之前因為好奇心使然,對所謂的雲端計算很感興趣,畢竟scalable computational skillset蠻重要的,自從之前有把整個研究室叢集電腦塞爆,不斷修改平行代碼,測試速度等,還是需要一個多月計算時間的經驗,如何短時間調用更多運算資源,可以解決未來遇到此類事件的scenario。
Google Cloud Platform之前因為使用SevenBridge的服務時有稍微碰到,因為SeverBridge的生資運算工具本質上是建立在Google Cloud Platform上的Genomic API,最近則是因為需要使用到GPU的效能,但暫時無法組一台桌機來做這件事,所以覺得可以試試看GCP,假如他的學習機會成本可接受,目前整體感覺GCP上手容易度頗高的,且在不使用客戶端下就能做很多事情。另外,GCP很大部分都是針對"資料工程"圍繞建立的服務,所以頗多針對機器學習常用工具推出的雲端服務,非常貼心!
相對於AWS一開始從很底層提供雲端計算服務,Google Cloud Platform的邏輯比較跟者應用層跑,相對起來很快可以轉移上去。
Goolge Cloud Platform基礎架構說明
Compute Engine: https://cloud.google.com/compute/
Storage: https://cloud.google.com/storage/
Pricing: https://cloud.google.com/pricing/
Cloud Launcher: https://cloud.google.com/launcher/
Pricing Philosophy: https://cloud.google.com/pricing/philosophy/
Google Cloud Platform中層服務
Cloud Pub/Sub: https://cloud.google.com/pubsub/
Cloud Dataflow: https://cloud.google.com/dataflow/
Cloud Datastore: https://cloud.google.com/datastore/
Cloud Bigtable: https://cloud.google.com/bigtable/
Google BigQuery: https://cloud.google.com/bigquery/
Google Cloud Platform高級服務(high level application level)
Cloud Datalab: https://cloud.google.com/datalab/
TensorFlow: https://www.tensorflow.org/
Cloud Machine Learning: https://cloud.google.com/ml/
Vision API: https://cloud.google.com/vision/
Translate API: https://cloud.google.com/translate/
Speech API: https://cloud.google.com/speech/
GCP部落格或應用
Reliable task scheduling on Google Compute Engine
Real-time data analysis with Kubernetes, Cloud Pub/Sub, and BigQuery
Processing logs at scale using Cloud Dataflow
Big data and machine learning blog: https://cloud.google.com/blog/big-data/
Google Cloud Platform blog: https://cloudplatform.googleblog.com/
Google Cloud Platform curated articles: https://medium.com/google-cloud