RAM | Reseller Advocate Magazine: Issue 61

By William Van Winkle

"Auld Lang Syne" is Scots AND roughly means "old long since." This leaves us with an opening verse to the old folk tune, as translated in Wikipedia, of: "Should old acquaintance be forgot, and never brought to mind? Should old acquaintance be forgot, and auld lang syne?" From there, the diddy descends into scenes of drinking, hills, streams, and more drinking, just like a Scottish country tune should. It's a good song for reflection, especially if you're inebriated enough not to care if the lyrics make sense.

I'm all for celebrating the end of this year—it certainly feels auld lang syne. The market roared up in the first quarter, dropped off a cliff from spring into summer, and is now on a rocketship ride into the ozone. Or so it seems. But that's the whole market. Here in the PC space, things get murkier. New numbers from Current Analysis show that PC sales increased 23% year-over-year for the week of Black Friday, but that's down from the 36% year-over-year growth reported the year before. Notebook sales grew by 51.6% while desktops were down 7.2 percent. That's after desktop ASPs dropped from $493 to $464. Notebooks tumbled like mad from $856 to $708 this year. Holy cow!

Admittedly, most of what Current Analysis is tracking comes from retail channels. I know from Intel that the average whitebook runs about $1,100, which tells me that system builders are targeting a higher-end market than retail. No surprises there. The channel needs to distance itself from pricing battles whenever possible. The real issue is whether this distancing is happening or not. Have we sailed beyond auld lang syne, or are we still living in it?

Let's ask this question from the perspective of notebooks—a worthy context since this is obviously the direction PC sales are going. The old way of selling notebooks is to move tier-one clamshells, Acer probably chief among them. I've even had a major component manufacturer exec tell me to my face that resellers should give up on whitebooks and sell Acer. This mentality is so old school that I can't help but look for starched uniforms and textbooks saying the Earth is flat. Sure—just sell Acer. Just put all of the power in Acer's hands so Acer (and its ilk) can dictate pricing, hardware standards, and lead the likes of Intel and AMD around by the nose. Brilliant. At least Intel is busting its hump to promote initiatives like VBI, put together back-end incentives, and give resellers a mobile leg to stand on. Anybody see Acer doing that for the channel? Anybody?

I know VBI isn't perfect. The designs are pretty hum-drum, and if I had a customer come at me wanting a kickass 17" media monster, I'd probably push him toward ASUS. I'd pick a channel-friendly company like ASUS or MSI over Acer any day. But if you want to be in the mobile mainstream—and I'd seriously question your reasons if you didn't—then you need VBI, especially VBI for the fourth-generation Centrino platform ("Santa Rosa") expected in the first quarter. VBI is the one major differentiator you've got with which to build your own brand and demonstrate real value far beyond what any tier-one now offers.

Next, consider the event of the year, the AMD-ATI merger. Talk about madness. First, the market is like, "Oh, yeah, this is the most brilliant thing since the invention of fire!" Then, a month later, everyone said, "Hey, wait, ATI's roadmap has disappeared. This new company has no idea what it's doing. The sky is falling!"

New numbers from Current Analysis show that PC sales increased 23% year-over-year for the week of Black Friday, but that's down from the 36% year-over-year growth reported the year before.

I agree to a point. Right now, the merger situation is chaos. I know of precisely one ATI PR person who survived the merger, and I sure wouldn't want his job right now. AMD's 4x4 platform (Quad FX) has not fared well against Intel's single-socket, quad-die Core 2 in initial reviews, which might explain why the Green Team isn't bashing on Kentsfield not being a "true quad chip" with the same enthusiasm we saw last summer. And as of the morning I write this in December, the first rumors of AMD's own "true quad" facing delays are floating about. The good news is that ATI's R600 GPU should ship on schedule with the Vista launch, and early buzz is that it will once again blow NVIDIA from the top of the GPU performance mountain.

Mergers suck for everyone. But I'm not worried about AMD. This is a company used to getting knocked down, and it knows from hard practice how to get up again, and it has more resources to do so now than ever before. Yeah, fourth quarter will be punishing for the chip maker, and the first quarter may be challenging, too. Yet I expect some amazing core logic to come from this new company soon, and AMD's discipline and manufacturing applied to ATI's ingenuity is going to rock the graphics world in 12 to 18 months. How AMD decides to involve the channel in these advances remains to be seen.

