Intel: Technology Leader, or Technology Copycat?

I recently decided to buy myself a new system after my hardworking IDEQ200N barebones with unlocked AthlonXP 2600+ died on me so I thought it would be a good time to jump onto the DDR2 bandwagon. Coincidently this happened amidst all the noise about Conroe and AM2 so obviously decisions had to be made whether I would finally jump to Intel or stick with AMD. Interestingly I choose to go for AM2 even though Conroe’s performances has been times and over again been dubbed stellar across the online comunitites. Could it be because I am a little bit of an AMD fanboy? Well, maybe, but not just blind devotion without reasons.

Intel is a giant compared to AMD. In terms of manufacturing capabilities and raw financial strength, Intel is superior. But being a giant also is what turns Intel to be the Goliath it is now. Despite claiming to be industry and technology ‘leader’ how many times has it been that Intel criticized AMD’s technology decisions but later red facedly tailing after with similar technology? How many times before had we read about Intel denying the need to move to 64-bit technology? How about bashing the advantages of integrated memory controller? How about AMD’s Torrenza initiative that caught Intel off guard?

Core2 processors, Duo and Quad may dominate for now but they are still on age old saturated FSB technology that has almost reached its limits. Despite this, Intel still refused integrated memory controller on the grounds that ‘memory standard changes’ and ‘it’s better to have larger cache on the processor itself’. This claim to ‘flexibility’ turns out to be not flexible at all. Currently Intel’s processors use the same socket LGA775 but it is very hard to determine what processor is compatible with what motherboard simply because of the different MCH chipsets used. This was evident when Conroe debuted a few months back. The only reason for MCH is that it costs one more chip for customers to use an Intel system thus bringing more money into Intel. Comparatively, socket AM3 processors in the future are guaranteed to work with current socket AM2 boards due to the integrated memory controller thus proving that it is more flexible for the memory controller to be integrated. The much anticipated CSI bus (successor to the old age FSB) is a similar point-to-point bus as HyperTransport that AMD had been using for years and even that is expected only by 2008 in Intel’s Tukwila chips in servers. Chances are it will also finally features integrated memory controller for Intel but who knows when it’s going to reach us laymen end-users, your guess as good as mine. EM64T is nothing more than a copy of AMD64’s instruction sets yet they dare to call it their own. At IDF recently Intel announced Geneseo as an answer to AMD’s Torrenza, opening up FSB and PCI-express bus for developers to build on, which I personally agree with the Inquirer’s article that is a waste of good R&D funds. Logically, why would people want to build on a soon-to-be-obsolete FSB when CSI is on the horizon? Then with recent AMD’s acquisition of ATI, suddenly news flew around that Intel is jumping into CPU-GPU integration. Not to mention they are actively hiring graphic engineers while firing HR and marketing personnel all over the world. Intel also will be shipping their quad cores ahead of AMD, even if that means slapping 2 dual cores on the same package. Guess who’s going to ship native quad cores first? Now, who’s exactly the technology ‘leader’?

AMD offers customers and users with choices. With AMD customers have the freedom to choose components, suiting the best of each and everyone’s needs. Use VIA’s components, or NVIDIA’s, ATI’s, Broadcom’s, use whoever’s chipsets, graphics, wireless chips, use whatever suits you, that’s the freedom of choice AMD brings. And AMD, being small, has technology partnerships with many other companies for example IBM giving them the advantage of innovation through open standards. Even HyperTransport is governed by a consortium. Sure, they might be behind in terms of process technology, but who cares if the processor is built on 65nm or 90nm if they perform just about the same? Do not forget that AMD64 with inferior process technology and only DDR out beat Pentium4 with DDR2 for years before Conroe came about.

With Intel, users have to stick with the so called ‘platforms’ to get the best performance. It always has to be Intel’s processor, Intel’s chipsets, Intel’s motherboard, Intel’s radio and soon, Intel’s graphics. While they argue that integrated solutions from one company benefits users in terms of cost, I have yet to see any Intel platforms that are priced cheaper than any competing solutions. Just look at VIIV which in fact, is a huge marketing scheme aimed at non-competent users. I attended a VIIV demo here recently and it was disappointing! The bloody thing costs 5 figures in my country, and my system at home can do exactly what VIIV can do for just a little over 3k! Not only has it cost a bomb, it also shuts the door for independent system builders like myself for VIIV certifications. Only OEMs can produce a VIIV compatible system, stick the VIIV sticker and guess who will pay for the overpriced sticker? Check out this commentary on the Inquirer on VIIV, my thoughts exactly.

Intel may have all the money in the world for R&D but AMD is more flexible and creative, and most of the time more future proof. What is the problem with Intel then? In my opinion, arrogance and pride lead to bad technological decisions. If only they will be humble enough to listen to what customers and people actually want instead of making decisions for them, I think they would be in a better position (and avoid lawsuits, of course). Technology needs to be open to create innovations and to be enjoyed by all, by not doing so Intel is dooming itself to be the Goliath.