June 5, 2023

SHSU Houstonian Online

Read all latest news headlines from USA, UK and around the world, get today's breaking news and live updates on politics, elections, business, sports, economy,​ …

Nvidia GeForce RTX 4060 Ti - The video card is suffering from shrinkage

Nvidia GeForce RTX 4060 Ti – The video card is suffering from shrinkage

The GeForce x60 series from Nvidia has always attracted a lot of attention. This class usually consists of video cards at a price that appeals to a much larger group of PC gamers, and thus offers a good performance-to-performance ratio. With the RTX 4060 Ti, the price is higher than its predecessors and the card comes out at a time when many are wondering if 8GB of video memory is still enough.

The RTX 4060 Ti has been added by Nvidia to the RTX 40 series series based on the Ada Lovelace architecture. While the RTX 4070 was dedicated entirely to the 1440p chip by Nvidia, the RTX 4060 Ti is tasked with doing the same for 1080p gaming. With this, after a long and gradual introduction of the RTX 40 series in the form of the RTX 4090, 4080, 4070 Ti, and 4070, attention is finally being paid to screen resolutions that are still widely used with just as much resolution. Popular video card of the x60 class. This generation has a hefty price tag, as the RTX 4060 Ti has a suggested retail price of €449 for the 8GB model. A 16GB version that will follow later will cost an additional €110.

The basis for the GeForce RTX 4060 Ti is the AD106 GPU, which is baked at 4nm by TSMC. This GPU contains 22.9 billion transistors on a surface of just 190 mm². These are significant differences with its predecessor, while the number of cores is somewhat less. Aside from the difference in the production process, it also comes into play here that the RTX 4060 Ti has a much larger cache memory than the RTX 3060 Ti, which has gone from 4 to 32MB.

See also  Canonical launches free version of Ubuntu Pro for small businesses - Computer - News

Following AMD with its RX 6000 series, Nvidia is also using a much larger cache memory than this generation to boost performance while cutting other specs at the same time. For example, the memory bus has been halved to 128 bits, bringing the slightly faster GDDR6 memory to 288 GB/s. That’s a big drawback, but thanks to the larger on-chip memory buffer, this shouldn’t cause problems right away. One consequence is that the video card can use power more efficiently. On the RTX 4060 Ti, tgp 160W is on the low side for modern video cards, despite significantly increased clock speeds.

video card RTX 4060 Ti RTX 3060 Ti RTX 2060
GPU m 106 GA104 TU106
building Ada Lovelace Ampere Turing
practical 4nm TSMC 8 nm Samsung 12 nm TSMC
die size 190 mm² 392.5 mm² 445 mm²
transistors 22.9 billion 17.4 billion 10.8 billion
CUDA cores 4352 4864 1920
Tensioner cores 136 152 240
RT cores 34 38 30
texture units 136 152 120
robs 48 80 48
base clock 2310MHz 1410MHz 1365MHz
clock increase 2535MHz 1665MHz 1680MHz
L2 cache 32 MB 4 MB 2 megabytes
memory speed 18 Gb/s 14 Gb/s 14 Gb/s
memory bus 128 bits 256 bits 192 bits
bandwidth 288 GB / s 448 GB / s 336 GB / s
Tgp 160 watts 200 watts 160 watts
(In the foreground)
449 euros (8 GB version) 419 euros 369 euros
released May 23, 2023 December 2, 2020 January 7, 2019