NVIDIA GeForce RTX 4060 Ti CUDA Benchmark Leaks Out, Close To 15% Faster Than 3060 Ti

Hassan Mujtaba
NVIDIA GeForce RTX 4060 Ti Listed By Russian Retailers: Up To 2685 MHz Clocks & 18 Gbps GDDR6 Memory 1

The first benchmarks of NVIDIA's GeForce RTX 4060 Ti graphics cards have started to leak out in the Geekbench database.

NVIDIA GeForce RTX 4060 Ti Benchmark Leak Shows 15% Faster Performance Than The 3060 Ti

The NVIDIA GeForce RTX 4060 Ti graphics card is the upcoming mainstream solution which will be launching later this month. The graphics card is going to be a direct replacement for NVIDIA's RTX 3060 Ti with similar pricing for the 8 GB variant. The graphics card was benchmarked within CUDA tests on Geekbench 5 but before that, let's recap the specifications quickly.

Related Story Mark Zuckerberg Meets With NVIDIA’s CEO Jensen Huang, Performing An Iconic Jersey Swap

NVIDIA GeForce RTX 4060 Ti Specifications "Rumored"

The NVIDIA GeForce RTX 4060 Ti is expected to utilize the AD106-350-A1 GPU core, a cut-down version of the full AD106 graphics chip, and based on previous rumors, it should pack 34 SMs or 4352 CUDA cores, 16/8 GB GDDR6 memory running at 18 Gbps across a 128-bit bus interface, providing the card with 288 GB/s of bandwidth. There's also 32 MB of L2 cache on board the GPU which is an 8x increase over the GeForce RTX 3060 Ti.

The 16 GB variant utilizes the PG190 SKU 363 PCB while the 8 GB variant utilizes the PG190 SKU 361 PCB. The NVIDIA GeForce RTX 4060 Ti graphics card will come in various SFF and compact ITX options making them great for small PC builds. The card will also sip much less power, requiring close to 150-160W while gaming or even lower which is 25% lower than its predecessor, the RTX 3060 Ti. The cards are expected to launch by the end of May and will cost between $399 - $499 US.

NVIDIA GeForce RTX 4060 Ti Graphics Card Benchmark

So coming to the performance now, the NVIDIA GeForce RTX 4060 Ti was benchmarked on a Z790 platform with the Intel Core i9-13900K. The Geekbench 5 database confirms it as a 34 SM part with 4252 cores and a maximum frequency of 2540 MHz. The memory is also reported as 8 GB running at 18 Gbps. We can't say for sure if this is an overclocked variant or a reference model but we have seen custom variants clocking well above 2.6 GHz.

NVIDIA GeForce RTX 4060 Ti CUDA (Geekbench 5) Benchmark
Score
0
25000
50000
75000
100000
125000
150000
0
25000
50000
75000
100000
125000
150000
RTX 3070
150k
RTX 4060 Ti
146.2k
RTX 3060 Ti
128k
RTX 3060
105k

The GPU scored 146,170 points within the Geekbench 5 CUDA benchmark. The GeForce RTX 3060 Ti scores around 125-130K points on average so this makes the 4060 Ti around 15% faster but at the same time, the RTX 3070 scores close to 150K points which means that the 4060 Ti might end up a bit slower or on par than the 3070 in actual gaming performance. We can't use AMD cards in this comparison since CUDA is optimized for NVIDIA cards. Only OpenCL and Vulkan metrics can be used to compare NVIDIA and AMD cards against each other.

So with that said, it looks like NVIDIA's GeForce RTX 4060 Ti would be a small upgrade over the 3060 Ti in native raster performance with ray tracing, DLSS 3, and the added features such as AV1, etc, being the main selling points.

Do you think the 4060 Ti is a good successor to 3060 Ti if it costs the same but adds 15% more performance & additional features (DLSS 3/ RT / AV1 etc)?

News Source: Benchleaks

Share this story

Deal of the Day

Comments