Vision Transformers Change the AI Acceleration Rules
Transformers were first introduced by the team at Google Brain in 2017 in their paper, “Attention is All You Need“. Since their introduction, transformers have inspired a flurry of investment and research which have produced some of the most impactful model architectures and AI products to-date, including ChatGPT which is an acronym for Chat Generative Pre-trained Transformer.
Transformers are also being employed for vision applications (ViTs). This new class of models was empirically proven to be viable alternatives to more traditional Convolutional Neural Networks (CNNs) in the paper “An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale“, published by the team at Google Brain in 2021.
Vision transformers are of a size and scale that are approachable for SoC designers targeting the high-performance, edge AI market. There’s just one problem: vision transformers are not CNNs and many of the assumptions made by the designers of first-generation Neural Processing Unit (NPU) and AI hardware accelerators found in today’s SoCs do not translate well to this new class of models.
What makes Vision Transformers so special?
To read the full article, click here
Related Semiconductor IP
- HDMI 2.1 Quad-Pixel Tx Controller
- Quadruple Capacitor Switch
- 10bit Quadruple DAC (Vcc/Vdd=3.3/1.8V, DNL:+/- 1/2LSB,INL:+/- 1/2LSB)
- M’ary Quadrature Amplitude Modulator IP in VHDL, FPGA and SDR
- 3-6 GHz to 0.075-3 GHz Quadrature former
Related Blogs
- From ChatGPT to Computer Vision Processing: How Deep-Learning Transformers Are Shaping Our World
- How AI Will Change Chip Design
- New Neural Processor Aids in Adding AI to Small, Low-Power Vision Processing Designs
- How the SYCLOPS project democratizes AI acceleration
Latest Blogs
- What’s Your Vector? Synopsys Introduces New ARC VPX6 Digital Signal Processor
- Rambus CXL IP: A Journey from Spec to Compliance
- Alphawave Semi is in Play!
- Vision Transformers Have Already Overtaken CNNs: Here’s Why and What’s Needed for Best Performance
- Tailoring Root Of Trust Security Capabilities To Specific Customer Needs