| Idea Name: |
AI Accelerator |
| Slogan: |
Empowering People with AI |
| Supervisor Name: |
Dr. Hussain Ali |
| Supervisor Designation: |
Assistant Professor |
| Supervisor School: |
MCS |
| Supervisor Department: |
Electrical Engineering |
| Contact number: |
03345440833 |
| Email ID: |
[email protected] |
| Abstract: |
| We are going to design an FPGA-based AI Accelerator which will help in accelerating the AI based algorithms running or calculating time span effectively. |
| What is the unmet need in society that your idea will fulfill ? |
| Machine Learning (ML) is widely used in Artificial Intelligence (AI) applications. ML algorithms requires execution in short time for real-time AI applications. The project will lead to the development of a platform for acceleration of ML algorithms. |
| Who needs it ? How many would benefit ? |
| Now-a-days every industry is shifting to AI based operations. So the major setback is the high computational complexity. So our project will help a lot in reducing these complexities. In this way it can be used in technological industries where ML algorithms computation has to be done regularly.
|
| How will the solution works |
| AI accelerator can be used in different places where ML algorithms are being applied. We will test an ML algorithm
by accelerating it through our hardware. We will be accelerating one ML algorithm and will compare the non-accelerated and accelerated outcomes. The accelerated outcome will have low computational complexity and low computational time span which will be less than the non-accelerated outcome. |
| Who are your competitors ? How is your solution different |
| Instead of GPU based accelerator we are going for FPGA based accelerator which will reduce the computational complexities more efficiently. We are developing an Indigenous solutions for Pakistan to enhance the local production of such accelerators which will help in reduction of importing cost. |
| Status: |
new |
| Entry Date & Time: |
2021-12-14 (0701) |
