There are many reasons why there is currently no new batch platform with iGPU on your motherboard, but the market changes, needs change, and businesses are trying to get the best value for money and resources. What happened to prevent this type of product from being built today?
Graphics cards embedded in motherboards, extinct for 10 years
The truth is that the last batches of this type of product were made at the end of 2012, but the end of its trajectory was conceived long before, on 2009, where Intel changed its business strategy in its processors and began to include a basic graphics card in these processors.
The objective was clear, to allow it to differentiate itself from the competition by offering a more complete product and at the same time to obtain simpler and cheaper motherboards for its manufacturers. Moving the iGPU to the processor simplified the connections of said board, you didn’t have to build in VRAM or take the one from the system and move it via a direct bus to it, lowered the cost for the board. user on the card and in addition off it got higher performance being much closer to cores with direct access and lower latency with PCIe controller and RAM.
Although AMD went this route two years and a bit later calling this type of product with integrated CPU + GPU APU, the truth is that Intel with the same approach is still talking about CPU as such, we assume that for marketing reasons.
Why aren’t they made if there are more GPU manufacturers?
For many reasons. The first is of course licensing, as AMD doesn’t allow GPUs outside of succulent and volume contracts for specific products. Intel doesn’t license outside of its CPUs either, and NVIDIA takes a different path and never really took an interest in this industry, although it did launch a few models at the time, but in the minority.
What about GOING THROUGH
Therefore, it is the industry itself that is currently rejecting this type of product for a more optimal product like APUs, which have a very important development which, in the case of Intel, affects the whole range of large processors. public and input, which AMD also seems to be doing in its next CPUs, leaving the HEDT and its specific matrices.