AMD Ryzen AI 7 PRO 350 vs M4 Pro (12C16G)
AMD Ryzen AI 7 PRO 350
M4 Pro (12C16G)
|
8 Cores
16 Threads
2GHz Clock
|
12 Cores
12 Threads
4.5GHz Clock
|
|
Socket
FP8
|
Socket
Unknown
|
|
iGPU
Radeon 860M 3000MHz
2.3 TFLOPS FP32
|
iGPU
16-Core M4 Pro
6.02 TFLOPS FP32
|
|
|
Ryzen AI 7 PRO 350 - 128.0GB/s
x1
M4 Pro (12C16G) - 273.1GB/s
x2.13
|
|
Ryzen AI 7 PRO 350 - 28W
x1
M4 Pro (12C16G) - 60W
x2.14
|
|
Ryzen AI 7 PRO 350 - 2.3 TFLOPS FP32
x1
M4 Pro (12C16G) - 6.02 TFLOPS FP32
x2.62
|
|
Ryzen AI 7 PRO 350 - 50 TOPS INT8
x1.32
M4 Pro (12C16G) - 38 TOPS INT8
x1
|
GB6S N/A
0%
|
GB6S 3,925
97%
|
|
GB6M N/A
0%
|
GB6M 20,605
77%
|
|
0%
|
0%
|
|
0%
|
0%
|
|
Manufacturer
AMD
|
Manufacturer
Apple
|
|
Architecture
Zen 5
|
Architecture
Apple Silicon 14
|
|
Family
Ryzen AI 300
|
Family
M Series
|
|
Instruction Set (ISA)
x86-64
|
Instruction Set (ISA)
AArch64
|
|
Codename
Strix Point
-
P-Core
Zen 5
E-Core
Zen 5c
|
Codename
Brava
M4 Pro
P-Core
AS14P
E-Core
AS14E
|
|
Market Segment
Laptop
|
Market Segment
Laptop
|
|
Release Date
1/6/2025
|
Release Date
10/29/2024
|
|
Foundry
TSMC
|
Foundry
TSMC
|
|
Node
N4P
|
Node
N3E
|
|
Die Size
233 mm²
|
Die Size
219 mm²
|
|
Transistor Count
28 Billion
|
Transistor Count
40 Billion
|
| FP8 |
Unknown
|
|
8 Cores
4 P-Cores
4 E-Cores
|
12 Cores
8 P-Cores
4 E-Cores
|
|
16 Threads
8 P-Threads
8 E-Threads
|
12 Threads
8 P-Threads
4 E-Threads
|
|
-
2GHz
5GHz Boost
|
P-Core
4.5GHz
-
|
|
-
-
|
E-Core
2.75GHz
|
|
Overclocking
Locked
|
Overclocking
Locked
|
|
Type
PCIe
|
Integrated Chipset
-
|
|
Transfer Rate
16GT/s
|
-
-
|
|
Lanes
4
|
-
-
|
|
Bandwidth
8GB/s
|
-
-
|
|
Bi-directional
16GB/s
|
-
-
|
|
L0 Instruction
|
L0 Instruction
|
|
L0 Data
|
L0 Data
|
|
L1 Instruction
32KB/Core
8-way set associative
-
-
|
L1 Instruction
192KB/P-Core
6-way set associative
128KB/E-Core
4-way set associative
|
|
L1 Data
48KB/Core
12-way set associative
-
-
|
L1 Data
128KB/P-Core
4-way set associative
64KB/E-Core
2-way set associative
|
|
L2
1024KB/Core
16-way set associative
-
-
-
-
|
L2
-
-
24MB Shared
16-way set associative
4MB Shared (E-Core)
16-way set associative
|
|
L3
16MB Shared
16-way set associative
|
-
-
-
|
|
-
|
24MB SLC Cache
|
|
Channels
2
|
Channels
4
|
|
Max Memory
256GB
64GB
|
Max Memory
32GB
|
|
ECC
Not Supported
|
ECC
Not Supported
|
|
Bus Width/Channel
64Bit
Bus Width
128Bit
|
Bus Width/Channel
64Bit
Bus Width
256Bit
|
|
Clock
4000MHz
2800MHz
Transfer Rate
8000MT/s
5600MT/s
|
Clock
4267MHz
Transfer Rate
8533MT/s
|
|
Bandwidth/Channel
64.0GB/s
44.8GB/s
Bandwidth
128.0GB/s
89.6GB/s
|
Bandwidth/Channel
68.3GB/s
Bandwidth
273.1GB/s
|
|
TDP
28W
15W cTDP Down
54W cTDP Up
|
TDP
60W
-
-
|
|
Temp
100°C Max
|
Temp
100°C Max
|
|
Included Cooler
|
Included Cooler
|
|
PCIe 4.0 x16 Lanes
|
No PCIe
|
|
Radeon 860M 3000MHz
384 Shaders
3000MHz
2.3 TFLOPS FP32
|
16-Core M4 Pro
2048 Shaders
1470MHz
6.02 TFLOPS FP32
|
|
XDNA 2
|
-
|
|
32 Cores
|
16 Cores
|
|
50 TOPS INT8
|
38 TOPS INT8
|
|
No Cellular
|
No Cellular
|
Copy Link