Apple M3 Pro

Apple M3 Pro
Apple M3 Pro processor review

Apple M3 Pro: New Cores and Graphics with Lower Memory Bandwidth

Apple M3 Pro launched in fall 2023 alongside the M3 and M3 Max. The higher-end configuration received 12 CPU cores, an 18-core GPU, and support for up to 36 GB of unified memory. Compared with the M2 Pro, single-threaded performance increased substantially, while the GPU moved to a new architecture with Dynamic Caching, mesh shading, and hardware ray tracing.

At the same time, Apple changed the CPU balance. The M2 Pro had eight performance cores and four efficiency cores, whereas the higher-end M3 Pro has six of each. This reduction in the number of performance cores limited the Multi-Core gains. The reduction in memory bandwidth from 200 to 150 GB/s was a separate change and has a greater impact on memory- and GPU-intensive workloads.

Two Versions of the M3 Pro

The M3 Pro was released in two main configurations.

The lower-end version has 11 CPU cores-five performance and six efficiency cores-along with a 14-core GPU. It was used in the base 14-inch MacBook Pro with M3 Pro.

The full version has 12 CPU cores in a 6P + 6E configuration and an 18-core GPU. It was installed in the 16-inch MacBook Pro and offered in higher-end configurations of the 14-inch model.

Both versions support up to 36 GB of unified memory.

CPU Performance

In Geekbench 6, the full M3 Pro scores around 3,100 points in Single-Core and 15,500 in Multi-Core. Compared with the 12-core M2 Pro, single-threaded performance increased by approximately 17-18%, while the Multi-Core advantage is around 6%.

For comparability, the table uses results from specific MacBook Pro models tested by Notebookcheck.

Device Geekbench 6 Single Geekbench 6 Multi Cinebench 2024 Multi Memory Launch Price
MacBook Pro 14 M2 Pro, 12-core 2,663 14,568 1,030 16 GB $2,499
MacBook Pro 14 M3 Pro, 11-core 3,113 14,412 908 18 GB $1,999
MacBook Pro 16 M3 Pro, 12-core 3,138 15,480 1,059 18 GB $2,499
MacBook Pro 14 M4 Pro, 12-core 3,844 20,458 1,403 24 GB $1,999

The prices refer to new computers at the time the corresponding generation was released.

The lower-end 11-core M3 Pro is approximately 7% behind the full version in Geekbench 6 Multi-Core and 14% behind it in Cinebench 2024 Multi-Core. There is practically no difference in Single-Core.

The full M3 Pro is only slightly faster than the M2 Pro when all cores are loaded, but it is noticeably ahead in Single-Core. The M4 Pro shows a different scale of improvement: the base 12-core version is approximately one-third faster than the full M3 Pro in Geekbench 6 Multi-Core and Cinebench 2024 Multi-Core.

Power Efficiency

The 6P + 6E configuration is accompanied by lower CPU power consumption under full load. In Notebookcheck tests, the full M3 Pro consumes around 26-27 W, while the M2 Pro can approach 34 W.

At the same time, the full M3 Pro is slightly faster than its predecessor in Cinebench R23 Multi. When performance and power consumption are compared for this workload, its performance-per-watt advantage is approximately 30%.

For extended compilation, CPU rendering, and other highly parallel workloads, the absolute gain over the M2 Pro is small. The generation’s main improvements here are higher single-core speed and lower power consumption under full load.

M3 Pro Graphics

The highest-end M3 Pro has 18 GPU cores, compared with 19 in the M2 Pro. The lower-end configuration has a 14-core GPU. The M3 generation also added Dynamic Caching, mesh shading, and hardware ray tracing.

Device GPU 3DMark Wild Life Extreme Unlimited Blender Classroom Metal Launch Price
MacBook Pro 14 M2 Pro 19 cores 12,983 97 s $2,499
MacBook Pro 14 M3 Pro 14 cores 12,094 73 s $1,999
MacBook Pro 16 M3 Pro 18 cores 14,545 69 s $2,499

In Wild Life Extreme, which does not use hardware ray tracing, the 18-core M3 Pro is approximately 12% ahead of the M2 Pro. The lower-end 14-core M3 Pro, by contrast, is approximately 7% slower than the 19-core M2 Pro.

The picture changes in Blender. Cycles on the M3 can use hardware ray tracing through MetalRT, so even the lower-end 14-core M3 Pro completes Classroom in 73 seconds, compared with 97 seconds for the M2 Pro. The full version takes 69 seconds-approximately 1.4 times faster than the M2 Pro.

Thus, the 14-core M3 Pro is noticeably behind the higher-end version in conventional raster workloads, but the gap between them is small in rendering. The GPU’s new hardware capabilities allow the M3 Pro to pull further ahead of the M2 Pro specifically in workloads that can use them.

