Intel Core i7 12700H vs AMD Ryzen 7 5800H

We compared two laptop CPUs: the 2.3 GHz Intel Core i7 12700H with 14-cores against the 3.2 GHz AMD Ryzen 7 5800H with 8-cores. On this page, you'll find out which processor has better performance in benchmarks, games and other useful information.


General overview and comparison of the processors
Single-Core Performance
Performance in single-threaded apps and benchmarks
Multi-Core Performance
Measure performance when all cores are involved
Power Efficiency
The efficiency score of power consumption
NanoReview Final Score
Generic CPU rating

Key Differences

What are the key differences between 5800H and 12700H
Advantages of Intel Core i7 12700H
  • Has 6 more physical cores
  • Newer - released 1-year later
  • Has 8192 KB larger L3 cache size
  • Newer PCI Express version – 4.0
  • 23% faster in a single-core Geekbench v6 test - 2409 vs 1964 points
  • 7% higher Turbo Boost frequency (4.7 GHz vs 4.4 GHz)
Advantages of AMD Ryzen 7 5800H
  • More modern manufacturing process – 7 versus 10 nanometers

Test in Benchmarks

Comparing the performance of CPUs across various tasks

Cinebench R23

Reputable cross-platform benchmark for high-performance processors
Cinebench R23 (Single-Core)
Cinebench R23 (Multi-Core)

GeekBench v6

The new version of this benchmark emulates common operations often used in real-world apps
Geekbench 6 (Single-Core)
Geekbench 6 (Multi-Core)
File compression 1009.1 MB/sec 739.8 MB/sec
Clang compilation 87.2 Klines/sec 61.5 Klines/sec
HTML 5 Browser 296.1 pages/sec 143.3 pages/sec
PDF Renderer 326.6 Mpixels/sec 259.8 Mpixels/sec
Text processing 237.7 pages/sec 205.6 pages/sec
Background blur 46.7 images/sec 38.2 images/sec
Photo processing 98.3 images/sec 59.6 images/sec
Ray tracing 18 Mpixels/sec 15.3 Mpixels/sec
Sources: GeekBench [1], [2]


Synthetic test that focuses on raw computational performance for low-level functions
Passmark CPU (Single-Core)
Passmark CPU (Multi-Core)
Integer math 93.1 GOps/sec 83.6 GOps/sec
Floating point math 67.1 GOps/sec 45.5 GOps/sec
Find prime numbers 98M Primes/sec 49M Primes/sec
Random string sorting 34.8M Strings/sec 29M Strings/sec
Data encryption 18.1 GBytes/sec 17.5 GBytes/sec
Data compression 309.2 MBytes/sec 277.5 MBytes/sec
Physics 1571 Frames/sec 816 Frames/sec
Extended instructions 18.4B Matrices/sec 18.5B Matrices/sec
Sources: PassMark [3], [4]


Full technical specification of Intel Core i7 12700H and AMD Ryzen 7 5800H


Vendor Intel AMD
Released January 4, 2022 January 12, 2021
Type Laptop Laptop
instruction set x86-64 x86-64
Codename Alder Lake Zen 3 (Cezanne)
Model number i7-12700H -
Integrated GPU Iris Xe Graphics (96EU) Radeon RX Vega 8


Performance Cores
P-Cores 6 8
P-Threads 12 16
Base Frequency (P) 2.3 GHz 3.2 GHz
Turbo Boost Frequency (P) 4.7 GHz 4.4 GHz
Efficient Cores
E-Cores 8 -
E-Threads 8 -
Base Frequency (E) 1.7 GHz -
Turbo Boost Frequency (E) 3.5 GHz -
Total Cores 14 8
Total Threads 20 16
Bus Frequency 100 MHz 100 MHz
Multiplier 23x 32x
L1 Cache 80K (per core) 64K (per core)
L2 Cache 1280K (per core) 512K (per core)
L3 Cache 24MB (shared) 16MB (shared)
Unlocked Multiplier No No


Fabrication process 10 nm 7 nm
TDP (PL1) 35-45 W (configurable) 35-54 W (configurable)
Max. Boost TDP (PL2) 115 W -
Socket BGA-1744 FP6
Peak temperature 100°C 105°C


Integrated Graphics Intel Iris Xe Graphics (96EU) Radeon RX Vega 8
GPU Base Clock 300 MHz -
GPU Boost Clock 1400 MHz 2000 MHz
Shading Units 768 512
TMUs 48 32
ROPs 24 8
Execution Units 96 -
TGP 15 W 15 W

Memory Support

Memory types - DDR5-4800
- DDR4-3200
- LPDDR5-5200
- LPDDR4x-4267
- DDR4-3200
- LPDDR4-4266
Memory Size 64 GB 64 GB
Max. Memory Channels 2 2
Max. Memory Bandwidth - 68.27 GB/s
ECC Support No Yes


Official site Intel Core i7 12700H official page AMD Ryzen 7 5800H official page
PCI Express Version 4.0 3.0
PCI Express Lanes 28 16

Cast your vote

Choose between two processors
929 (72.6%)
351 (27.4%)
Total votes: 1280


So which CPU will you choose: AMD Ryzen 7 5800H or Intel Core i7 12700H?
Nnkk 11 April 2023 01:06
5800H with RTX 3060 or i7 12700H with 3050ti?
+162 Reply
modal 06 July 2023 20:16
i7 12700H with RTX 3060
+31 Reply
André 15 December 2022 13:59
Go for the better gpu if you want to game. but you don't need a good CPU or gpu for ML. You can do ML with free cloud computing using tools like google collab for example.
+30 Reply
Omen 26 April 2022 18:31
Ryzen 7 5800h RTX 3070 8GB |OR| i7 12700h RTX 3060 6GB (Both with 16GB ram) Want to use pc for machine learning and a bit of casual gaming.
+222 Reply
Omen 18 May 2022 05:21
Take RTX 3070 because you need a better GPU than a better CPU for machine learning-related stuff.
+142 Reply
Noob 18 May 2022 09:40
If you want to learn ML and AI, I recommend the i7-12700h cause it has more core and thread. And you can run any game in this with medium settings or even ultra and high in some games. But anyone purchasing a laptop for Gaming then goes with AMD (gamers' choice with more power and 7 nm processor).
+61 Reply
Schokolade 08 July 2022 17:28
Noob, I believe Tensor cores are much more important for ML, so should go for the better GPU, instead of the CPU.
+46 Reply
Mr. Anon Imous 23 June 2023 17:45
Omen, If you're still reading and seeking further insights: Astrodynamics and Machine Learning engineer here. A GPU is not really "necessary". Most TensorFlow or PyTorch models for major tasks can be initially trained with CPUs before you deploy them on a cloud service like AWS or Azure. It's more efficient to use AWS or Azure than to rely on a laptop-based ML architecture. As for single-threaded performance, which is critical in my case for tasks like orbit propagations (which, if you are aware, cannot be multi-threaded), Ryzen 7/9 used to be the best. However, Intel seems to have taken the lead now, albeit being more power-hungry, with the same codebase. If you prefer working on the go, I would recommend going for Ryzen 5, 7, or 9. You won't be missing out on much, and your battery will last longer than it would with Intel (speaking from experience, as this is my line of work). Additionally, I would suggest using alternative cloud-based sources for ML, such as Google Colab with TPUs.
+26 Reply