Intel® Arc™ Pro A40 – Compact Professional Graphics Card
The Intel® Arc™ Pro A40 is a low-profile, single-slot professional graphics card designed for compact workstations and SFF (Small Form Factor) systems. Featuring Intel’s Xe HPG architecture, hardware ray tracing, AI acceleration, and AV1 hardware encode/decode, it’s optimized for CAD, content creation, and AI-assisted workflows, all while fitting into space-limited builds.
Key Specifications
-
GPU Architecture: Intel® Xe HPG
-
Execution Units: 8 Xe-cores (128 vector engines, 128 XMX engines)
-
Ray Tracing Units: 8 hardware-accelerated
-
VRAM: 6GB GDDR6
-
Memory Interface: 96-bit
-
Memory Bandwidth: 192 GB/s
-
FP32 Performance: Up to 3.5 TFLOPs
Display & Media Support
-
Max Displays: 4
-
Outputs: 4× Mini DisplayPort 2.0 (UHBR10)
-
Max Resolution:
-
DP 2.0: up to 7680×4320 @ 60Hz (HDR)
-
4K @ 240Hz
-
-
Media Codec Support: AV1, HEVC, VP9, H.264 hardware encode & decode
Professional & AI Features
-
API Support: DirectX® 12 Ultimate, OpenGL® 4.6, Vulkan® 1.3
-
AI Acceleration: XMX engines for machine learning workloads
-
ISV Certification: Verified for leading professional software suites
-
Ray Tracing: Hardware-accelerated for enhanced realism in supported apps
Power & Form Factor
-
Typical Board Power: 50W
-
Cooling: Active single-slot cooler
-
Interface: PCIe 4.0 x8
-
Form Factor: Low-profile, single-slot
Intel Arc Pro A40 – Specifications
Architecture / Codename | Xe HPG (Alchemist), DG2-128 |
---|---|
Process Node | 6 nm (TSMC N6) |
Release Date | Q3 2022 (Aug 2022) |
Execution Units (EUs) | 128 |
Shading Units | 1024 |
TMUs | 64 |
ROPs | 32 |
Ray-Tracing Units | 8 |
Tensorth / AI Units (XMX) | 128 |
Base Clock | 1,500 MHz |
Boost Clock | 1,700 MHz |
FP32 Performance | ~3.48 TFLOPS |
FP16 Performance | ~6.96 TFLOPS |
INT4 (AI) Performance | ~111 TOPS |
Pixel Fillrate | 54.4 GPixel/s |
Texture Fillrate | 108.8 GTexel/s |
Memory | 6 GB GDDR6, 96-bit, 2,000 MHz / 16 Gbps |
Memory Bandwidth | 192 GB/s |
L2 Cache | 4 MB |
TDP | 50 W |
PCIe Interface | PCIe 4.0 ×16 (8 lanes) |
Display Outputs | 4× mini-DisplayPort 2.0 |
Max Resolution | 7680×4320 @ 60 Hz |
Supported APIs | DirectX 12 Ultimate, Vulkan 1.3, OpenGL 4.6, OpenCL 3.0 |
Video Codec Support | AV1 / H.264 / HEVC encode & decode |
Form Factor | Single-slot, blower-style fan |
Deep Link Features | Hyper Compute, Hyper Encode, Dynamic Power Share |