QUETZAL: Vector Acceleration Framework for Modern Genome Sequence Analysis Algorithms

QUETZAL is a hardware-software co-designed vector acceleration framework that significantly outperforms other algorithms in genome sequence analysis. It offers high performance and energy efficiency, capable of accelerating both modern and classical algorithms. With features like custom hardware and reduced access latency, QUETZAL showcases remarkable speedups and improved performance, especially when processing long reads like WFA and Smith-Waterman.

• Genome Analysis
• Vector Acceleration
• QUETZAL
• High Performance
• Modern Algorithm

brea_iah Follow

Uploaded on Nov 16, 2024 | 0 Views

Download Presentation

Please find below an Image/Link to download the presentation.

The content on the website is provided AS IS for your information and personal use only. It may not be sold, licensed, or shared on other websites without obtaining consent from the author. Download presentation by click this link. If you encounter any issues during the download, it is possible that the publisher has removed the file from their server.

Presentation Transcript

1. QUETZAL: Vector Acceleration Framework for Modern Genome Sequence Analysis Algorithms Lead authors: Julian Pavon, Ivan Vargas Valdivieso, Carlos Rojas, Cesar Hernandez Co-authors: Mehmet Aslan, Roger Figueras, Yichao Yuan, Joel Lindegger, Mohammed Alser, Francesc Moll, Santiago Marco Sola, Oguz Ergin, Nishil Talati, Onur Mutlu, Osman Unsal, Mateo Valero and Adrian Cristal

2. Accelerating the Genome Analysis Pipeline ? Custom Accelerator High performance and Energy efficiency. They are tight to a single algorithm or a single sequence length. High design and entry-cost. CPU/GPU Flexible. Low entry-cost. Generality and flexibility limits their performance. Answer: We proposed QUETZAL, a hardware-software co-designed vector acceleration framework 6

3. QUETZAL QBUFFERs VPU data encoder Issue queue QUETZAL supports bit-encoded operations without extra instructions. access ctrl Operand Bypass count count count QBUFFERS reduce the access latency from 22 cycles to only 2 cycles. count ALU ALU VRF ALU ALU ALU QUETZAL features custom hardware to calculate the maximum number of exact matches. 15

4. Evaluation QUETZAL significantly outperforms all the evaluated algorithms. Speedup: 5.7x better performance compared to other vectorized algorithms. QUETZAL is capable of accelerating both modern and classical genome sequence analysis algorithms. 20

5. Evaluation When processing long reads, QUETZAL outperform by 2.7x and 1.1x for WFA and Smith-Waterman (Classical algorithm), respectively. 22