Memory

The M3 Pro uses a 192-bit unified-memory bus with 150 GB/s of bandwidth. By comparison, the M2 Pro had a 256-bit bus and 200 GB/s.

With this memory configuration, the M3 Pro is offered in capacities unusual for previous Macs: 18 and 36 GB instead of 16 and 32 GB. The maximum capacity increased, but bandwidth simultaneously fell by 25%.

This change is barely noticeable in typical CPU workloads. The difference can be more pronounced in GPU computations and other workloads that constantly access large amounts of data.

Media Engine

In terms of capabilities, the M3 Pro’s media engine is close to the base M3’s. It supports hardware decoding and encoding of H.264, HEVC, and ProRes, ProRes RAW and AV1 decoding, as well as one video encode engine and one ProRes encode/decode engine.

The larger M3 Max differs by having two video encode engines and two ProRes encode/decode engines. Therefore, for particularly demanding video editing, the Max’s advantage is not limited to the CPU and GPU.

M3 Pro in 2026

The main reference point for the M3 Pro today is the M4 Pro. The base 14-inch MacBook Pro with M4 Pro launched at the same $1,999 price as the M3 Pro a year earlier, but it came with 24 GB of memory instead of 18 GB and a significantly faster 12-core CPU.

Therefore, the M3 Pro MacBook Pro is primarily interesting when available at a substantial discount. This is especially true of the lower-end configuration with 11 CPU cores and 14 GPU cores: with only a small price difference, the base M4 Pro is significantly stronger in both CPU performance and memory capacity.

For an M2 Pro owner, upgrading to the M3 Pro also provides limited gains in multi-core CPU workloads. The move from an M1 Pro or weaker Apple Silicon chip is more noticeable.

Conclusion

Apple M3 Pro received significantly faster CPU cores and a more efficient graphics architecture, but changing the CPU configuration to 6P + 6E limited the Multi-Core gains over the M2 Pro.

In graphics, the result depends on the workload. In the conventional Wild Life Extreme test, the lower-end 14-core M3 Pro is actually slower than the 19-core M2 Pro, whereas in Blender, hardware ray tracing allows both M3 Pro versions to be noticeably faster.

In 2026, the M3 Pro should primarily be evaluated based on the price of the specific MacBook Pro. With only a small price difference versus the M4 Pro, the newer generation is significantly more attractive, while the M3 Pro becomes more appealing with a substantial discount.

CPU Comparison

Compare Apple M3 Pro with other CPUs

Please select CPU from the dropdown list.

Basic

Label Name
Apple
Platform
Laptop
Launch Date
October 2023
CPU Architecture
Apple M3 performance cores + efficiency cores
CPU Name
Apple M3 Pro
Model Name
The Intel processor number is just one of several factors - along with processor brand, system configurations, and system-level benchmarks - to be considered when choosing the right processor for your computing needs.
Apple M3 Pro
Foundry
TSMC
Generation
Apple M3 series

CPU Specifications

Total Cores
Cores is a hardware term that describes the number of independent central processing units in a single computing component (die or chip).
12
Total Threads
Where applicable, Intel® Hyper-Threading Technology is only available on Performance-cores.
12
Performance-cores
6
Efficient-cores
6
Efficient-core Max Turbo Frequency
Maximum E-core turbo frequency derived from Intel® Turbo Boost Technology.
2.75 GHz
Instruction Set Extensions
NEON
Performance-core Max Turbo Frequency
Maximum P-core turbo frequency derived from Intel® Turbo Boost Technology.
4.06 GHz
Extended Instruction Set
ARMv8-A, NEON
L1 Cache
3 MB total L1 cache
L2 Cache
20 MB L2 cache
L3 Cache
12 MB system level cache
Technology
Lithography refers to the semiconductor technology used to manufacture an integrated circuit, and is reported in nanometer (nm), indicative of the size of features built on the semiconductor.
3 nm
Instruction Set
The instruction set is a hard program stored inside the CPU that guides and optimizes CPU operations. With these instruction sets, the CPU can run more efficiently. There are many manufacturers that design CPUs, which results in different instruction sets, such as the 8086 instruction set for the Intel camp and the RISC instruction set for the ARM camp. x86, ARM v8, and MIPS are all codes for instruction sets. Instruction sets can be extended; for example, x86 added 64-bit support to create x86-64. Manufacturers developing CPUs that are compatible with a certain instruction set need authorization from the instruction set patent holder. A typical example is Intel authorizing AMD, enabling the latter to develop CPUs compatible with the x86 instruction set.
ARMv8-A
Transistor Count
37 billion

