With the introduction of the GeForce4 Ti and MX line of video cards, NVIDIA has followed the tradition started with the GeForce3 Ti line of cards. That is, bombarding the marketplace with six different versions of their new GeForce4, to which NVIDIA hopes to produce a card in the budget and performance needs of every potential buyer out there.
However, to supply the more popular and more profitable value segments of the market a vendor may often clock their product under its true capability; this is especially true when they are getting such good yields on a chip they have no choice but to remark it as a slower one. This is common with processors, and is much the same with 3D video chips. If you can remember back to the launch of the GeForce3 Ti500 and Ti200, a very high percentage of the Ti200s produced could overclock to match the core speed of the Ti500, and with a little luck these could even match the memory speed as long as the vendor did not skimp on the memory used.
The caused the GeForce3 Ti 200 to become a very popular card, since it not only filled the market need for a lower-cost GeForce3 variant, but it gave overclockers the chance to get a lot more value for their gaming buck. This last part is a key element to this article, since the new GeForce4 Ti and MX cards may also hold some extra MHz in reserve.
Overclocking a graphics card can be just as significant as overclocking a processor. While a 20 MHz core speed jump on a graphics card may not catch your eye like a 300 MHz overclock with a Pentium 4 or Athlon XP, the additional 20 MHz can go a long way. This is especially true if your processor is fast and the video portion is the limiting factor when it comes to high-end game framerates.
Although there are hobbyists who overclock for pure fun, the prime motivation is value. Why buy the more expensive version of the hardware when a less expensive, and lower clocked version of the same hardware may well overclock to the same speeds as a higher-end, and many times more expensive, model. Through the years, many examples of this come to mind, such as the infamous Celeron 300A, Pentium III 600E and the current overclocking rage over the Pentium 4 1.6A chips. Even general users are buying these new Pentium 4 chips like hot cakes and taking them all the way up to 2.1GHz with no problem.
Some of the lower-end models of the GeForce4 line, most apparently the GeForce4 MX 440 and the Ti4400, have the best chance of being underclocked, and not reaching their fullest potential right out of the box. The GeForce4 Ti 4400 is simply a lower clocked version of the GeForce4 Ti 4600 core, but with 3.6ns DDR SDRAM instead of the Ti 4600's 2.8ns DDR SDRAM. The same goes for the GeForce4 MX 440, which has the added benefit of many models running at their standard core speeds even without the aid of active cooling.