-
The Start of Geforce
The Gpu had a core clock of 120 mhz & a memory clock of 166 mhz. it was made on a 220 nm process with 128 bit bus that supported 32mb of VRAM -
Tesla is 2nd to the Fam
It supported DX 10 on ddr3 600mb of VRAM on the first Geforce 8800 GTS with a 90nm process with a 320 bus size. clocked at 516 with a memory clock of 1600mhz. also support SLI
more multiple gpus -
Fermi takes over with 480
the gtx 480 used an even smaller 40nm process for more transistors. 3 billion in total. clocked at 700 with a Memory clock of 3696 mhz much improved over older gen. bus size of 384 bit using the lasted GDDR5 ram 1.5gbs of it. -
Nividia makes the double gpu with the 590
the 590 had 2 580s which use the Fermi architecture both gpus had 40nm for a combined 80nm process & combined 6 billion transitors. 2 384 bit buses but not a combined VRAM staying at a mear 1.5 gbs -
Kepler soon to follow
Kepler came in under a year with a disapointment becuase half the cards were rebrands anyway. but 680 and 690 were something else. the 690 was basiclly 2 680s. the 680 had a super small 28 nm process. 3.5 billion transitor had a core clock of a 1 ghz a new type of boost clock up to 1150 mhz but still mear 2 gbs of VRAM
