Snapdragon 8 Elite (Gen 4) vs Tensor G5

We compared two 8-core processors: Qualcomm Snapdragon 8 Elite (Gen 4) (with Adreno 830 graphics) and Google Tensor G5 (PowerVR DXT-48-1536). Here you will find the pros and cons of each chip, technical specs, and comprehensive tests in benchmarks, like AnTuTu and Geekbench.

Review

General comparison of performance, power consumption, and other indicators
CPU Performance
Single and multi-core processor tests
Gaming Performance
GPU performance in games and OpenCL/Vulkan
Battery life
Efficiency of battery consumption
NanoReview Score
Overall chip score

Key Differences

Main differences and advantages of each chip
Pros of Qualcomm Snapdragon 8 Elite (Gen 4)
  • Performs 2x better in floating-point computations
  • Shows significantly better (up to 2x) AnTuTu 11 score – 3126K vs 1549K
  • Supports 24% higher memory bandwidth (84.8 against 68.2 GB/s)
  • 14% higher CPU clock speed (4320 vs 3780 MHz)
Pros of Google Tensor G5
  • Better instruction set architecture

Benchmarks

Performance tests in popular benchmarks

AnTuTu 11

The AnTuTu Benchmark measures CPU, GPU, RAM, and I/O performance in different scenarios
1549798
CPU 893474 734567
GPU 1133808 183228
Memory 427210 222379
UX 671910 409624
Total score 3126402 1549798
Samples [541 tests] [16 tests]

GeekBench 6

The GeekBench test shows raw single-threaded and multithreaded CPU performance
Single-Core Score
Multi-Core Score
Asset compression 367.8 MB/sec 281.8 MB/sec
HTML 5 Browser 223.5 pages/sec 162.5 pages/sec
PDF Renderer 257 Mpixels/sec 207.4 Mpixels/sec
Image detection 234.4 images/sec 149.5 images/sec
HDR 315.8 Mpixels/sec 151.4 Mpixels/sec
Background blur 32.7 images/sec 24.4 images/sec
Photo processing 90.6 images/sec 46.5 images/sec
Ray tracing 12.4 Mpixels/sec 6.79 Mpixels/sec
Samples 148 tests 8 tests
Compute Score (GPU)
Background blur 45.3 images/sec 4.77 images/sec
Face detection 21.6 images/sec 2.41 images/sec
Horizon detection 0.73 Gpixels/sec 0.17 Gpixels/sec
Edge detection 0.91 Gpixels/sec 0.27 Gpixels/sec
Gaussian blur 1.75 Gpixels/sec 0.2 Gpixels/sec
Feature matching 0.33 Gpixels/sec 0.05 Gpixels/sec
Stereo matching 96.8 Gpixels/sec 12.6 Gpixels/sec
Particle physics 3848 FPS 634.9 FPS
API Vulkan Vulkan
Samples 100 tests 7 tests

3DMark

A cross-platform benchmark that assesses graphics performance in Vulkan (Metal)
3DMark Steel Nomad Light
Score 2386 1012
Stability - 96%
Graphics test 18 FPS 7 FPS
3DMark Solar Bay
Score 10874 -
Graphics test 41 FPS -
Section 1 47 FPS -
Section 2 41 FPS -
Section 3 35 FPS -
3DMark Wild Life Extreme
Score 5664 -
Stability 51% -
Graphics test 33 FPS -

PCMark 3.0

Benchmark for evaluating overall system performance in everyday workloads
PCMark 3.0
Web score 21945 17237
Video editing score 8458 7203
Data manipulation score 16086 14036
Photo editing score 39118 24584
Writing score 22647 20213
Phones tested 36 models 4 models

