GPU Comparison
GPU
Intel Arc A310E
CORE STATE
DG2-128
VRAM
4 GB
CLOCK SPEED
2000 MHz
TDP
75 W
BUS WIDTH
64 bit
ARCHITECTURE
Xe-HPG
PROCESS
6 nm
LAUNCH DATE
2024
VS
GEFORCE
H20 NVL16
CORE STATE
GH100
VRAM
96 GB
CLOCK SPEED
1980 MHz
TDP
400 W
BUS WIDTH
6144 bit
ARCHITECTURE
Hopper
PROCESS
5 nm
LAUNCH DATE
2025
DETAILED SPECIFICATIONS
SPECIFICATION
A310E
H20 NVL16
Core Specs
Shading Units
768
9,984
+1200.0%
Shaders
768
9,984
+1200.0%
TMUs
32
312
+875.0%
ROPs
16
24
+50.0%
SM Count
—
78
Execution Units
96
—
Clocks
Base Clock
2000 MHz
1830 MHz
Boost Clock
2000 MHz
1980 MHz
Memory Clock
1937 MHz 15.5 Gbps effective
1313 MHz 5.3 Gbps effective
Memory
Memory Size
4 GB
96 GB
VRAM (MB)
4,096
98,304
+2300.0%
Memory Type
GDDR6
HBM3
Memory Bus
64 bit
6144 bit
Bandwidth
124.0 GB/s
4.03 TB/s
Cache
L1 Cache
—
256 KB (per SM)
L2 Cache
4 MB
60 MB
Performance
Pixel Rate
32.00 GPixel/s
47.52 GPixel/s
Texture Rate
64.00 GTexel/s
617.8 GTexel/s
FP32 (TFLOPS)
3.072 TFLOPS
39.54 TFLOPS
FP64 (TFLOPS)
768.0 GFLOPS (1:4)
19.77 TFLOPS (1:2)
FP16 (TFLOPS)
6.144 TFLOPS (2:1)
79.07 TFLOPS (2:1)
AI/RT
RT Cores
6
—
Tensor Cores
—
312
XMX Cores
96
—
Power
TDP
75 W
400 W
TDP (W)
75
400
+433.3%
Suggested PSU
250 W
800 W
Power Connectors
None
—
Architecture
Architecture
Xe-HPG
Hopper
GPU Name
DG2-128
GH100
Generation
Alchemist (Arc 3)
Server Hopper (Hxx)
Process Size
6 nm
5 nm
Transistors
7,200 million
80,000 million
Die Size
157 mm²
814 mm²
Foundry
TSMC
TSMC
Density
45.9M / mm²
98.3M / mm²
API Support
DirectX
12 Ultimate (12_2)
—
OpenGL
4.6
—
Vulkan
1.4
—
OpenCL
3.0
3.0
CUDA
—
9.0
Shader Model
6.6
—
Physical
Slot Width
Single-slot
SXM Module
Length
168 mm 6.6 inches
—
Height
69 mm 2.7 inches
—
Outputs
4x mini-DisplayPort 2.0
No outputs
Bus Interface
PCIe 4.0 x8
PCIe 5.0 x16
Other
Production
End-of-life
Active
Predecessor
Xe Graphics
Server Ada
Successor
Battlemage
Server Blackwell