We are back with part two of the PM Forum. Part one covered the RDRAM support situation, the possible future of integrated graphics in the high-end market, the Athlon64 benefits, upgrade cycles of Intel chipset based motherboards, and the ability of the Opteron processor to make it in the high-end mainstream market. In part two, product managers give us feed back on topics such as VIA’s market perception, diversification of motherboard companies into other technology mediums, the roles of Intel and AMD, features to expect in motherboards down the road, and the timeline of HyperThreading to make it in the mainstream market.

Participant - Official Title, Company.

Hunter Lee – Product Manager, ABIT Computer
Darryl Chan – Marketing Division Director, Albatron Technology Co., LTD.
Gerald Wang – Sr. Director, AOpen Inc.
Richard Liu – Product Manager, Asustek Computer Inc.
Steven Kuo – Product Manager, Chaintech Computer
Fanny Chen – Product Manager, DFI
Eric Kuo – Deputy Manager, Elitegroup Computer Systems
Calvin Yen – Product Manager, Epox Computer Co., LTD.
Chris Wang – Director of Product Marketing Division, Giga-Byte Technology
Scott Yang – Senior Product Manager, Micro-Star International Co., LTD.
Jonathan Yi – V.P. of Product Marketing, Shuttle Inc.
Tom Yang – Section Manager of Product Marketing Division, SOYO Computer Inc.
John Nguyen – Product Manager, Tyan Computer

1. Do you see VIA as a successful...

1. Intel Pentium 4/Pentium5 chipset maker?

2. AMD Athlon64/Opteron chipset maker?

PM #1:

1. VIA lost timing in Pentium4/Pentium5 chipset develop. No advanced spec. can persuade customer adopt for VIA chipset platform MB.
2. Good reputation in AMD platform and good quality of compatibility than NVIDIA

PM #2: VIA got the license from Intel and will launch PT800 and PT880. Some signs show the market will accept VIA for its brand and features. In the near future, VIA will get some market share from SiS in P4. For K7, the VIA is the best seller in the mainstream while NVIDIA is good at high-end. For K8, it’s not clear now and NVIDIA could be the best at first. After 1-2 quarters, we have to watch the situation.

PM #3: VIA p4x series always sells at the low end market. NVIDIA puts too much competition on VIA at K7/K8.

PM #4:

1. Intel’s technology is always a generation ahead of others. Customers will still prefer Intel for high end boards.
2. VIA’s success in Athlon 64/Opteron boards will depend on success of the CPU itself.

PM #5:

For the Intel platform, VIA stay in value segment well, but their chipset’s reliability still is an issue.
For the AMD platform, actually, K8 has integrated the memory controller; north bridge will not be a focus for every chipset provider. VIA still is a leader for south bridge.

PM #6: VIA has been losing ground since the patent war began in court not too long ago. Now that the dust has cleared and the case is settled, it’s in the best interests for a competitive market to have VIA still co-exist with other chipset vendors as a major player. Unfortunately as it stands now, VIA is spread too thin: it needs to re-think its strategies to focus on the core competencies, and keep aware of dangerous competitors to themselves such as NVIDIA who have managed to take the mainstream K7 market from them.

PM #7:

The launched schedule of PT800 is late considering VIA got the license from Intel. I think it is going to be hard to gain the P4 market from competitors.
The K8T800 chipset supporting AMD 64 is ready now. It looks like VIA could lead the market to promote it very successfully.

PM #8: VIA Technology Inc. is one of the leading technology providers in the world; we respect and believe in their expertise.

PM #9: VIA is in a very tough situation for 03 and 04, and if they can not settle down their strategy and internal management issues, it will be worse.

The market perception of VIA is important to both VIA and their customers, because this will filter down an inherent image for motherboards and motherboard companies using their chipsets, thus effecting sales. Market perception is nothing more than the “mantra,” but this alone carries a lot of weight. As it is widely known, VIA got the license situation with Intel settled and is now ready to relaunch a more full scale line of chipsets that support the Intel platform. With the Pentium 4 now widely adopted, VIA’s future market position will have much to do with their success with Intel supported chipsets. VIA is currently working on rebranding themselves, and the market is taking this in mixed reactions. In our manufacturer inquires, we have learned that VIA is losing AMD production ground to NVIDIA’s nForce2 in Q2 and Q3. This relatively new competition by NVIDIA is making it hard for VIA to maintain their previous image of performance and stability. The blood-letting price cuts, as one CEO referred to in our CEO Forum, has only made this even fiercer and has somewhat lead to the industry’s new perception of VIA as the “SiS of AMD platforms” [basically, a price player].

