Difference between revisions of "Edge AI SDK/Accelerator/RTX-A5000"

From ESS-WIKI
Jump to: navigation, search
 
(20 intermediate revisions by 3 users not shown)
Line 1: Line 1:
  
−
{{DISPLAYTITLE:RTX-A5000_rc}}
+
{{DISPLAYTITLE:RTX-A5000}}
  
 
{| border="0" cellpadding="5" cellspacing="0" style="width:100%;"
 
{| border="0" cellpadding="5" cellspacing="0" style="width:100%;"
 
|-
 
|-
 
| style="width: 300px; high: 200px; text-align: center;vertical-align: top;" |  
 
| style="width: 300px; high: 200px; text-align: center;vertical-align: top;" |  
−
[[File:RTXA5000.png|250x150px|RTXA5000.png]]
+
[[File:EAS NV RTX A5000.jpg|500x400px|RTXA5000.png]]
  
 +
<span style="font-size:larger;">The NVIDIA RTX A5000 is a high-performance GPU for professional AI, data science, and graphics applications, designed for workstation use<span style="color: rgb(77, 77, 77); line-height: 20px;">.</span></span>
 +
 +
| style="vertical-align: top;" | &nbsp;
 
| style="vertical-align: top;" |  
 
| style="vertical-align: top;" |  
−
<span style="font-size:larger;">The NVIDIA RTX A5000 is a high-performance GPU for professional AI, data science, and graphics applications, designed for workstation use<span style="color: rgb(77, 77, 77); line-height: 20px;">.</span></span>
+
{| class="wikitable" style="width:100%; margin:auto; border:2px solid #ffffff; border-collapse:separate; border-spacing:10px; text-align:center;"
 +
|-
 +
! style="background-color:#3664B1; color:white; width:150px; text-align:center; border:1px;" | AI Applications
 +
| colspan="5" style="padding:0; border:1px solid black;" |
 +
{| class="wikitable" style="width:100%; border-collapse:separate; border-spacing:5px; border:0px;"
 +
|-
 +
| style="border:1px solid gray; text-align:center; padding:10px; width:25%;" | Vision AI Apps
 +
| style="background-color:#D3D3D3; border:1px solid gray; text-align:center; padding:10px; width:25%;" | Pre-trained Models
 +
| style="border:1px solid gray; text-align:center; padding:10px; width:25%;" | Monitoring
 +
| style="border:1px solid gray; text-align:center; padding:10px; width:25%;" | Benchmark
 +
|}
 +
 
 +
|-
 +
! style="background-color:#3664B1; color:white; text-align:center;" | Utility
 +
| colspan="5" style="padding:0; border:1px solid black;" |
 +
{| class="wikitable" style="width:100%; border-collapse:separate; border-spacing:5px; border:0px;"
 +
|-
 +
| colspan="1" style="background-color:#74B71B; border:1px solid gray; text-align:center; padding:10px; width:50%;" | <span style="font-size:larger; color:white;">nvidia-smi</span>
 +
| colspan="1" style="background-color:#74B71B; border:1px solid gray; text-align:center; padding:10px; width:50%;" | <span style="font-size:larger; color:white;">trtexec ( AI Benchmark )</span>
 +
|}
 +
 
 +
|-
 +
! style="background-color:#3664B1; color:white; text-align:center;" | AI SDK
 +
| colspan="5" style="padding:0; border:1px solid black;" |
 +
{| class="wikitable" style="width:100%; border-collapse:separate; border-spacing:5px; border:0px;"
 +
|-
 +
| colspan="1" style="background-color:#74B71B; border:1px solid gray; text-align:center; padding:10px; width:33%;" | <span style="font-size:larger; color:white;">TensorRT 8.6.1.6</span>
 +
| colspan="1" style="background-color:#74B71B; border:1px solid gray; text-align:center; padding:10px; width:33%;" | <span style="font-size:larger; color:white;">CUDA 12.2</span>
 +
| colspan="1" style="background-color:#74B71B; border:1px solid gray; text-align:center; padding:10px; width:33%;" | <span style="font-size:larger; color:white;">Deepstream 6.4</span>
 +
|}
  
−
*<span style="font-size:larger;">CUDA Cores: 8192</span>
+
|-
−
*<span style="font-size:larger;">Performance: Up to 27.8 TFLOPS (single precision)</span>
+
! style="background-color:#3664B1; color:white; text-align:center;" | Driver
−
*<span style="font-size:larger;">Power Efficiency: 230W</span>
+
| colspan="5" style="padding:0; border:1px solid black;" |
−
*<span style="font-size:larger;">Interface: PCIe Gen 4.0 x16</span>
+
{| class="wikitable" style="width:100%; border-collapse:separate; border-spacing:5px; border:0px;"
−
*<span style="font-size:larger;">Frameworks: TensorFlow, PyTorch, ONNX, Keras</span>
+
|-
−
*<span style="font-size:larger;">OS: Linux, Windows</span>  
+
| colspan="5" style="background-color:#74B71B; border:1px solid gray; text-align:center; padding:10px; width:100%;" | <span style="font-size:larger; color:white;">Nvidia Driver R535.104.12</span>
 +
|}
  
