Blog Post

Qualcomm adds to a mobile chip fragmentation issue

You may not care who makes the processing chip in your smartphone or tablet yet, but if chip makers have their way, you will in the future.

Qualcomm(s qcom) announced on Wednesday that it is expanding its Snapdragon GamePack to more than 100 mobile apps, some of which are exclusive to devices running on Qualcomm Snapdragon chips. Consumers can learn about GamePack apps in Qualcomm’s new Snapdragon GameCommand, a central place to find them.

Both the new games and the central location to find them will appear in 2012, says Qualcomm, although there are already some Android games available today in the GamePack, which launched in June. These software titles are specifically optimized to take advantage of the graphics capability in Qualcomm’s new Snapdragon chips, now available in four performance brands called S1, S2, S3 and S4, the latter being the most advanced.

Forgive me for coming down harshly here, but do we really need more fragmentation in the market for Android devices? I think not. Between Google pushing out new versions of Android quickly and carriers being slow to forward them to consumer devices, the Android market is already made up of a wide range of devices with differing capabilities. How exactly does the release of chip-specific, exclusive games help with the problem?

The fact is, it doesn’t, and it sets the precedent for even more component fragmentation in the future. Imagine if certain Android applications only ran on touchscreen displays made by Samsung and not those from LG, for example? Heck, most consumers don’t know — or care — who made the display, or chip, for that matter, on their device. In another industry, say for automobiles, this situation would be akin to choosing a car that can only drive on certain roads if it has a particular engine.

Qualcomm isn’t alone on this path: Nvidia (s nvda), one of Qualcomm’s largest competitors, offers similar software and a storefront for Tegra-optimized games called TegraZone. I didn’t like this situation when Nvidia announced it earlier this year, and I don’t like it any more today.

At one point, some Samsung Galaxy S 2 handsets were reportedly running on a Tegra chip while others weren’t, allowing for the same consumer handset not being able to run a consistent set of apps. True, non-optimized versions of the same apps were available with fewer game levels or features, but that doesn’t solve the potential problem.

Ironically, Qualcomm is holding an analyst day on Wednesday, when it will continue to push out news and talk about its chip portfolio. One such future product that could be running on Qualcomm chips is a new smart TV, according to CNet. Let’s hope a Snapdragon TV doesn’t have access to exclusive channels and connected apps.

6 Responses to “Qualcomm adds to a mobile chip fragmentation issue”

  1. curdriceaurora

    Let’s say QCOM decides to ship one chip that is available to a plethora of Android OEMs and everyone makes only the fastest smartphones. Surely as time progresses, the OEM design teams would like to add some customizations based on their own plans. Do they look elsewhere or do they talk to QCOM about each of their customizations? Instead of spending resources on each OEM’s customization, it is best to turn the product portfolio into one that has feature based hierarchy. Once you include mid-level smartphones into the picture, obviously you would see more levels in the said hierarchy.

    Not everyone needs the fastest phone. Not every OEM has the same vision about those phones. Doing what they are, QCOM is making sure it becomes the first go to manufacturer when a phone’s design is being chalked up.

  2. I don’t think you have much cause to worry about Gamepack harming the industry.
    Just look at how much of a failure Tegrazone has been.

    Nvidia brought tegrazone to the marketplace because the Tegra2 did not support standard SIMD instruction sets, and their benchmarks suffered because of it. Tegrazone was their way of convincing manufacturers that those benchmarks didn’t matter. If that was the entire purpose of Tegrazone, then it worked… they were able to get Tegra2 processors into a large amount of devices.

    However, if you look at the service itself, there are still only 23 games available through tegrazone, some of which are available in the android market to all devices anyway. It is for all intents and purposes a complete failure as a service. Such a failure in fact that nVidia has put a NEON decoder that supports SIMD into the Tegra 3.

    Qualcomm isn’t in the same situation as nVidia was when it started Tegrazone, so really I’m not sure what they hope to gain with GamePack. Maybe they simply see Tegrazone as a “feature” and have feature envy?

    The fact remains that content creators like standards even more than consumers, so I have little doubt that GamePack will be just as big of a failure as Tegrazone.
    Of course if they’re just looking for checkbox on a comparison sheet, then they probably don’t even care if it fails.

  3. who cares

    You are not forgiven. If the computer industry had followed in the same steps you seem to want the cell phone industry to follow, we would still be held hostage to one (or two) chip designs. I even doubt the graphic card industry would be where it is at now. COMPETITION is always good. Stop thinking otherwise. Fragmentation is bad, OBVIOUSLY, but the best will survive.You really really want only one android phone, from one company? WTF?!

      • As long as there are performance differences, the brand/model of the chips will sometimes matter, and 3D performance can vary drastically.

        The question is: do these games require a Snapdragon (more specfically, the Adreno GPU) or are they simply coded to take to make the most out of Snapdragons’ GPU performance profile.

        Mobile 3D is recapitulating the desktop PC 3D experience of the late 90s.

      • Fair points, Sam. As I understand it, the games are coded to take advantage of the Adreno and such benefits from that coding are initially exclusive. I like your mobile 3D point in regards to the desktop 3D scenario in the ’90’s. I remember that time well and didn’t like it. ;)