700 logos in index Logo index | Add new logo | Add to favorites

Centrino logo

Centrino mobile technology first saw the light of day on March 12, 2003 . In its original form , the platform's 1st generation featured a WLAN module based on the 802.11b standard. The L2 cache of the original Pentium M (Banias) was only 1 MB in size, and worked with two chipsets: the Mobile 855GM and the Mobile 855PM. The Mobile 855GM had an integrated graphics core and supported memory with a peak bandwidth of up to 2100 MB/s (DDR266). The Mobile 855PM, hardly changed since March of 2003, needed its own dedicated AGP graphics chip.

In the final version of Carmel, the chipset with integrated graphics is called Mobile 855GME and supports DDR333 memory with a peak bandwidth of up to 2700 MB/s. The graphics core of the Mobile 855GME differs from that of the Mobile 855GM mainly in its higher maximum core frequency, and resulting better performance. Like the chipset, the CPU also got a makeover in May of 2004. Since then, the Pentium M has gone by the name of Dothan; it is made using a 90nm manufacturing process and now has a 2 MB L2 cache. Unfortunately, instead of having plain old clock speeds to go by, users must guess which CPU model they require based on relatively obscure processor numbers.

Innovations In The Mobile 915 Express Chipset Family
The official name of the new chipset generation for notebooks is itself an indication that one of the most important innovations here is the introduction of PCIe. In notebooks of the future, there will no longer be any graphics subsystems tied to the system via an AGP port. Instead, the new chipset family will only support PCI Express connections. In addition, GPU producers ATi and Nvidia will move to outfitting both current and future graphics processors with only one interface, for cost reasons if no others. The user simply will not have a choice between AGP and PCIe.

Whether PCI Express is absolutely necessary in a notebook is still unclear. In terms of performance, there's as little reason for urgency here as there is for desktop PCs. The reason is that there's still not one 3D application that can profit measurably, not to mention tangibly, from the bandwidth differential that PCIe offers over AGP.

If you examine the PCIe interfaces in terms of energy requirements, a theoretical advantage does appear compared to AGP, because with PCIe individual lanes can be switched off to save energy. This advantage, however, is negated in other areas. Compared to earlier AGP-connected GPUs, the current graphics chips' power consumption has increased considerably. Both ATi and Nvidia are packing their processors with more and more transistors, functional units, pixel pipelines and share units, all of which need lots of power and high clock speeds for core and memory. Even if the graphics card manufacturers do their homework in terms of power management, they will have only limited success in countering the trend of growing power requirements caused by both higher clock speeds and an ever-increasing number of transistors.
(http://www.tomsguide.com/us/sonoma,review-380-3.html) www.intel.com