When Nvidia unveiled the new GeForce graphics cards with the Blackwell architecture in January, only four models were announced—the already released RTX 5090, 5080, and 5070 Ti, and the GeForce RTX 5070, which will be out next week. Now, information has surfaced about what comes next—the GeForce RTX 5060 Ti. This will be the first of the more affordable models (as even the GeForce RTX 5070 starts at a relatively high price).
According to WCCFtech, Nvidia is preparing two versions of the GeForce RTX 5060 Ti, with one (PG152 SKU 15) featuring 8GB of memory. The second model (PG152 SKU 10) will come with 16GB of memory. This mirrors the approach taken with the previous generation’s RTX 4060 Ti, offering a choice in memory capacity. The 16GB version will again be more expensive, though it’s unclear if the price difference will be as significant as with the RTX 4060 Ti (+100 USD).
The memory bus width is 128 bits, but both versions are expected to use GDDR7 memory with an effective clock speed of 28.0 GHz, delivering a bandwidth of 448 GB/s—over 50% higher than the RTX 4060 Ti.
It’s not yet confirmed whether this GPU will use the smaller GB206 chip (which is a native 128-bit bus die) or whether the card might instead be based on a cut-down version of the GB205 chip, which is already used in the GeForce RTX 5070. In the first case, the GPU could feature up to 4608 shaders (36 SM blocks) at most, as this is supposed to be the full specification of the GB206 chip according to earlier leaked specifications. Using a cut-down configuration of GB205 instead would open the possibility of providing the card with more units. For comparison: the RTX 4060 Ti model features the AD106 chip with 4352 shaders (34 SM blocks) active.
The card will likely rely on higher clock speeds compared to its predecessor to drive up performance, as WCCFtech reports a TDP of 180 W. This represents an increase of 20 W over the RTX 4060 Ti (or 15 W over the 16GB model), though this doesn’t seem like a major drawback in this GPU category (the RTX 5070, with a TDP of 250 W, is still significantly more power hungry).
These cards may feature a 12+4-pin power connector, but manufacturers will also be allowed to release models with 8-pin power connectors. Given the lower power consumption of graphics cards like the RTX 5070 and RTX 5060 Ti, the risk of burning out the 12V-2×6 connector (either on the card or the PSU side) should be minimal. However, one advantage of 8-pin connectors is that such cards will be fully compatible with older power supplies without the need for adapters.
Tip: Blackwell: GeForce RTX 5000 architecture and innovations [Analysis]
April release
Recently, reports have surfaced suggesting that the more affordable GeForce RTX 5000 cards may be delayed compared to the original plan to release them in Q1 2025. WCCFtech writes that the 8GB GeForce RTX 5060 will launch in April, while the RTX 5060 Ti 16GB will arrive in March.
However, this is something that VideoCardz specifically disagrees with, stating that Nvidia currently has an April release date listed for the GeForce RTX 5060 Ti in its NDA materials, with no distinction made between memory sizes. So both cards are more likely to go on sale at the same time (as with the RTX 4060 Ti) and only sometime after the start of Q2.
Sources: VideoCardz, WCCFtech
English translation and edit by Jozef Dudáš
⠀
