Is There A Future For Enthusiasts?
Alan: I disagree. You and I have desktops, laptops, and mobile devices. But we're techies. If you had Office on an ARM device, casual users aren't going to be buying PCs anymore. If the PC market shrinks, there's going to be less economies-of-scale or R&D investment. That's ultimately going to hurt enthusiasts.
Chris: But that's like saying if the same Zelda game showed up on the 3DS you wouldn't bother with the U. There's no way I could edit spreadsheets, write fluidly, or build a PowerPoint presentation without wasting tons of time on a small ARM-based device. I maintain folks will own both before they go one way or the other.
Alan: But they already own both. The question is if they're going to buy another PC "next time" or if they're happy with what they have spreadsheet, word processing, and PowerPoint-wise. If they're happy with the HP or Dell they just bought in the last couple of years, and they're not planning on buying another PC in the future, then HP is right for jumping ship. But it still means that enthusiasts lose because we become a niche industry. In the 70s and 80s, having an audiophile setup was common. You could get great equipment for not-a-lot-of-money. Nowadays, very few people are buying pure music setups. Could desktop PCs go the way of the compact discs to make room for thin clients, Quadro-in-the-cloud, etc?
Chris: So it sounds like you’re anticipating a bad situation for enthusiasts, no matter which way the wind blows? The thing is, we’ve been hearing that for years. And yet, AMD and Nvidia derive their entire product lineups from high-end GPUs. Intel’s CPUs are fast and overclockable enough to stay ahead of multi-GPU configurations, SSDs are simultaneously quicker and more affordable than ever, and you can find $600 motherboards if you really want such a thing. Just as there’s a demand for sports cars from the guys who can afford them, so too will there remain a market for high-margin, low-volume enthusiast gear. The key to success is taking that Formula 1 technology and using it to improve Honda-class components.
Alan: Funny that you used Honda as your example. Honda's F1 team ended when the company decided it couldn't support the team financially. Now look at what's happened to the NSX replacement.
Chris: Another unfortunate consequence of the economic down-turn. But hey, at least there's still rumored to be a successor in the works.
Alan: Most people think that HP hasn't done much for enthusiasts since the Articooler, and even then, HP detractors will say that it was really Agilent that decided to sell the cooler to consumers. Still, that set the stage for countless Thermaltake and Zalman coolers which, only now, are bested by Thermalright and Noctua-style heatsink designs.
But I think HP's experiments with VoodooDNA and Tom Szolyga's and Rahul Sood's concept-designs-as-reality still pushed enthusiast-level engineering, some of which we haven't seen trickle down to things that individual system builders can access.
The pedestal-based Blackbird still demonstrates solid system cooling principles and I think Rahul Sood still uses that chassis for his personally-built system, since it was ATX-compliant. You can actually see some of that HP engineering expertise in its current workstation line. The Z600 and Z800 workstations have the function-is-form design that comes from the BMW DesignWorks design, but the HP PSU design is something pretty awesome that replicates the old SGI Octane chassis, where the PSU spans the length of the case and gets its own supply of fresh air from the front.
The also short-lived FireBird lineup brought us a 350 W external PSU with a reasonably powerful (at the time) Core 2 Quad and GeForce 9800S.
In 2011, I'd love to build a small system like the FireBird. Three hundred and fifty watts doesn't sound like a lot, but it's plenty for a Core i5-2500K or i7-2600K, a Radeon HD 6870 or GeForce GTX 560 Ti, and an SSD or two. I'm not sure where I can find a cheap, reliable external PSU with that kind of power.
Chris: That "tier-one advantage" comes from the resources to build unique enclosures that wrap around hardware in a specific way, which is what the FireBird embodied. Incidentally, this is something becoming more available to the do-it-yourself crowd in 2011. Most people missed it, but the Intel DH61AG motherboard is actually intended to drop into empty all-in-one shells from companies like Gigabyte, ECS, and Mitac. We're not even close to achieving tier-one parity in that regard (with the powerful, external PSU, even), but you can go out and build a mainstream all-in-one with HD Graphics 3000, if that really tickles your fancy. Here's the thing: it takes a push from a company as big as Intel to set the gears into motion. You need an organization large enough to jump-start ODM production, to set up distribution, and to promote reseller awareness. That's not an issue when you're HP or Dell.
Alan: HP was much more aggressive than Dell in trying to push the envelope of PC power supplies. If you think about it, our PSUs really haven't changed in form factor since the 80286 era. Sure, we've gone from AT, ATX, ATX12V, EPS, etc. But PC power supplies are almost always a box at the back of the chassis with a bunch of wires that travel everywhere. Guys like SGI and HP were daring enough to try to develop better PSUs. If you look at what they did, it was all ultimately to improve airflow to the PSU and to the main motherboard compartment itself. SGI actually went further with its Octane, putting the GPU is a fully separate compartment, something that not even Thermaltake's BMW Level 10 replicates. SGI did this in 1997. On the enthusiast level, we have awesome power supplies like the Seasonic X series, but the best we've been able to do is sleeve the cables to limit turbulence and to have modular PSUs so that we can take out unused cables.
Both the Firebird and Blackbird were liquid-cooled. So are the HP Z400 and Z800 workstations. All of these systems were liquid-cooled by Asetek.There's no question that HP's investment in Asetek made a difference. Sure it was Asetek's engineers who did the heavy lifting, but with HP as its very first OEM design win in 2007, the company saw its venture capital doubled over the next two years. Now we have Corsair- and Antec-badged Asetek CPU coolers that go head-to-head against the best Thermalright and Noctua heatsink/fan combos.
People often think that the MacBook Air was the first product from a major corporation to push the idea of a SSD mainstream in January 2008. In fact, a week earlier, HP started building business desktop PCs with SSDs. I think SSDs would have caught on without HP, but it's worth noting that HP was among the first to recognize the potential.
HP rarely engaged in the high-end enthusiast market, but when it did, deeper pockets helped fund ideas that ultimately benefited enthusiasts. As the Honda/Acura NSX, Ford GT, Mitsubishi Evolution, Subaru WRX STi and Nissan GT-R have all proven, a company does not need to have an exclusive focus on enthusiast-products to do something special. Good engineering is good engineering.
Five years from now, we'll probably see the return of high-end case and power supplies being sold together as a bundle. What will change is that we will see the "new innovation" of having a non-standard ATX power supply, spanning the entire length of the chassis with dedicated airflow. When this comes out, remember that SGI was doing this in 1997 and HP was doing it with its Z workstations since 2009.
R&D by big corporations like HP affects system-building enthusiasts in ways that people cannot immediately see. If HP pulls out of the market, the major innovator with deep pockets is gone.
When HP becomes a pure software and services company, how much will it still invest in HP Labs? How much of the engineering resource pool will the spun-off or bought-out HP PC company have without access to the riches of the software and services division? That's the worry that I have when I hear that HP is backing out. Unlike IBM/Lenovo, there aren't a lot of buyers who can make the HP acquisition something worthwhile. I think Todd Bradley, one Mark Hurd's first recruits, is lobbying to become that buyer by getting HP to spin off the company.