57 lines
1.2 KiB
Markdown
57 lines
1.2 KiB
Markdown
## Introduction
|
|
### Energy Demand of Applications
|
|
<hr/>
|
|
|
|
|
|
- Total compute energy approaches world's energy production
|
|
|
|
--> Drastic improvements in energy efficiency needed
|
|
|
|
<div class="flex justify-center">
|
|
<img src="/world_energy.svg">
|
|
</div>
|
|
|
|
<Footnotes separator>
|
|
<Footnote>
|
|
SRC. „Decadal Plan for Semiconductors“, Januar 2021. https://www.src.org/about/decadal-plan/.
|
|
</Footnote>
|
|
</Footnotes>
|
|
|
|
<!--
|
|
- compute 2x every two years
|
|
- energy production 2% per year
|
|
|
|
- to meet future compute demands, drastic improvements in energy efficiency
|
|
-->
|
|
|
|
---
|
|
|
|
## Introduction
|
|
### Memory Bound Workloads
|
|
<hr/>
|
|
|
|
- AI applications become increasingly memory-bound
|
|
|
|
--> Instead of bringing the data to the processing, bring the processing to the data
|
|
|
|
<br>
|
|
|
|
<div class="flex justify-center">
|
|
<img src="/gpt.svg">
|
|
</div>
|
|
|
|
<Footnotes separator>
|
|
<Footnote>
|
|
Ivo Bolsens. „Scalable AI Architectures for Edge and Cloud“, 2023.
|
|
</Footnote>
|
|
</Footnotes>
|
|
|
|
<!--
|
|
- Emerging AI applications become increasingly memory-bound
|
|
- Roofline model
|
|
- Not limited by compute power but by memory
|
|
- researchers begin to consider PIM to circumvent memory bottleneck
|
|
|
|
- (drastically more parameters in GPT-3, operational intensity goes down)
|
|
-->
|