Memory Specifications

Memory Bus Width
192-bit
Memory Type
Intel® processors come in four different types: Single Channel, Dual Channel, Triple Channel, and Flex Mode. Maximum supported memory speed may be lower when populating multiple DIMMs per channel on products that support multiple memory channels.
Unified LPDDR5-6400
LPDDR5 Speed
LPDDR5-6400
Max Memory Size
Max memory size refers to the maximum memory capacity supported by the processor.
36 GB
Max Memory Bandwidth
Max Memory bandwidth is the maximum rate at which data can be read from or stored into a semiconductor memory by the processor (in GB/s).
150 GB/s
Maximum Memory Speed
6400 MT/s

GPU Specifications

External Display Standard
Thunderbolt 4, HDMI 2.1
GPU Name
Apple M3 Pro GPU
Max External Display Resolution
Up to two 6K 60Hz displays over Thunderbolt; or one 6K 60Hz display plus one 4K 144Hz display over HDMI; HDMI up to 8K 60Hz or 4K 240Hz
Video Concurrency
Multiple streams of 4K and 8K ProRes video; AV1 decode
Video Decode
H.264, HEVC, ProRes, ProRes RAW, AV1 decode
Video Encode
H.264, HEVC, ProRes, ProRes RAW
Video Processing Unit
Apple media engine with ProRes acceleration and AV1 decode
GPU Max Dynamic Frequency
1296 MHz
Graphics Core Count
18
Number of Displays Supported
Up to 2 external displays
GPU APIs
Metal, OpenCL
Graphics Performance
Up to 4.6 TFLOPS FP32
Media Engine
Hardware-accelerated H.264, HEVC, ProRes, ProRes RAW, and AV1 decode
Video Decode Engines
1
Video Encode Engines
1
ProRes Encode/Decode Engines
1
OpenCL Support
OpenCL (Open Computing Language) is a multi-platform API (Application Programming Interface) for heterogeneous parallel programming.
Yes

AI Specifications

AI Engine
16-core Apple Neural Engine
Neural Engine Core Count
16
NPU Name
Apple Neural Engine
NPU Performance
18 TOPS

Connectivity

Bluetooth Support
Yes
Bluetooth Version
Bluetooth 5.3
Wi-Fi Standard
Wi-Fi 6E (802.11ax)

Interfaces and Ports

Thunderbolt Support
Yes, Thunderbolt 4 up to 40 Gb/s
USB Version
USB4
USB4 Support
Yes, USB4 up to 40 Gb/s

Miscellaneous

Hardware-Verified Secure Boot
Yes, Apple silicon secure boot chain of trust
Runtime Anti-Exploitation
Kernel Integrity Protection, Pointer Authentication Codes, Fast Permission Restrictions

Benchmarks

Cinebench R23
Single Core Score
1928
Cinebench R23
Multi Core Score
15061
Geekbench 6
Single Core Score
3100
Geekbench 6
Multi Core Score
15263
Geekbench 5
Single Core Score
2314
Geekbench 5
Multi Core Score
15235
Passmark CPU
Single Core Score
4240
Passmark CPU
Multi Core Score
23993
Cinebench 2024
Single Core Score
142
Cinebench 2024
Multi Core Score
1059
Cinebench 2024 GPU
Score
5961
Blender
Score
228

Compared to Other CPU

Cinebench R23 Single Core
2634 +36.6%
2042 +5.9%
1928
1674 -13.2%
1373 -28.8%
Cinebench R23 Multi Core
32197 +113.8%
18920 +25.6%
15061
12370 -17.9%
8830 -41.4%
Geekbench 6 Single Core
4698 +51.5%
3219 +3.8%
3100
Geekbench 6 Multi Core
18335 +20.1%
16667 +9.2%
15263
14548 -4.7%
13595 -10.9%
Geekbench 5 Single Core
2974 +28.5%
2314
1875 -19%
1793 -22.5%
1705 -26.3%
Geekbench 5 Multi Core
20469 +34.4%
17349 +13.9%
15235
13665 -10.3%
12390 -18.7%
Passmark CPU Single Core
4342 +2.4%
4240
4155 -2%
4091 -3.5%
Passmark CPU Multi Core
26045 +8.6%
25039 +4.4%
23993
23042 -4%
22480 -6.3%
Cinebench 2024 Single Core
142
141 -0.7%
M3
137 -3.5%
Cinebench 2024 Multi Core
1360 +28.4%
1203 +13.6%
1059
1025 -3.2%
960 -9.3%
Cinebench 2024 GPU
7768 +30.3%
6139 +3%
5961
5851 -1.8%
5415 -9.2%
Blender
1154 +406.1%
314 +37.7%
228
151 -33.8%
71 -68.9%