Google Tensor G4: Unveiling Its Geekbench Debut and Surprising CPU Configuration

Earlier reports suggested that Google would take an approach, with the Tensor G4 aiming for it to be a modified version of the Tensor G3. The plan is for it to have the GPU. Updated CPU cores. This system on a chip (SoC) is expected to power the Pixel 9 Pixel 9 Pro and Pixel Fold 2 which are scheduled for release this year. Recently a purported Geekbench listing has surfaced, providing some insight into its specifications.

All About the Google Tensor G4

On Geekbench, the Tensor G4 has made its appearance alongside a device codenamed “Google Tokay.” It’s not clear yet which of the three aforementioned devices this SoC is intended for. However, since there is always a possibility of manipulation in Geekbench listings there could be doubts about its authenticity.

Upon examination of the Geekbench listing details it seems likely to be genuine. The listing mentions an Arm Mali G715 GPU and an octa-core configuration with one Cortex X4/X5 core clocked at 3.1 GHz three Cortex A7xx cores at 2.6 GHz, four Cortex A5xx cores at 1.95 GHz along with 8 GB of RAM.

At a glance, these specifications might seem puzzling as they depict the Tensor G4 as an eight-core chip compared to its predecessor, the Tensor G3 which had nine CPU cores. Typically, in upgraded components like these, we expect an increase in core count, rather than a reduction.

Additionally, the performance of the Tensor G4 (1,082 in single-core and 3,121 in multi-core tests) is slightly lower compared to the Tensor G3 (1,268/3,510) according to Geekbench scores. It is worth noting that this discrepancy may be due, to the fact that the tested Tensor G4 model is a release sample.

Previously disclosed information indicated that the Tensor G4 would be manufactured using Samsung’s 4LPP+ node, the same one utilized by the Exynos 2400. Thus far, the Exynos SoC has demonstrated competitiveness with flagship offerings from Qualcomm and MediaTek, raising curiosity about the Tensor G4’s performance. Those anticipating a more substantial performance boost may need to await the Tensor G5, purportedly developed entirely by Google on an unspecified TSMC node.