Recent User Tests

The latest benchmark tests that have been submitted by users
Qualcomm Snapdragon 8 Elite (Gen 4)
DateBenchmarkResult
πŸ“˜ 2026-09-25 (Asuna)AnTuTu 113203192
πŸ“˜ 2026-09-25 (Ramirez)AnTuTu 123484313
πŸ“˜ 2026-09-24 -> MaximumAnTuTu 123347504
πŸ“˜ 2026-09-24 (Icecream)AnTuTu 123297029
πŸ“˜ 2026-09-24 -> iannnnnAnTuTu 123255862
Google Tensor G5
DateBenchmarkResult
πŸ“˜ 2026-09-25 (Dolbaeb)AnTuTu 121549159
πŸ“˜ 2026-08-22 -> PixeluserAnTuTu 111536042
πŸ“˜ 2026-08-02 -> bangAnTuTu 111719659
πŸ“˜ 2026-08-02 -> bangCPU Throttle74%
πŸ“˜ 2026-07-31 -> bangGeekbench 6 GPU4116

Specifications

Full list of technical specifications of Snapdragon 8 Elite (Gen 4) and Tensor G5

CPU

Architecture 2x 4.32 GHz – Oryon (Phoenix L)
6x 3.53 GHz – Oryon (Phoenix M)
1x 3.78 GHz – Cortex-X4
5x 3.05 GHz – Cortex-A725
2x 2.25 GHz – Cortex-A520
Cores 8 8
Frequency 4320 MHz 3780 MHz
Instruction set ARMv8-A ARMv9.2-A
L1 cache 192 KB -
L2 cache 12 MB -
L3 cache 8 MB 8 MB
Process 3 nanometers 3 nanometers
TDP (Sustained Power Limit) 8 W -
Manufacturing TSMC TSMC

Graphics

GPU name Adreno 830 PowerVR DXT-48-1536
Architecture Adreno 800 -
GPU frequency 1100 MHz 1100 MHz
Pipelines 3 6
Shading units 512 128
Total shaders 1536 768
FLOPS 3379.2 Gigaflops 1689.6 Gigaflops
Vulkan version 1.3 1.4
OpenCL version 3.0 3.0
DirectX version 12.1 -

AI Accelerator

Neural processor (NPU) Hexagon Yes

Memory

Memory type LPDDR5X LPDDR5X
Memory frequency 5300 MHz 4266 MHz
Bus 4x 16 Bit 4x 16 Bit
Max bandwidth 84.8 Gb/s 68.2 Gb/s
Max size 24 GB 16 GB

Multimedia (ISP)

Storage type UFS 4.0, UFS 4.1 UFS 3.1, UFS 4.0
Max display resolution 3840 x 2540 3840 x 2400
Max camera resolution 1x 320MP 1x 200MP
Video capture 8K at 60FPS, 4K at 120FPS 8K at 30FPS, 4K at 120FPS
Video playback 8K at 60FPS, 4K at 120FPS 8K at 30FPS, 4K at 120FPS
Video codecs - H.264
- H.265
- AV1
- VP8
- VP9
- H.264
- H.265
- AV1
- VP9
Audio codecs - AAC
- AIFF
- CAF
- MP3
- MP4
- WAV
- AAC
- AIFF
- CAF
- MP3
- MP4
- WAV

Connectivity

Modem Snapdragon X80 Exynos 5400c
4G support LTE Cat. 24 -
5G support Yes Yes
Download speed Up to 10000 Mbps -
Upload speed Up to 3500 Mbps -
Wi-Fi 7 7
Bluetooth 6.0 6.0
Navigation GPS, GLONASS, Beidou, Galileo, QZSS, NAVIC GPS, GLONASS, Beidou, Galileo, QZSS

Info

Announced October 2024 August 2025
Class Flagship Flagship
Model number SM8750-AB GS501
Official page Qualcomm Snapdragon 8 Elite (Gen 4) official site -
Successor - Qualcomm Snapdragon 8 Elite Gen 5 -
Further details
Notes on Snapdragon 8 Elite (Gen 4):
    - The SM8750-AC model has a higher frequencyβ€”4,470 MHz for performance cores versus 4,320 MHz in the SM8750-AB and SM8750-3-AB modelsβ€”and its GPU is 100 MHz faster.

