GPU Comparison
RADEON
AMD Radeon Instinct MI25
CORE STATE
Vega 10
VRAM
16 GB
CLOCK SPEED
1500 MHz
TDP
300 W
BUS WIDTH
2048 bit
ARCHITECTURE
GCN 5.0
PROCESS
14 nm
LAUNCH DATE
2017
VS
GEFORCE
A2
CORE STATE
GA107
VRAM
16 GB
CLOCK SPEED
1770 MHz
TDP
60 W
BUS WIDTH
128 bit
ARCHITECTURE
Ampere
PROCESS
8 nm
LAUNCH DATE
2021
PERFORMANCE BENCHMARKS
geekbench_opencl
DETAILED SPECIFICATIONS
SPECIFICATION
Instinct MI25
A2
Core Specs
Shading Units
4,096
1,280
-68.8%
Shaders
4,096
1,280
-68.8%
TMUs
256
40
-84.4%
ROPs
64
32
-50.0%
Compute Units
64
—
SM Count
—
10
Clocks
Base Clock
1400 MHz
1440 MHz
Boost Clock
1500 MHz
1770 MHz
Memory Clock
852 MHz
1704 Mbps effective
1563 MHz
12.5 Gbps effective
Memory
Memory Size
16 GB
16 GB
VRAM (MB)
16,384
16,384
0.0%
Memory Type
HBM2
GDDR6
Memory Bus
2048 bit
128 bit
Bandwidth
436.2 GB/s
200.1 GB/s
Cache
L1 Cache
16 KB (per CU)
128 KB (per SM)
L2 Cache
4 MB
2 MB
Performance
Pixel Rate
96.00 GPixel/s
56.64 GPixel/s
Texture Rate
384.0 GTexel/s
70.80 GTexel/s
FP32 (TFLOPS)
12.29 TFLOPS
4.531 TFLOPS
FP64 (TFLOPS)
768.0 GFLOPS (1:16)
70.80 GFLOPS (1:64)
FP16 (TFLOPS)
24.58 TFLOPS (2:1)
4.531 TFLOPS (1:1)
AI/RT
RT Cores
—
10
Tensor Cores
—
40
Power
TDP
300 W
60 W
TDP (W)
300
60
-80.0%
Suggested PSU
700 W
250 W
Power Connectors
2x 8-pin
None
Architecture
Architecture
GCN 5.0
Ampere
GPU Name
Vega 10
GA107
Generation
Radeon Instinct
(MIx)
Workstation Ampere
(Ax000)
Process Size
14 nm
8 nm
Transistors
12,500 million
8,700 million
Die Size
495 mm²
200 mm²
Foundry
GlobalFoundries
Samsung
Density
25.3M / mm²
43.5M / mm²
API Support
DirectX
12 (12_1)
12 Ultimate (12_2)
OpenGL
4.6
4.6
Vulkan
1.3
1.4
OpenCL
2.1
3.0
CUDA
—
8.6
Shader Model
6.7
6.8
Physical
Slot Width
Dual-slot
Single-slot
Length
267 mm
10.5 inches
—
Height
111 mm
4.4 inches
—
Outputs
No outputs
No outputs
Bus Interface
PCIe 3.0 x16
PCIe 4.0 x8
Other
Production
End-of-life
End-of-life
Predecessor
FirePro Data Center
Quadro Turing
Successor
—
Workstation Ada