Google Tensor
Select processor 2
Select video card 1
Select video card 2
User rating
star star star star star
Released
Q4/2021
Google Tensor

Processor Google Tensor, specifications and benchmarks

Google Tensor Processor is being reviewed, the release of which was held in Q4/2021. Here are all the Google Tensor’s specifications and performance estimation in benchmark. We offer to study all the data and compare it with a competing model. The base clock frequency of the model is 2.80 GHz, total of cores is 8 / 8. The CPU supports LPDDR5-5500 and is installed in the motherboard on Socket N/A. The TDP for CPU is 10 W.

Specifications

Technical data
CPU generation and family
  • CPU group
    Google Tensor
  • Family
    Google Tensor
  • Generation
    1
  • Name
    Google Tensor
  • Predecessor
    —
  • Segment
    Mobile
  • Successor
    —
CPU Cores and Base Frequency
  • A-Core
    2x Cortex-X1
  • B-Core
    2x Cortex-A76
  • Base Frequency
    2.8 GHz
  • C-Core
    4x Cortex-A55
  • Core architecture
    hybrid (Prime / big.LITTLE)
  • CPU Cores
    8
  • CPU Threads
    8
  • Hyperthreading / SMT
    No
  • Overclocking
    No
  • Turbo Frequency (1 Core)
    No turbo
  • Turbo Frequency (8 Cores)
    No turbo
Internal Graphics
  • DirectX Version
    12
  • Execution units
    20
  • Generation
    Vallhall 2
  • GPU Frequency
    760 MHz
  • GPU name
    ARM Mali-G78 MP20
  • GPU Turbo
    No turbo
  • Max. displays
    1
  • Max. GPU Memory
    —
  • Release date
    Q4/2021
  • Shader
    320
  • Technology
    5 nm
Hardware codec support
  • AV1
    Decode
  • H.264 / AVC
    Decode / Encode
  • h265 / HEVC (8 bit)
    Decode / Encode
  • h265 / HEVC (10 bit)
    Decode / Encode
  • JPEG
    Decode / Encode
  • VC-1
    Decode / Encode
  • VP8
    Decode / Encode
  • VP9
    Decode / Encode
Memory & PCIe
  • ECC
    No
  • Max. Memory
    12 GB
  • Memory channels
    2
  • Memory type
    LPDDR5-5500
  • PCIe
    —
Encryption
  • AES-NI
    No
Thermal Management
  • Maximum Operating Temperature
    —
  • TDP (PL1, legacy label)
    10 W
  • TDP (PL2)
    —
  • TDP down
    —
  • TDP up
    —
Technical details
  • Architecture
    —
  • Instruction set (ISA)
    ARMv8-A64 (64 bit)
  • ISA extensions
    —
  • L2-Cache
    8 MB
  • L3-Cache
    —
  • Part Number
    —
  • Release date
    Q4/2021
  • Socket
    —
  • Technology
    5 nm
  • Virtualization
    None
Positions in all rankings

Common positions Google Tensor CPU in popular benchmarks, for comparison with other models.

  • Geekbench 5, 64bit (Single-Core)
    385 place
  • Geekbench 5, 64bit (Multi-Core)
    620 place
  • iGPU - FP32 Performance (Single-precision GFLOPS)
    104 place

Benchmarks

Performance tests CPU

Compare the scores of popular benchmarks Google Tensor with other processors to understand the real speed of the CPU when processing content.

Tests include mathematical calculations, creation of 3D models, cryptocurrency mining, tests in single and multi-core modes.

These are saved results from the earlier catalog. Measurement dates and test configurations were not recorded. Models with no comparable results are omitted.

Recent models: test results with identified primary sources — Cinebench R23 · Cinebench R20 · Cinebench R15

Geekbench 5, 64bit (Single-Core)
Geekbench 5 SC is a popular cross-platform performance test for desktop or mobile processors that uses system memory intensively

Higher values are better in this metric.

1038
Google Tensor
1049
1048
Geekbench 5, 64bit (Multi-Core)
Geekbench 5 MC is a popular cross-platform performance test for desktop or mobile processors that uses system memory intensively

Higher values are better in this metric.

2810
Google Tensor
2819
2816
iGPU - FP32 Performance (Single-precision GFLOPS)
The performance of the iGPU - the internal GPU in games

Higher values are better in this metric.

1943 GFLOPS
Google Tensor
1946 GFLOPS
1946 GFLOPS

Comparable benchmark results are required to suggest performance alternatives.

*

PL1 is the sustained processor power limit; PL2 is the short-term turbo power limit. They are not measured performance or efficiency.

The values are used to select the optimal cooling system and power supply.

Rate the processor
Google Tensor

Latest comparisons