Cast your vote

So, which SoC would you choose?
564 (83.2%)
114 (16.8%)
Total votes: 678

Comments

Please give your opinion on the comparison of Tensor G5 and Snapdragon 8 Elite (Gen 4), or ask any questions
🌐 Register your profile and become part of NanoReview community!
Avatar
HalasKYO July 28, 2026 at 11:55 AM
I don't know, as the pixel are much less score, in any test even in AI which supposedly should be much better then competititon But in camera they are unbeatable, and reliable, in camera performance (not like Honor) I would have choosen the HOnor Magic 8 Pro against the Pixel 10 pro XL, as it has better battery, much better soc, camera specs are better (exept the wide angle) but the sofware the winng recipe on the Pixel I will buy 10 pro XL soon or wait for the 11... no games tho... but hard user... with a lot of camera needs.
0 Reply
Avatar
Mamataro December 22, 2025 at 04:33 AM
They say Pixel phones are great for cameras but not for gaming. But actually, you need powerful, fast hardware not just for gaming, but also for editing and rendering 360, 1080p, and high-bitrate 4K videos. If rendering video is going to take hours anyway, I might as well just choose the Redmi Note 14 4G version (SD 685). The rendering time is about the same (hours), but the price is 40x cheaper! 😜🀭🀭
+4 Reply
Avatar
Q.C August 30, 2025 at 07:49 AM
The Tensor G5 is trash.
+33 Reply
Avatar
swapnil August 20, 2025 at 10:20 AM
let the processor be launched 1st before comparing with the speculative features.
+6 Reply
Avatar
babayaga August 27, 2025 at 08:04 PM
did not changed anything. its still shit.
+8 Reply
Avatar
Enthusiast January 1, 2025 at 04:40 AM
First of all Google's upcoming 2025 Tensor G5 is supposed to be compared to Qualcomm's upcoming 2025 snapdragon 8 Gen 5. Anyways as you can see on the performance benchmarks specs. even when compared to 8 Gen 4, the Tensor G5 still gets badly outperformed. That said, the upcoming 8 Gen 5 is guaranteed to destroy Tensor G5 in performance benchmarks by an even greater margin. To sum it up, Google's ditching Qualcomm in 2021 for their in-house Tensor chip is the worse thing they could have done which is why Pixel's 6-9 have been far below the premium level despite charging a premium price.
+21 Reply
Avatar
Pixel January 13, 2025 at 08:27 AM
After seeing OnePlus 13.. I wish I wish Pixel 10 Pro XL gets that's beast 8 elite chip 😭. It's so sad! Tensor is just curse now...
+13 Reply
Avatar
Adnan March 24, 2025 at 05:16 PM
Google doesn't want to rely on Qualcomm chip . They're from usa and they want to create their own hardware and software. Even with tensor g3 is enough for me for daily needs. We don't need a beast chip because most of people don't use Google for gaming . Just 7 years of update quick system and best camera.
+4 Reply
Avatar
TG5 March 26, 2025 at 12:23 AM
Adnan, That midrange chip is not gonna last for 7 years, heck not even for 3 years. Choosing shitty AI over Speed or raw performance is absolutely pathetic, that is the reason no one buys garbage Pixel's and will remain shit until their performance matches Qualcomm...
+21 Reply
Avatar
Best Camera??? May 16, 2025 at 01:56 PM
Adnan, Your Pixel can't even stitch pictures properly in panorama mode. LOL
+3 Reply
Avatar
Ygg July 15, 2025 at 10:54 PM
Best Camera???, Do you know how to produce phone lol πŸ˜‚
+3 Reply
Avatar
darknight799 December 3, 2025 at 08:16 AM