VIA has a very good chance now to solidify their future, as the market is still waiting on hearing their strategy. This is evident from the 54% of product managers giving VIA the thumbs up for the Intel side. The support that these product managers lend are on the premise that performance, compatibility, etc… has been worked out for future VIA chipsets for the Intel platform. Their past experiences with the P4X400 chipset wasn’t remembered too fondly, which is why the rest of the product managers are still raising an eyebrow. Either way, VIA is going to face an up hill battle. They can ability to win, but this will depend largely on their market strategy and the ability/capacity of their Intel supported chipsets.

Feedback was more positive on the AMD platform side, where 69% of product managers see VIA as a successful Athlon64 and Opteron chipset maker. But as we noted above, VIA is facing harder competition with NVIDIA. Their market perception will depend largely on what takes place between the two chipset makers as time goes forward.

We followed up by asking, “What are your thoughts on VIA’s current market position, particularly with the recent announcement of the KT600 and PT800?”

PM #1:

I think everyone will agree that VIA is one key player in the chipset market, especially in AMD platform. Regarding the kt600/pt800 chipset:
1. Based on these chipset’s spec and performance. You can say they are not high end products.
2. Flexibility

PM #2: VIA needs to seriously consider quality control on their products. Customers will want stability and assurance that the product won’t fail over a short period of time. Having been given the taste of competitors’ products, which have better support and higher quality control, it will be an uphill battle for VIA.

PM #3: VIA is currently work very hard on get back to the mainstream market position by developing new Intel & AMD platforms. The lead-time is the main-factor to the judgment. We believe the market will figure that out with in two quarters.

PM #4: They will win back some market share and get their good performance image back but will still behind NVIDIA due to lack of support for dual channel DDR with KT600. As for PT880, development schedule is too late for introduction to channel.

PM #5:

So far, KT600 was positioned for the high-end market because it can support FSB 400 & DDR 400. As to PT800, it just can compete with 648FX in the mainstream market.

PM #6: There are in the value to mid segment.

PM #7: VIA still enjoys its market share & brand-name in the AMD platform market. Regarding Intel platforms, the future seems quite tough.

PM #8: The performance of these (PT800 and KT600) is limited due to its older 0.22micron process. This makes VIA chipsets suitable for the lower end segment, no change from its position in the past. Introducing PT800 for the high end P4 and KT600 for K7 will not help raise VIA’s perceived value.

PM #9: We believe VIA is still a major player for AMD platform. VIA has great expertise, as well a price advantage compared to competitors. Both KT600 and PT600 have good cost/performance ratio for end-users.

PM #10: Too late for the market, VIA will stay at entry level/low end segment with KT600/PT800, but an integrated graphics solution may be better for this segment.

PM #11: KT600 is good for the mainstream. PT800 is still not exposed in the market. We need to watch the reaction from user.

2. Many companies have begun to diversifying into other markets, such as Small Form Factor PCs and wireless technology. Does your company plan on pursuing...

1. Small Form Factor

2. Wireless

3. Consumer Electronics (LCDs, PDAs, MP3 players, etc…)

4. Optical Media/Storage

PM #1: a, b, and c are more related to motherboard makers.

PM #2: We believe the SFF will be next big thing for IT industry. The SFF will provide better integration for both functional abilities and life style. We will introduce our SFF in early September and focus aggressively on this market.

PM #3:

... we’ll pursue this market (SFF) for sure, and wireless is the future with no doubt, but there are already tons of competitors, and the industry standard are still changing, not too good for a late comer, the situation applies to optical disk drive too.
As for consumer electronics, it’s the game for consumer electronic giants such as Sony and Toshiba, M/B manufacturers may have chance in OEM business.

PM #4: Diversity is important for all m/b makers for a better future.

PM #5: We will adopt various IA and XC PC more for market now.

PM #6: Channel research and resource build up right now.

