NVIDIA RTX A1000 Embedded GPU MXM Module Based On Ampere architecture.
- NVIDIA RTX A1000 GPU.
- 4GB GDDR6 Memory.
- Up to 6.66 TFLOPS.
- Up to 7680 x 4320 Resolution.
- 2048 CUDA Cores.
- 0°C to 55°C Operating Temperature.
The Aetina M3A1000-PP features advanced NVIDIA RTX™ A1000 GPU with NVIDIA Ampere Architecture in the MXM 3.1 Type A, bringing excellent performance and power efficiency into a wide variety of embedded systems.
- GPU Engine
-
GPU Architecture:
NVIDIA Ampere.Graphics Processing Unit:
NVIDIA RTX A1000.Nvidia CUDA Cores:
2048.Tensor Cores:
64.RT Cores:
16.Floating Point Performance:
4.67 TFLOPS (M3A1000-PPA-A2 & M3A1000-PPW-A2).
6.66 TFLOPS (M3A1000-PPA-A1 & M3A1000-PPW-A1).Core Clock (MHz):
Base= 630 / Boost= 1140 (M3A1000-PPA-A2 & M3A1000-PPW-A2).
Base= 1192 / Boost= 1627 (M3A1000-PPA-A1 & M3A1000-PPW-A1).
- Memory
-
Size:
4GB GDDR6.Clock:
11Gbps (M3A1000-PPA-A2 & M3A1000-PPW-A2).
14Gbps (M3A1000-PPA-A1 & M3A1000-PPW-A1).Interface Width:
128-bit.Bandwidth (GB/sec):
176 (M3A1000-PPA-A2 & M3A1000-PPW-A2).
224 (M3A1000-PPA-A1 & M3A1000-PPW-A1).
- Feature Support
-
PCIE Gen4.0 x8.
DirectX 12 Ultimate.
Open GL 4.6.
Vulkan 1.2.OS Support:
Windows 10/11 (64-bit).
- Display
-
Maximum Displays per Board:
4 x DisplayPort.Resolution:
7680 x 4320.
- Power
-
35W TGP (M3A1000-PPA-A2 & M3A1000-PPW-A2).
60W TGP (M3A1000-PPA-A1 & M3A1000-PPW-A1).
- Environmental
-
Operation Temperature (Standard):
0℃ to +55℃ (M3A1000-PPA-A2 & M3A1000-PPW-A2).Operation Temperature (Extended):
-40 to + 85°C (M3A1000-PPA-A1 & M3A1000-PPW-A1).Storage Temperature:
-40°C to 85°C.Operation Humidity:
95% @ 40°C Related Humidity (Non-condensing).
- Dimensions
-
82 x 70 mm.
MXM 3.1 Type B Module Powered By 4 x Hailo-8™ AI Inference Processors
- Powered by 4x Hailo-8 AI Processors
- MXM Type B small form factor accelerator platform
- Deliver up to 104 TOPS of AI performance at a typical power consumption of 25W
NVIDIA RTX A500 Embedded GPU MXM Module Based On Ampere Architecture.
- NVIDIA RTX A500 GPU.
- 4GB GDDR6 Memory.
- Up to 7.28 TFLOPS.
NVIDIA Quadro T1000 Embedded GPU MXM Module Based On Turing Architecture.
- NVIDIA Quadro T1000 GPU.
- 4GB GDDR6 Memory.
- Up to 2.6 TFLOPS.