The GeForce GTX 1050 is said to come with 640 CUDA cores, which by far makes its the weakest member of the Pascal family. The base clock is dialed in at 1354 MHz, while the boost clock goes up to 1455 MHz. When all the numbers are tallied, the GeForce GTX 1050 will have FP32 compute performance of 1.8 TFLOPs.
The GeForce GTX 1050 Ti, on the other hand, has 768 CUDA cores along with base and boost clocks of 1290 MHz and 1382 MHz respectively. FP32 compute performance is said to come in at 2.1 TFLOPs.
Of course, these clock speeds are likely to change between now and when the cards are released, and we already seen minor variations in clock speeds from different sources. Both cards feature a 128-bit memory bus, don’t require PCIe power connectors and have a TDP of 75 watts.
We’re assuming that the GeForce GTX 1050 series is probably nearing the lower limit when it comes to the Pascal architecture, as we’re not sure how far NVIDIA is willing to compromise performance in order to reach lower price points.
Both the GeForce GTX 1050 and the GeForce GTX 1050 Ti are expected to bow sometime next month, priced at $119 and $149 respectively.