PM #7: It will be important to diversify on the capabilities of one’s product lines as new players enter fresh markets in search of expanding margins. The diversification strategy should be considered as a tie-in to core competencies rather than a revolutionary tangent that may not bring the best return on investment in the long run.

PM #8: It all depends on one company’s core competence.

This reflects what we asked in the CEO Forum – Q3/2003, where 92% of CEOs felt that their company’s future depends on their diversification. According to the responses, it seems that everyone from ABIT to Tyan is looking into the Small Form Factor. However, we should note that at least 31% of the product managers stated that they were pursuing SFF cautiously, which means it is very possible that they will never get to market. Irregardless of that another 31% of product managers reported that they will be pursuing SFF very aggressively. The rest of the product managers (38%) weren’t specific in which manner they would pursue SFF.

Of all the four markets that we inquired about, wireless was the one in which we got the widest spectrum of information. Over half of the product managers (54%) stated that they were planning on pursuing wireless, but only 23% (8% - cautiously and 15% - very aggressively) mentioned the manner in which they would pursue it. The rest of the respondents either answered no or gave no response.

Consumer Electronics is a market segment that many motherboard manufacturers are very new in. This seems to be the main reason behind our responses, and the fact that the largest cluster of responses were in the “yes, but cautiously” segment seemed to reinforce this. Meanwhile, 38% of product managers claimed that consumer electronics wasn’t even on the discussion table at their company. Optical media/storage brought even more hesitation, as the majority of PMs (62%) wrote that they had no plans to pursue this market segment. Only 30% of the respondents answered “yes,” and half of that figure stated they planned on pursuing “very aggressively.” This means that competition will likely be lower here among motherboard manufacturers than compared to the other three segments we inquired about.

3. Intel and AMD are the two largest PC processor companies in the world. How do you see each company’s position in the industry moving forward?

PM #1:

There is no doubt Intel is a leading company among processor companies, I also think sooner or later Intel will migrate the desktop processor to 64 bit computing to compete with AMD64.
AMD might gain some market share from Intel next year with AMD64, but it takes times to catch up Intel.

PM #2: AMD will co-exist with Intel and continue to remain a small player. The only question is how small will AMD be?

PM #3: Intel still dominates the overall market with all of its marketing power. AMD is only good at some of the particular markets.

PM #4:

Intel will still gain the moment with its excellent strategy.
AMD will stock in the pool of lacking strong new technology for the user that seek for better performance from AMD.

PM #5:

Intel is definitely the leader for the PC processor market.
But AMD has their own market share because they always provide the better price/performance processor.

PM #6:

Intel: Excellent manufactory ability and technology leader.
AMD: Not solid planning in new generation CPU.

PM #7: The portion of AMD & Intel CPU market shares might keep the same ratio in the near future. But Intel has better chance to get stronger.

PM #8:

Intel is the largest semiconductor company in the world, as well as a developer, and they have been executing very well in the past years. We still see Intel as the dominant solution provider in the PC industry, and it strongly, continuously, owns the position.
People see AMD as the best cost/performance solution provider in the market. However, by introducing the new K8 family (Opteron, Athlon64FX, AThlon64), AMD should be able to extend their business into the next level.

PM #9: I believe Intel will remain dominant in the foreseeing years, while AMD has to fighting hard for surviving, even with the Athlon64.

PM #10: In the long run, we believe both AMD and Intel will enjoy the benefits from different target markets. Intel will still occupy the major consumer PC market, but AMD will use its newest AMD 64 / Opteron CPU to win more niche markets such as gamer/server market.

PM #11:

For the desktop market share: Intel 80~85%, AMD15~20%
AMD will move forward to server and IA market more aggressively

PM #12: Intel will likely gain market share into the server chipset segment with their ability to offer a wide variety of products for that type of customer, even though other providers still hold a responsibility to many of the larger tier1 OEMs. Intel will still continue to hold a good share of the workstation market, although it is possible that AMD may also take some share especially with the support from third party vendors such as NVIDIA who can help to offer upcoming technologies such as PCI Express. In the desktop processor market, Intel will still have sizeable control, although they are likely to be sharing the market with AMD still.

