In general, my prediction is in the single-digit area. Consumer Electronics is currently enjoying a renaissance with HDTV, DVD, digital photography, and a host of other neat gadgets. The efficiencies that the big tech companies have achieved with workforce reductions will bolster the bottom line, which will eventually lead to growth. The growth will be steady and modest over some period of time.

Paul McNamara: We expect to see solid double-digit growth in this segment over the next few years. Accordingly, this will be a focus for SGI that we believe will drive overall revenue growth for SGI's Visual Systems business.

John Metcalf: Semis as a whole 10% to 15%.

Dave Orton: PC growth should be OK, but not great. My sense is that the DT will be high single digits, and the notebook will be near 20%. DTV will be huge, as will HD-DTV, but from a small base.

Bob Raikes: There will be around 15% growth in revenues for all display devices, with LCDs up by 40% plus.

Dirk Schunk: That is depending on the segment but in general I believe that 7% to 10% is a fair expectation.

Ross Smith: Vis-Sim, etc.: 10% to15%.

Consumer: Who knows?

Neil Trevett: The PC market will be relatively flat while the growth in embedded and handheld 3D graphics will be very strong—as the industry begins to transform the humble cell phone into an all-pervasive personal computing device.

Dan Vivoli: It feels like we are poised for a good year. The economy is rebounding. In the business world, a huge number of PCs will need updating after years of tight IT budgets. At the consumer level, more users than ever are strongly motivated to upgrade their pc to support the requirements of their new digital media devices such as MP3 music players, digital cameras, digital camcorders, and DVDs.

---

What is the most troubling or scariest thing the industry is facing in 2004?

Paul Allen: The whole area of DRM is evolving, although "scary" isn't the right word.

The world economy continues to be subject to downturn because of possible terrorist activity.

Bob Bennett: The number of jobs in 3D may not keep pace with the number of students being trained.

Rick Bergman: The easy answer is capacity shortages (and price increases) that could stifle the growth rates that are returning. But a close second is the continued slipping of next generation OSs (Longhorn, X86-64) and APIs (DX10, OGL2.0). We can't move to a 5-year cycle on these things.

Bob Bishop: Lessening gross margins.

Richard Brown: Privacy and security are the two major threats facing the industry, and if they are not dealt with effectively they could seriously hamper growth in the future. By privacy, I mean the ability to combat spam and other intrusions into your system. Email used to be a very simple and effective communications tool, but this is becoming less and less the case as hundreds of unwanted emails flood into your inbox on a daily basis. As for security, we have to find better ways of protecting our corporate and personal networks from hackers and virus threats.

Louis Burns: The scariest scenario at this important time would be a failure of innovation by the vendors to deliver usable products that solve real business problems or make people's lives better. This could be caused by an extreme cost-reduction orientation, or an assumption that usage models, business processes and consumer behaviors will not evolve with advancing technologies. Intel will continue to focus on innovation both from a building block standpoint with our silicon producst and from a platform standpoint with our reference designs and form factor and ease of use initiatives.

Pat Gelsinger: We remain concerned that a number of negative trends could work to undermine the progress and impact of our industry. Piracy concern could lead regulators to "control" hardware designs, privacy concerns, and spam could force regulatory responses, viral internet growth could be thwarted by outdated telecom regulation brought to the Internet due to VoIP emergence, fears of China's growth or jobless economic recoveries could lead to mature countries taking extreme isolationist stances. Many such lay on the horizon but a prudent approach encouraging innovation, limited regulation, and maximum opportunity for market forces needs to temper all these risks.

Alex Herrera: Lack of killer apps spurring purchase of new apps and upgrade cycles. Who besides hard-core gamers really needs to buy a new PC? Looks like Longhorn is really 2005, not 2004, so that won't help spur software and hardware upgrades.

Jen Hsun Huang: Most challenging will be to get all of these devices to actually work. PC media center XP will work for sure. Everything else will take a few years.

