NVIDIA set to confuse the hell out of people by changing its model naming.
In what seems like a lifetime ago, AMD made a brave move and changed the naming systems for its processors. Branding by megahertz was the industry way, with fights for milestones like 1GHz a big deal. However, that all changed in response to the Pentium 4.
Because the Pentium 4 didn’t have that linear relationship with megahertz that the Pentium 3 did, AMD initially enjoyed the headlines about the Athlon punching above its weight, with 1.2GHz CPUs smashing higher speed Intel offerings. Enthusiasts knew that megahertz meant very little, but as the popularity of AMD’s products grew, the problem of lower numbers started to hurt with the undereducated public.
By the time of the Athlon 64, the relationship between core speed and performance was so out of alignment that AMD made the brave (and much derided) move of abandoning megahertz for a new system called Performance Ratings. These PR numbers were theoretically based on an extension of the megahertz numbering of the original Athlon. For example, a 3200+ Athlon 64 CPU should deliver the same ‘speed’ as a vanilla Athlon 3200+ would.
In reality, these ratings were more about equivalency with Intel’s persistence on the megahertz scheme and an attempt to provide a comparable range of numbers. A legitimate attempt to level the playing field for the end consumer.
It worked, despite the critics. So much so that Intel had to come in and ruin it all a few years later by introducing its own number scheme, which slapped several levels of obfuscation onto the whole thing. Trying to grasp the difference between the two brands returned to an exercise in futility that only really settled when the massive leap that was Core 2 effectively pushed AMD into insignificance in the retail PC market.
This is an example of a system that, for a while, worked. The graphics market spiralled out of control at a much quicker rate when it came to naming, with major shifts happening in every new generation of product, leading to complete and utter confusion for anyone not totally up on developments.
Say 9800 around me and I don’t think NVIDIA, I think of the RADEON 9800. These cards were several generations apart and light years different in performance, yet, for some bizarre reason, they share a number. It has been even worse in the realm of suffixes, with LE, GT, GTS, GTX, HD, and Ultra all meaning different things at different times. For a while, GT was ATI’s suffix for low end cards, yet it has been used by NVIDIA for its lower high end models.
Perhaps worst of all is the horrible tendency to rename half of the previous generation to fit with the new system, something it appears NVIDIA is about to do with its graphics chips. After jumping from the 9800-type naming on the GeForce 9 series, VR Zone has evidence from early drivers that this naming is going to move to one more similar to the GT 280 naming on its GeForce 10 series of cards.
This would be an attempt to simplify the lineup going forward, but it perpetuates this horrible situation we have in the hardware world where names really do mean nothing. Instead of being a convenient means of assisting our decisions, they are just the labels we search for when we try and work out how things are supposed to actually match up.
Supporting our habits:
Post new comment anonymously (Anonymous Gibbon)