 +
|-
 +
! style="background-color:#3664B1; color:white; width:150px; text-align:center; border:1px;" | OS / Kernel
 +
| colspan="5" style="border:1px solid black; text-align:center; padding:15px;" | <span style="font-size:larger;">Ubuntu 22.04 ( 5.15.0-1027-intel-iotg )</span>
 +
|}
 +
<div style="font-size:large; color: rgb(77, 77, 77); line-height: 20px; text-align: center;">AI Software Stack on RTX-A5000 ( v3.0.0 )</div> <div style="text-align: right; margin-top: 10px;">
 +
{| style="border-collapse: separate; border-spacing: 10px; float: right; margin-right: 0"
 +
|-
 +
| style="width:20px; height:20px; background-color:#ffffff; border:1px solid black;" | &nbsp;
 +
| style="text-align: left;" | Advantech
 +
| style="width:20px; height:20px; background-color:#74B71B; border:1px solid black;" | &nbsp;
 +
| style="text-align: left;" | NVIDIA
 +
| style="width:20px; height:20px; background-color:#D3D3D3; border:1px solid black;" | &nbsp;
 +
| style="text-align: left;" | Open Source
 +
|}
 +
</div>
 
|}
 
|}
 +
 +
&nbsp;
  
 
{| border="0" cellpadding="0" cellspacing="5" style="width:100%;"
 
{| border="0" cellpadding="0" cellspacing="5" style="width:100%;"
Line 30: Line 80:
 
|-
 
|-
 
| style="vertical-align: top;line-height:20px;" |  
 
| style="vertical-align: top;line-height:20px;" |  
−
[http://ess-wiki.advantech.com.tw/view/Edge_AI_SDK/AI_System/AIR-510_rc AIR-510]
+
[http://ess-wiki.advantech.com.tw/view/Edge_AI_SDK/AI_System/AIR-510 AIR-510]
  
 
|}
 
|}
Line 42: Line 92:
 
|-
 
|-
 
| style="vertical-align: top;line-height:20px;" |  
 
| style="vertical-align: top;line-height:20px;" |  
−
[http://ess-wiki.advantech.com.tw/view/Edge_AI_SDK/AI_Framework/RTXA5000_AI_Suite_rc RTXA5000_AI_Suite]
 
  
−
[https://ess-wiki.advantech.com.tw/view/Edge_AI_SDK/AI_Framework/Hailo#Applications Applications]
+
[[Edge_AI_SDK/AI_Framework/Nvidia_x86_64#CUDA|CUDA]]
 +
 
 +
[https://ess-wiki.advantech.com.tw/view/Edge_AI_SDK/AI_Framework/Nvidia_x86_64#DeepStream DeepStream]
 +
 
 +
[https://ess-wiki.advantech.com.tw/view/Edge_AI_SDK/AI_Framework/Nvidia_x86_64#TensorRT TensorRT]
 +
 
 +
[https://ess-wiki.advantech.com.tw/view/Edge_AI_SDK/AI_Framework/Nvidia_x86_64#Edge_AI_SDK.C2.A0.2F_Application Applications]
  
 
|}
 
|}
Line 54: Line 109:
 
|-
 
|-
 
| style="vertical-align: top;line-height:20px;" |  
 
| style="vertical-align: top;line-height:20px;" |  
−
[http://ess-wiki.advantech.com.tw/view/Edge_AI_SDK/AI_Framework/Hailo#Benchmark Benchmakrks]<br/> [http://ess-wiki.advantech.com.tw/view/Edge_AI_SDK/AI_Framework/Hailo#HailoRT_CLI HailoRT CLI]
+
[https://ess-wiki.advantech.com.tw/view/Edge_AI_SDK/AI_Framework/Nvidia_x86_64#Benchmark Benchmarks]
 +
 
 +
[https://ess-wiki.advantech.com.tw/view/Edge_AI_SDK/AI_Framework/Nvidia_x86_64#NVIDIA_System_Management_Interface Nvidia-smi]
  
 
|}
 
|}
Line 61: Line 118:
  
 
|}
 
|}
−
 
−
[[Category:Editor]]
 

Latest revision as of 09:55, 25 December 2024


RTXA5000.png

The NVIDIA RTX A5000 is a high-performance GPU for professional AI, data science, and graphics applications, designed for workstation use.

 
AI Applications
Vision AI Apps Pre-trained Models Monitoring Benchmark
Utility
nvidia-smi trtexec ( AI Benchmark )
AI SDK
TensorRT 8.6.1.6 CUDA 12.2 Deepstream 6.4
Driver
Nvidia Driver R535.104.12
OS / Kernel Ubuntu 22.04 ( 5.15.0-1027-intel-iotg )
AI Software Stack on RTX-A5000 ( v3.0.0 )
  Advantech   NVIDIA   Open Source

 

Edge AI System

AIR-510

 

AI Inference Run-time SDK

CUDA

DeepStream

TensorRT

Applications

AI Benchmark & Tools

Benchmarks

Nvidia-smi

Retrieved from "https://ess-wiki.advantech.com.tw/wiki/index.php?title=Edge_AI_SDK/Accelerator/RTX-A5000&oldid=37948"