TG5, I know people who are still using pixel 6a and pixel 7 pro. Both are still smooth and getting better with updates.. also cameras, speakers, haptics on those are very good compared to most nee midrange phones. If you dont game and just want smooth phone with nice cameras, pixel will be fine for 4-5 years easily. Dont get pixel if you are mobail gamer. As simple as that
0 Reply
Avatar
Dr mujeeb hamid December 4, 2024 at 11:51 AM
No of efficiency cores : A17 pro - 4 A18 pro - 4 8 elite - 6 8 gen 3 - 7 (5+2) Dimensity 9400 - 4 Dimensity 9300 - 4 Tensor g5 - 7 (5+2) (Dimensity 9400 & 9300 have a difference of only 1 large core, everything else is same, yet Dimensity 9400 is 15 % faster & scores 25 % more efficient upto 12w) Geekbench Scores of above chips @ 3w : D9300 : 3600 8 gen 3 : 3925 A17 pro : 4140 D9400 : 4400 8 elite : 4475 A18 pro :4500 Scores @ 11w : 8 gen 3 : 7250 D9400 : 8600 A18 pro : 9000 8 elite : 9400 Scores of processors on Samsung foundry @ 3w: Tensor g3 - 2500 (also 4400 @ 9w) Exynos 2400 - 2700 (also 7100 @ 12w) Expected scores of tensor G5 : Scenerio 1 tsmc 4nm (less likely )- 3900 @ 3w & 7400 @ 12w Scenerio 2 tsmc 3nm ( more likely)- 4600 @ 3w & 7600 @ 8.5w PS : 3w score matters more. It deduces the overall efficiency of a chip and its sustained performance. Eg, Exynos 2400 & 8 gen 3 score almost same (7200) @ 11w yet in real world it is 33 % less efficient than 8 gen 3. Wonder why ? Because it is 33% less efficient than 8 gen 3 @ 3w. Similarly, the multicore score of tensor g5 i.e, 7600-8000 might seem less than 8 elite (9500-10500), but it will actually be slightly more efficient than it @ 3w. Infact it will be even more efficient than A18 pro @ 3w. So u can expect same level of performance from tensor g5 as 8 elite, Dimensity 9400 & a18 pro. (The 2 ultra small cores on tensor g5 will help it to run less demanding apps on them, so it'll help it to achieve more SOT than other 3 processors at a same battery capacity).
+26 Reply
Avatar
Halo January 4, 2025 at 12:11 PM
Thanks for the comparison
+2 Reply
Avatar
Pixel January 13, 2025 at 08:32 AM
Are just speculating? If that happens to be true, G5 can be the best chip then and will be a game changer.
+5 Reply
Avatar
TG5 January 14, 2025 at 06:59 PM
Not possible! Tensor G5 is using last years Cortex X4 as main core. It is not going to match snapdragon 8 Elite. Stop with your false narratives!
+8 Reply
Avatar
Facts April 4, 2025 at 06:47 PM
TG5, Depends in battery the person is correct if google optimises the apps the g5 could keep up and maybe even surpass the 8 elite in real performance but not in games ofc
+3 Reply
Avatar
linux August 26, 2025 at 06:06 PM
yeah, I'm thinking that the pixel WON
0 Reply
Avatar
babayaga August 27, 2025 at 08:06 PM
linux, they didnt won anything. benchmarks show its same performance for cpu with a iphone 14 pro. 3 year old phone this month. gpu is even worse tho. around iphone 13-13 pro level
+2 Reply
Avatar
anonymous September 7, 2025 at 02:49 AM
My friend, performance and efficiency cores are VERY different. The 8 Elite has 0 efficiency cores and all big cores (6 performance, 2 prime).
+3 Reply
Avatar
Joe February 3, 2026 at 11:00 PM
anonymous, The interesting thing is that Qualcomm wins in battery life, despite having few or no efficiency cores. Qualcomm is doing a great job optimizing those performance cores to provide efficient performance in certain situations.
+1 Reply