Michael James: The home entertainment industry needs a new OS. Microsoft's monopoly in OSs can and will stifle innovation. Our industry has accepted that no one can produce an OS except Microsoft. This accepted truth can delay the revival of tech. Microsoft acts in Microsoft's best interests not the industry's. Microsoft has lost the ability to make UIs that are easy to use and convenient. They make UIs that are powerful and flexible. Who can fill the vacuum in home entertainment OSs?

I bet on an Open Source solution that will be universally adapted by the Japanese and USA.

Chris Lin: Continue growing in demand but declining in ASP (Average Sales Price).

Gerry Liu: The scariest thing that our industry is facing is the same threat that all industries are facing except the duct tape and plastic sheeting industry, global political instability and terrorism. The tech industry is a beautiful thing in general but quite fragile as recent events have shown. We can only hope for the kind of stability and peace that allows the advancement of the human condition.

Paul McNamara: Despite seeing double-digit growth in Reality Center business, I think the overall growth rate for commercial computer visualization systems will remain flat for the next 2 years as more basic engineering work moves to lower cost PCs from Unix workstations. 

Now, let me describe the biggest issue that we face. The emergence of a mass market for graphics cards has driven prices down dramatically. Today, the companies that make graphics cards are focused exclusively on the needs of mass market gamers. The needs of professionals are not being served. There is no economically viable model to create graphics cards for high end professionals because the volumes are not sufficient to justify the development cost. At SGI, our approach will be to provide professional features via software. So, we will use mass market cards and write software (fragment shaders and the like) to provide the features that our customer need.

John Metcalf: Industrial terrorism? Al Qaeda et al. move away from the ever more diffcult targets to soft, but high-impact targets like fabs.

• Death by communication / spam on your phone.

• To continue to grow the industry needs a "next big thing" (after PC and mobiles) and I don't think I've seen it yet.

Dave Orton: We need the Corporate World to begin to grow, but I still see profits coming more from expense cuts than from real growth. The PC industry needs to have the corporate side to begin to spend again.

I am concerned about some aspects of our business slowing down in terms of innovation, and this will hurt our segment: 1) The DX world is moving slower. Microsoft is focused on Longhorn. 2) The process technologies are coming at a slower pace. 3) Intel continues to look for more silicon growth areas and this has its threats on companies like ATI, NV, Broadcom, Atheros, and others. They could do a Centrino on the Desktop, etc.

Bob Raikes: The scariest thing is if there proves to be a significant resistance to the adoption of LCDs in TV applications outside Japan. If this happens, then there will be a lot of new large LCD capacity that will end up in the LCD monitor business, which would be driven to new levels of low pricing.

However, this doesn't seem to be happening!

Dirk Schunk: The continuous price war of the manufacturers as well as the retailers/discounters!

Ross Smith: Vis-Sim... the gov't keeps buying the same old ways from the same old guys, despite the rhetoric.

Consumer—Lack of competition... Microsoft, Intel, NVIDIA, ATI ... no one new. Only in the handheld space does there appear to be real competition happening.

Neil Trevett: After all this time—we still have no compelling use for 3D outside games or design applications—no interactive use for 3D that enhances the life of the majority computer users. Even the early Longhorn demos, with all the power of 3D at the disposal of the GUI, have not created any new usage paradigms that leverage the power of 3D beyond new gimmickry. Perhaps we really are doomed to live with the 3D chasm, with most people not caring about 3D, until the advent of genuine three dimensional displays, which will finally liberate us from peering at 3D through a 2D window.

Dan Vivoli: To me, one of the scariest things is the increasingly complex PC buying experience. Consumers buy based solely on CPU speeds or disk capacity and know little about the rest of the PC. With the explosion of media-rich usage mentioned above, it is crucial that the consumer be more educated about the critical components in the system. The wrong graphics or core logic, for example, can leave someone with a PC that can't play games or videos well.

###