TECH NEWS – It’s not enough that there’s a Ti edition of the 4070; Nvidia is also preparing something similar one level down…
With the correct pricing, the GeForce RTX 4060 Ti could be one of the company’s mainstream video cards: mid-range, not too outstanding in performance, but seems to be a product fit for this purpose based on the rumors. Its computing power could be 22 TFLOPS, according to Videocardz. Insider kopite7kimi leaked the card’s specs, but now TechPowerUp’s GPU database editor TFC Fantasy has posted the card’s clock speeds (which Videocardz based their calculations on). The base clock speed of the card will be 2310 MHz and 2535 MHz in boost mode, but premium models (depending on the manufacturer) can be bumped up to 2685 MHz, which will bring the computing performance up to around 24 TFLOPS.
The GeForce RTX 4060 Ti (plus, reportedly, the regular 4060 model) is based on the AD106 GPU. The 4060 Ti will use the AD106-350-A1 GPU, a cut-down AD106. 34 SMs, 4352 CUDA cores, and 8 GB of GDDR6 VRAM at 18 Gbps on a 128-bit interface so the bandwidth will be 288 GB/s. The card has 32 MB of L2 cache, eight times more than the predecessor 3060 Ti. According to leaked images, like the 4070, the 4060 Ti will be powered via the dividing 12VHPWR PCIe Gen5 connector, as Nvidia is looking to standardize it across its product line. With a more compact PCB design, its power consumption won’t be too high either: it can have 20% less TGP (160W) than the RTX 360 Ti.
If it is indeed the specification of Nvidia’s GeForce RTX 4060 Ti, it will be able to compete with the RTX 3070 Ti. That card launched for $600, so the company could be expected to hit the stores for around $400, as the 3060 Ti also had the same price tag. Except knowing Nvidia, it will be more like $450-500.
The RTX 4060 Ti might be good for 1080p, but the 128-bit bus and GDDR6X instead of GDDR6 would cause problems at higher resolutions…
Leave a Reply