The feedback is basically unanimous on the Intel front. The large success of the Pentium 4, Centrino technology, and a good marketing strategy are only the latest reasons to assume that Intel will be around for a good deal of time. The mainstream market is going to be largely Intel saturated, but we may see something happen with the Athlon64. One of the product managers offered a bleaker perspective into AMD’s future, “AMD will co-exist with Intel and continue to remain a small player. The only question is how small will AMD be?”

We at AnandTech see the co-existence of AMD and Intel to be ultimately important for consumers and the market at hand. A viable alternative to any product can only bring better prices and better products, because it doesn’t leave either side complacent. AMD will no doubt try and push forward their products in the market, but how the length of their opportunity window will depend on the availability and ability of their upcoming products. The good news is that AMD’s market perception is not viewed as based on a poor market strategy or that they have a poor product. Instead, AMD’s market perception is mainly seen as the by-product for the hold out over the next generation and [for the moment] the competition’s better product. Additionally, AMD has a good sized market base for their own niche market, which they can use to propel the introduction and adoption of their next generation products.

4. Profit margins for motherboards have shrunk in the recent past, and with chipsets integrating more and more features on-die it has become exceedingly difficult for motherboard makers to differentiate their products from competitors. Onboard sound and LAN are standard nowadays; even the lowest-end motherboard comes with these features. Serial ATA RAID, IDE RAID, and even wireless technologies are now starting to become standard as well. What should consumers expect to see one year from now? How will companies distinguish their products?

PM #1: From next year, new technologies like DDR-II & PCI Express will be the major changes to desktop motherboards. In order to make differentiation, we will continue improving on ease-of-use, developing useful hardware & software features.

PM #2: One year from now consumers could expect to see more of the same, although the technology will have evolved to such architecture as UAA, PCI-E, and integrated 802.11b. Companies will have to focus on price, ease of use, and functionality to stand out among competitors. More than likely, strong, highly visible marketing will play more of a role in pushing desktop solutions to the mainstream audience.

PM #3: We believe that besides the technology trends, there is something more that is important and appreciated by our customers: more creative ideas, outstanding performance, more user friendly applications, a great experience with a company’s products, impressive product packaging/outlook, plus extremely high quality and service. These are key elements for us to be successful in the future.

PM #4:

Develop some special features to distinguish product is necessary. More focus on:
a.) Easy to use
b.) Good quality more stable
c.) Excellent performance…
Only easy to use and add on value features will adopt for end user.

PM #5:

We are trying to make our product satisfy particular user groups. Like making a series product for PC gamers, another series product focuses on multimedia player
Also to deeply understand behavior of these users’ group and work on user friendly features for them, hence, even with the same chipset spec., but different series product will have different features for each field.

PM #6: Wireless LAN, Dual LAN, High speed SATA

PM #7:

Besides the productivity, one good thing for technology is bring down the cost. “Affordable technology” is the trend for the PC industry. A year from now, “faster (better productivity)” & “cheaper (lower total cost of ownership) still are the factors for consumers. As you mentioned: it’s hard to differentiate products from manufacturer to manufacturer…we already foresaw the trend while back ago. The solution for it has been three points:
1. How to develop a product more “humanized (user-friendly)”...
2. Understand the market/consumer needs...
3. Customer-oriented Service...

PM #8: PCI Express is the next industry revolution that will influence the market, not only the M/B, the graphics solution and other add-on cards will need to be re-designed, I see there will be quite a lot compatibility issues with PCI Express for a while, and that is the challenge and chance for M/B manufacturers to find their niche.

PM #9:

Consumers always pay the same price but get more features due to the competition among the IT industry. Next year, they can take advantage of some new specifications, such as PCI express, DDR2, 64 bit CPU and so on.
Our products will be separated by two segments, one is price oriented product, and the other one is spec-oriented.

PM #10: The next area of growth for motherboard manufacturers will be the front panel and mini PC (SFF)...

PM #11: There is no doubt about the trend of chipset integrating more and more function on-die. Consumers can expect that the motherboard manufacture won’t only highlight on the product spec and technology side but also the user friendly, plentiful accessories, outward appearance design and so on. Those will be the points where motherboard makers differentiate their products from competitors.

This is a very interesting question to ask and watch answered, because motherboards are slowly getting differentiated by brand name instead of the features. It seems many of the companies are focusing on the “whole customer experience.” This isn’t just making a product easier to use, but also making the product interface more “humanized,” as one product manger noted.

