NVIDIA is following up on the GeForce GT 220 with yet another affordable 40nm process technology based graphics solution - the GeForce GT 240. Positioned by NVIDIA as a low-cost graphics solution that lie in between the GeForce 9800 GT and 9600 GT, it has 96 CUDA cores (as NVIDIA likes to call them now), which is actually pretty decent, considering it is just 16 cores short of the 9800 GT. Other crucial tech specs include 32 texture mapping units and 16 raster operating units, which are pretty much in-line with other cards of its caliber.

We have also learnt that the GeForce GT 240 will come in either 512MB or 1GB flavors, and vendors have the choice of choosing between GDDR3 and GDDR5 memory to equip the cards.

In terms of clock speeds, the GeForce GT 240 in ’stock’ configuration will come clocked at 550MHz at the core, 1340MHz at the shaders and depending on the memory type equipped, 2000MHz DDR for GDDR3 variants and 3400MHz DDR for GDDR5 ones.

Despite the rather healthy specifications, don’t for a moment let them fool you into thinking that this is a mainstream gaming card, because they are signs that point towards the new GeForce GT 240 being a basic SKU. For one, it doesn’t require a PCIe power connector for power. And secondly, it lacks an SLI connector for future performance boost options. These are some of the ‘features’ commonly noted on less powerful cards and sadly the GT 240 is amongst them.

Fortunately, to sweeten the deal, the new GeForce GT 240 brings about full PhysX and CUDA support, and is also DirectX 10.1 compatible. So it’s at least a tad more updated than its GeForce 9 brothers, but not yet DirectX 11 compliant for which we’ll have to wait for the Fermi architecture based offspring sometime next year.

Popularity: 57% [?]