TheBigFatClown
Posts: 1,110 +495
I've looked at some RX 550 graphics cards for sale on NewEgg. And some of them are described as above. "PCI-E 3.0 x16 (uses x8)". I want to understand this completely (or at least enough to not get ripped off).
From the few cards I looked at on NewEgg it appeared that there was either no difference in price between a regular x16 vs an x16(uses x8) graphics card or the x16(uses 8) was actually priced higher.
Are the 2 different types of cards being offered because the RX 550 graphics chipset is fully capable of saturating all 16 lanes under proper conditions of memory and clock speeds or is the (uses x8) version being offered to save a few pennies in manufacturing cost because no configuration of the RX 550 could possibly saturate all 16 lanes of the PCI-E 3.0 interface?
I guess my question is very similar to the question of whether the chicken or the egg came first.
Or to put it another way (this may be a different question all together), is there a way to tweak the characteristics of the graphics card itself so that an x16 graphics card's maximum theoretical bandwidth would be equal to another graphics card using only x8 PCI-E 3.0 lanes? Not "purposefully gimping it" necessarily but by way of logical tradeoffs? In other words, maybe because of criminal high prices on memory, a card manufacturer uses less of it and more of something else to save on manufacturing costs?
What was the reason for releasing x8 graphics cards into an x16 slot to begin with? I know that AMDs new AMD 2200/2400G only offers x8 lanes but this just got to market. How long have x8 graphics card been around?
Anyone who can shed any light on this issue to help me understand better is much appreciated.
From the few cards I looked at on NewEgg it appeared that there was either no difference in price between a regular x16 vs an x16(uses x8) graphics card or the x16(uses 8) was actually priced higher.
Are the 2 different types of cards being offered because the RX 550 graphics chipset is fully capable of saturating all 16 lanes under proper conditions of memory and clock speeds or is the (uses x8) version being offered to save a few pennies in manufacturing cost because no configuration of the RX 550 could possibly saturate all 16 lanes of the PCI-E 3.0 interface?
I guess my question is very similar to the question of whether the chicken or the egg came first.
Or to put it another way (this may be a different question all together), is there a way to tweak the characteristics of the graphics card itself so that an x16 graphics card's maximum theoretical bandwidth would be equal to another graphics card using only x8 PCI-E 3.0 lanes? Not "purposefully gimping it" necessarily but by way of logical tradeoffs? In other words, maybe because of criminal high prices on memory, a card manufacturer uses less of it and more of something else to save on manufacturing costs?
What was the reason for releasing x8 graphics cards into an x16 slot to begin with? I know that AMDs new AMD 2200/2400G only offers x8 lanes but this just got to market. How long have x8 graphics card been around?
Anyone who can shed any light on this issue to help me understand better is much appreciated.
Last edited:
