Intel's Atom: AMD Plays Defense
Intel's Atom: AMD Plays Defense
You might have heard of the recent NVIDIA-VIA alliance, in that both are working together to offer a US$45 platform to combat Intel's own Celeron and Intel 945G combo. So far not many details have surfaced on this platform, but it's probably going to be targeted at very low-cost compact desktop systems (likely Nettops) and in the Netbook segment. Of course in this same segment is Intel's new Atom processors that will span the UMPC, netbook and nettop product groups as well.
Now here's where AMD doesn't exactly have a ready-made platform option to offer the industry. While it has low-cost and low-TDP processors like Semprons and the Turion processor, they are not yet low-power enough to fit these segmentations. This is where AMD's previously disclosed Bobcat platform will cater to these needs, an x86 class chip that suits devices more powerful than PDAs, but less powerful than full-fledged notebooks - which is exactly where UMPCs and netbooks position themselves. This was made known in 2007 and AMD brings to light that Bobcat that will most likely come into play in 2009. At this late juncture, it's hard not to dismiss the possibility that Bobcat might have a Fusion-like core for greater integration and power reduction.
While several of these mini notebook vendors like HP, ASUS and more are ramping up their netbook class systems, AMD begs to differ that this doesn't yet signal a strong demand for such systems. At least till early this year, many analysts too were quite divided and uncertain if this category would be in critical demand. Look at it this way - there are high-end handhelds like the Nokia Communicator series or the multimedia oriented N95 phones as well as other heavyweights like the HTC TyTN II that have great PIM functionality (including e-mail), have simple office suites and have speedy Internet connectivity. Best of all, these fit in your pocket. With products like these, the only few advantages netbooks have is their screen size, larger keyboard and better processing power for better productivity. They are however not much cheaper than full-fledged notebooks that are larger and more comfortable to use but lose out on the ultra-light and portability factor. The point John drove here is that netbooks don't solve a critical need/desire that traditional notebooks or handhelds offer at the moment. In a way, netbooks have sort of become a sought after niche segment, but not yet large enough a demand to make it a mainstream volume consideration for all vendors.
Of course we can also conversely say that AMD doesn't yet have a solution to this area (the AMD Geode family isn't powerful enough as seen from some Kohjinsha netbooks), thus AMD's current stance of this growing segment. When asked to comment on AMD's position against the Intel Atom processors, this is what John has to say at the moment:-
"They (Intel) may be able to deliver a lot of these Atom-based systems, but what is the total system experience/architecture such that Intel can compete and deliver a really rich experience such that the device/system has a chance to compete and not feel like a highly compromised device. We think this is where we have the advantage. The AMD device would always look better and deliver better performance than other competitive devices."
Probably true for the next class of options such as ultra value notebooks, but with AMD out of an immediate solution to the Netbook class of products, they might have to sit out and watch at the sidelines for sometime. By the time Bobcat arrives to compete in this space, they might have difficulty penetrating this segment as Intel's expecting Moorestown (successor to the current Intel Centrino Atom or Menlow platform) to be available in 2009. At the moment, Moorestown is expected to be made up of an SoC chip (Lincroft, that integrates the 45nm Silverthorne core with graphics, video and memory controller on a single chip) along with an I/O hub (Langwell). Where AMD might have a chance to succeed is if Bobcat debuts as Platform on a Chip (PoC) solution and/or have significantly better graphics capabilities. So this will be an interesting face-off next year. For now, AMD feels that there's not enough demand to carve a solution for this segment immediately.
As for Mobile Internet Devices (MIDs), which is related to Intel's Centrino Atom platform initiative, AMD's approaching this segment with a wait-and-see attitude. They clearly see room for smartphones with interesting things happening in this space (which also includes the advancements made in their own Imageon series offerings) and we have to agree that this segment is flourishing. There hasn't been any prototype or demo substantial or persuasive enough for us to consider MIDs a serious segment. It seems like yet another product that's vying for your bag/pocket space, but doesn't address any particularly clear market segment that existing smartphones, PMPs or other handheld devices already take care of.
The next segment above this is the ultra value notebooks and desktops and is one of AMD's favorite. It's very obvious when you check any retail outlets because at the low price points, you'll most often find an AMD-powered machine or the Intel's equivalent doesn't technically offer enough merit as AMD does. So for the near-term, we see AMD holding this area strong as it's able to offer solutions at low cost with reasonable performance as well (something that the Celeron series has never been as competitive in).