Update README.md

qdrant/README.md

@@ -1,8 +1,8 @@
 # Qdrant vector database example
 
-WasmEdge is emerging as a lightweight, portable, secure and cloud-native runtime for large language models (LLMs). LLM inference applications, such as RAG chatbots and AI agents, can be developed on Mac or Windows, compiled to Wasm once, and then deployed across Nvidia / GPU / NPU powered devices or servers. 
+WasmEdge is emerging as a lightweight, portable, secure and cloud-native runtime for large language models (LLMs). LLM inference applications, such as RAG chatbots and AI agents, can be developed on Mac or Windows, compiled to Wasm once, and then deployed across Nvidia / AMD / ARM-powered devices or servers, fully taking advantage of on-device GPUs, NPUs, and accelerators.
 
-However, besides the LLM inference runtime, a key function of those LLM applications is to manage vector embeddings in vector databases. The [qdrant-rest-client](https://crates.io/crates/qdrant_rest_client) crate allows you to access the Qdrant vector database from your portable Wasm apps!
+Hence, besides the LLM inference runtime, those LLM applications also need to manage embeddings in vector databases. The [qdrant-rest-client](https://crates.io/crates/qdrant_rest_client) crate allows you to access the Qdrant vector database from your portable Wasm apps!
 
 ## Quick start