When rumors of the GeForce GTX 1660 Ti began to spread, there was also speculation that lower-end GeForce GTX 1660 and GTX 1650 models were also in the works. While there is no official confirmation from Nvidia on the GeForce GTX 1660 yet, VideoCardz's industry sources have confirmed that Nvidia is currently preparing the GeForce GTX 1650.
(Image credit: Tom's Hardware)
The GeForce GTX 1650 is purportedly the direct successor to the current GeForce GTX 1050. Just like its other GTX 16-series brethren, the GeForce GTX 1650 is built around Nvidia's Turing architecture, but lacks the real-time ray tracing feature. It's rumored to utilize the smaller TU117 silicon, which is also produced by Taiwanese foundry TSMC with its 12nm production process.
So far, there haven't been any leaks surrounding the GeForce GTX 1650. The graphics card's specifications are anybody's guess at this point, but the consensus is that the GeForce GTX 1650 could come equipped with 896 or 1,024 CUDA cores. According to TUM_APISAK's latest tweet, the graphics card purportedly runs with a 1,485MHz base clock, which is only 15MHz slower than the GeForce GTX 1660 Ti.
The GeForce GTX 1650 might carry up to 4GB of GDDR5 memory. It's feasible that the memory could operate at 2,000MHz (8,000MHz effective) across a 128-bit memory interface. We'll have to wait for more leaks to confirm. Assuming that Nvidia can maintain same TDP (thermal design rating) as the previous GeForce GTX 1050, the GeForce GTX 1650 could have a 75W TDP as well and draw its power directly from the PCIe slot without the need for external PCIe power connectors.
A HardOCP article from last month claims that Nvidia will release the GeForce GTX 1650 in late March at a price point of $179 (~£137.05). The GeForce GTX 1050 launched at $109, so it'll be interesting to see whether the GeForce GTX 1650's level of performance justifies the higher asking price.
Zhiye Liu is a news editor, memory reviewer, and SSD tester at Tom’s Hardware. Although he loves everything that’s hardware, he has a soft spot for CPUs, GPUs, and RAM.