Purely from a motherboard standpoint, we will almost always see a spectrum of motherboard selection. A good example is the approach of Albatron and Chaintech, who are arguably two of the fastest growing non-tier 1 motherboard companies. Albatron has focused their strategy on providing high quality but very low priced products, which is why their motherboards aren’t feature laden. Chaintech, on the other hand, is focused on feature rich products with less regard to the price point. The result has been good sales to both, because they fit well into their respective niche markets. However down the road, the question becomes with integration will the spectrum of motherboard selection become smaller? For instance, the new chipsets from Intel has more USB controllers than previous generations. Companies like Albatron can cut price points by simply not putting less USB ports on their motherboard, but the cost of the chipset with added USB controllers can’t be avoided.

Price can do a lot for the differentiation of two products, especially if it is by brand name, but we don’t see it as the only issue for the two quarters. The blood letting price cuts have for the most part ended, and the next two quarters will still provide opportunities for other companies to find their life-saver, like Shuttle’s adoption and pioneer of the Small Form Factor.

5. How far off is HyperThreading to the mainstream market? When will the average PC user need this type of computing?

PM #1: Don’t think the general publics were educated enough about the true benefits which HT is offering...

PM #2: I guess it would be around the end of 2003 ~ 1H 2004, depend on what CPU price can offer then.

PM #3: HyperThreading ready processors will likely be transitioned into mainstream starting early next year. HyperThreading for single processor machines will experience mixed results though, as the technology only applies to specific applications designed to handle such capabilities.

PM #4: Actually, it is not so important, but Intel’s strong recommend to the end user will force the market to take it as a must.

PM #5: Theoretically, every user will benefit from HT since multi-tasking is quite common in current PC usage models. For HT to actually make it into mainstream market, it will highly depend on its price premium.

PM #6: According the Intel’s plan, the shipment of FSB 800 CPU is going to account for 50 percent till Q4/03’, so at that time the CPU with FSB 800 and Hyper-threading will be the mainstream product.

PM #7: In our view, end-users are already enjoying the benefit of HyperThreading, especially for background virus scanning. Why does AnandTech think Hyper Threading is not available to the user today?

PM #8: Intel will push all their P4 CPU to incorporate the HT. I feel this is a common feature that will not be emphasized at the end of this year.

PM #9: As everybody knew, Hyper-Treading technology became a standard feature with all of new Intel FSB800 P4 CPUs. The technology was pioneered on Intel's advanced server processors, helps your PC work more efficiently by maximizing processor resources and enabling a single processor to run two separate threads of software simultaneously. Even though, only some of users enjoy the benefit by using particular software\environments at this moment, but we believe the population are significantly growing.

PM #10: Hard to say based on the motherboard supply; HT is basically more a market gimmick than a real performance boost for me.

PM #11: You might see the Hyper-Threading technology occupy the mainstream market by the year’s end. In the PC industry, sometimes the product spec and feature aren’t driven by consumer side, but drive by producer side. Consumers may not need this spec, but when most producers produce this kind of product and offer reasonable prices, I don’t see any reason why a customer won’t want to buy it.

PM #12: At this moment, we haven’t seen great success for HT yet. But don’t get me wrong, the HT will be a trend in the future especially more and more programs run simultaneously and use multi-tasking. The user who has HT will enjoy the benefits for sure.

Overall, the feedback is that HyperThreading is going to really get its full steam in the last half of 2003. Q3 ends in September, and so it looks like some is discussed by the product managers will be left for Q4. As one of the product managers wrote, “According the Intel’s plan, the shipment of FSB 800 CPU is going to account for 50 percent till Q4/03.” As a result of 50% shipment now, it reinforces our thoughts of Q4, due to timing issues. The adoption of HyperThreading (HT) in the mainstream market will be effected by the outcome of Athlon64, but since HT will because a standard for Intel, it cannot be avoided, only delayed.

It should be made clear that AnandTech does believe HT is available for the everyday consumer, and will be a good feature for those that take advantage of it. However, we still do feel that the average consumer walking into a consumer electronics outlet will not need the level of computing that HT brings for applications of this day. While the need will not fully be addressed till software catches up, we will likely see full adoption of HT for the average consumer in Q4 and Q1/2004.