HiSilicon Kirin 985 5G vs Unisoc Tanggula T760 5G

When comparing the HiSilicon Kirin 985 5G and the Unisoc Tanggula T760 5G processors, several key specifications should be taken into consideration.

In terms of CPU cores and architecture, the HiSilicon Kirin 985 5G boasts a more diverse setup. It includes 1x 2.58 GHz Cortex-A76 core, 3x 2.4 GHz Cortex-A76 cores, and 4x 1.84 GHz Cortex-A55 cores. On the other hand, the Unisoc Tanggula T760 5G features 4x 2.2 GHz Cortex-A76 cores and 4x 1.8 GHz Cortex-A55 cores.

Both processors have 8 cores in total, but the Kirin 985 5G has a higher clock speed on its highest-performing Cortex-A76 core. This could potentially result in better performance under heavy workloads for the Kirin processor.

In terms of instruction sets, both processors support ARMv8.2-A, ensuring compatibility with the latest software and optimizations.

There is a slight difference in lithography between the two processors. The Kirin 985 5G utilizes a 7 nm process, while the Tanggula T760 5G employs a 6 nm process. A smaller lithography generally translates to better power efficiency and potentially improved performance.

In terms of power consumption, the Tanggula T760 5G has a lower thermal design power (TDP) of 5 watts compared to the 6 watts of the Kirin 985 5G. A lower TDP indicates that the Unisoc processor may consume less power and generate less heat.

Lastly, when comparing neural processing capabilities, the Kirin 985 5G utilizes the Ascend D110 Lite and Ascend D100 Tiny technologies based on the HUAWEI Da Vinci Architecture. The Tanggula T760 5G, on the other hand, features its own Neural Processing Unit (NPU). The specific features and performance of these NPUs would need to be compared separately to assess their effectiveness.

In conclusion, the HiSilicon Kirin 985 5G processor excels in terms of its CPU core diversity and potentially higher performance under heavy workloads. The Unisoc Tanggula T760 5G, on the other hand, offers a slightly smaller lithography and a lower TDP, indicating better power efficiency. The specific neural processing capabilities of each processor would require further investigation to determine their respective advantages.