NVIDIA RTX 4060 Compute Benchmark Leaks: ~15% Faster than the RTX 3060 [GB6]

The first OpenCL compute benchmark of the GeForce RTX 4060 has leaked out (via Benchleaks). The budget Ada Lovelace GPU scores 105.6K points in the Geekbench 6 GPU benchmark, edging past its predecessor, the RTX 3060, by a marginal 14%. The GeForce RTX 4060 leverages the AD106 GPU, packing 3,072 shaders across 24 SMs with a boost clock of 2.6GHz. The 8GB memory buffer (18Gbps) is paired with a 128-bit bus and 24MB of L2 cache for an external bandwidth of 288GB/s.

The RTX 4060 scores 105,630 points in OpenCL, beating the RTX 3060 by 14.6% despite using fewer shaders. Much of this improvement comes from the higher core clock on the Lovelace parts. In this case, the RTX 4060 tops at 2.6GHz while the RTX 3060 is limited to just 1.84GHz. Combined with microarchitectural refinements, this 700MHz+ uplift in core frequency is sufficient to surpass its predecessor despite having only 69% of the die size.

It’s worth noting that Horizon and Edge Detection run faster on the RTX 3060, while everything else scores better on the RTX 4060. This is likely due to the larger VRAM buffer, as a quick search revealed lower scores on the 6GB variant.


Computer hardware enthusiast, PC gamer, and almost an engineer. Former co-founder of Techquila (2017-2019), a fairly successful tech outlet. Been working on Hardware Times since 2019, an outlet dedicated to computer hardware and its applications.

Related Articles

Back to top button