There is no strict standard. The difference between the mobile version of the 4850 and the desktop version of the 4850 is very small, about 12%, which is only reflected in a slightly lower frequency. For example, the mobile version of the GTX260M and the desktop version of the GXT260. The difference is relatively large. The mobile version has 96 stream processors and 192 bits.It is equivalent to the reduced frequency version of 8800GS. Its performance is only more than 60% of that of the GTX260, and there is no difference between the same model and the same model.For example, ATI's 9600 standard desktop version in 2005 is 325/440 and 4X1.The mobile version has the same parameter performance
"Technically, the frequency of ddr5 is twice that of ddr3." - In fact, it has not doubled. It's 10% faster. "Bandwidth determines performance. Bandwidth=frequency * bit width. If you add the unit conversion, it is/8/1000. For example, on the highway, frequency is the speed and bit width is the lane. This is the role of dual memory channels. When the bit width is doubled, the performance will naturally double." - This has nothing to do with dual channels. High bit width and performance are good, but it will not double. "If the actual frequency of 1g video memory is the same... and the bit width is 128... it will be one third worse..." - I don't know what to say. "In short, the graphics card looks at the bit width, the frequency, and the size and type of video memory (DDR5 or DDR3)." - The most important thing is to look at the GPU type of the graphics card, just like the GT250 is 10 times higher than the GTX460 in bit width, frequency, and video memory.
For the same model, it is conservatively estimated that the performance of the graphics card will be improved by about 20%, which is not reflected in normal use, but is more intuitive when playing large games