No, I'm more worried about Intel on this front. My gut says that Intel is going to make a play for the discrete graphics space by 2008, and Intel has never been a strong graphics player or done especially well outside of its CPU/motherboard core competencies. Remember Intel digital cameras, MP3 players, microscopes, Xscale processors, and other ghosts of Santa Clara past? Admittedly, GPUs are a lot closer to Intel's heart than MP3 players, and one could increasingly argue that the difference between GPUs and CPUs is eroding faster than bulldozed Brazilian topsoil, so Intel may have no choice but to master GPUs if it wants to maintain its market dominance.

What will this mean for NVIDIA in the intermediate future? I maintain that Intel is ultimately either going to have to buy or bury NVIDIA, and every month that goes by makes the latter case seem more likely. This uneasy alliance we have today is simply too tense and awkward. The telltale is SLI. Intel can't keep selling CrossFire technology once the D975XBX2 runs its course, and there's no reason except politics why that board couldn't have come out with SLI. Google it for yourself: Even the original D975XBX runs SLI with hacked drivers. The fact that we got the BadAxe 2 without SLI is a bad omen. But hey, even S3 came out with multi-slot graphics, so it doesn't seem far-fetched for Intel to tackle discrete graphics in 2007, especially with a design targeted somehow at businesses.

I'm also concerned that the "platform" concept took such a nosedive in 2006. AMD's LIVE! still shows several signs of being on life support (if it was ever alive to begin with), and Intel's Viiv took a 2x4 to the head and is now in a coma. In hindsight, neither of these medical emergencies is surprising. Without digital cable tuning and HD premium content, the masses will never abandon their set-tops, and I'm really worried that if AMD, Intel, and Microsoft don't get their acts together pronto on this point, Apple is going to inherit the world via iTunes and the forthcoming iTV box.

Centrino needs a big overhaul with Santa Rosa's arrival. The name Centrino still carries huge weight, but the core message of "great processor, integrated graphics, longer runtime, easy wireless connectivity" is old news. Every notebook has this now.

Centrino needs to grow. Maybe that means incorporating messages about upgradeability that start with VBI elements and embrace modular graphics designs like NVIDIA's MXM. Maybe that means building on Intel's dual-core campaign and establishing certain performance baselines for multitasking. Maybe Centrino should adopt the same chip used on Intel's virtually unknown 600SM VoIP PCI card and start building the concept of notebooks as digital communication devices. Something. Centrino was a huge success because it took existing elements that everyone wanted and packaged them together in a compelling way. We need to do that again, because right now, I don't feel the need to upgrade from my original Centrino whitebook built three years ago. And I should.

Let's throw vPro in the same boat. Resellers who even know what vPro is often tell me that they don't understand how vPro is any different than the remote management platforms they've been selling for months or years. But it is different, if only because there's more value from the brand credibility and single-vendor simplicity behind it. (And yes, there's more benefit to vPro than just that.) But the point is that both Intel and AMD stopped talking about platforms this year, and all we hear about is chips, chips, chips. This is a bad trend for the channel because everybody sells the same chips. If the message is just about Core 2 Duo, I can buy that from Dell or HP as easily as Equus or my corner system builder. But tier-ones stink at selling platforms because it's much harder to convey a total value proposition in a 30-second commercial than a list of components with a price at the bottom.

I don't know why the channel has struggled so much with platforms. If anything, I think it must be a case of wallowing in auld lang syne. Everybody understands the old ways. They used to work, and for the big brands they still work. But they're not working for the channel. We need to figure out new models to move forward with in 2007 or the consequences will be dire.

Some people say that the idea of a system builder is dead, that channel resellers are all going to be VARs. You hear this echoed by that guy who told me to just sell Acer. I disagree. I don't think resellers can thrive on services alone any more than they can thrive just on hardware sales. You need both, and it's a lot easier to do both at once when you're effectively pushing platforms.

The reason why Intel took a break from marketing platforms to the channel is because only the auld lang syne message of chips, chips, chips was going to stem the market share bleeding to AMD within a 12-month time window. Any smart business manager would have made the same decision. You see AMD already gearing up with the same approach for quad-core Opterons in mid-2007. But again, this strategy oversimplifies your value and solutions.

Auld lang syne means "old long since." For old time's sake. We've moved a long way from the old days. Had a few drinks, run some hills, picked some daisies. Put a lot of seawater between then and now. And yet...we're still here, toasting the old ways. Many of us are still living them. Old acquaintance should not be forgot; it should be learned from. But don't get stuck in it. Move on. Build. Only by devising a better sales model than the traditional one will resellers achieve real, lasting success.