成立於 2010 年，總部位於美國紐約市，研發中心位於以色列特拉維夫，擁有多達 10 項平行運算國際專利技術。SQream 重新定義了全球巨量資料關聯分析的技術門檻，架構於 GPU 之上的運算能力提供極速、高彈性、成本效益絕佳的 GPU 加速 SQL 資料倉儲與分析技術，專事 Terabytes、Petabytes 巨量資料進行即時查詢與連接分析。全球電信、半導體、零售、醫療、金融等產業，均已仰賴 SQream 之技術加速解決以往無法達成的商業效益。
企業組織之數據資料正以指數級速度成長，而傳統 CPU-based 資料倉儲與分析系統，有高達90%的數據價值尚未被充分發揮，面對動輒以 Terabyte 為計的巨量資料時更顯得左支右絀，Cubing、Pre-computation、Indexing 等繁瑣冗長的資料準備作業，不僅造成數據資料分析結果緩不濟急且錯失先機，更易造成有形的商業損失。
SQream DB 大規模多重平行 GPU 關聯式資料庫具備以 NVIDIA® GPU 為核心的 MPP 大規模多重平行分析專利技術，專事提供巨量、精確且極速的 TB/PB 資料分析，運算速度與傳統 CPU 相較優化高達 100 倍之多。SQream DB 採用標準 ANSI SQL，無需繁複轉換程序即可透過 ODBC 或 JDBC 等連結方式無縫整合資料視覺化 BI 工具，如：Tableau、 Jupyter、 H2O、Spotfire、 Qlik 等。SQream DB 擁有多項國際專利技術，打破以往必須仰賴龐大的 server farm 進行巨量資料關聯分析的舊迷思，巨幅降低導入成本，立即脫離持續以硬體對抗資料成長速度的困境。
Mobile services subscribers use their connected devices for everything from voice, text, video, gaming, and more, creating enormous and fast-growing quantities of data. While this data carries great potential, Telecoms bear the difficult challenge of storing, navigating, and turning these huge data stores into actionable insights. Those telecom organizations that succeed in unlocking the intelligence gems hidden in these treasure troves of data often find themselves better poised to compete in an increasingly cluttered and competitive market.
The financial landscape has changed drastically over the last twenty years, as technological advancements and the internet revolution have enabled new and more complex financial business models. These changes brought with them new ways of interacting with customers and other financial institutions, and facilitated game-changing competition by narrowing certain barriers to entry. The digital transformation of financial services produced never-before-seen volumes of data growing at an unprecedented rate.
Hadoop came to prominence when the web exploded with unstructured data. The use of unstructured data is common for web analytics, where flexibility is required for unknown or compound fields (arrays, nested objects, or just unknown). The popularity of Hadoop for these use-cases led to its adoption, also for structured use-cases. For these cases, SQL query engines have been bolted on Hadoop, and convert relational operations into map/reduce style operations.
In the past, data was small, as were the number of data consumers. Most datasets were relatively simple, coming from a handful of ERP, CRM and other transactional sources. Traditional data warehouses were built to support this type of data. As computing hardware advanced, these databases got faster ‘for free’. However, they have by now become legacy technology incapable of utilizing new parallelized computing paradigms
SQream DB brings a new technological approach that eliminates data professionals’ struggles at the source, resulting in fast, accurate, up-to-the-minute dashboards from which vast new insights can be extracted. SQream’s GPU-accelerated data warehouse is designed for fast, unrestricted access to an organization’s full scope of data, even when data grows exponentially. SQream’s technology enables Tableau users to explore and productize interactive dashboards that power many aspects of modern data-driven organizations.
SQream DB is a fully-featured GPU-accelerated data warehouse, capable of handling the most complex queries. SQream DB has all of the features you expect from a relational database system, like comprehensive ANSI SQL support. Anyone can use SQream DB to load, store, and analyze data up to 100x faster than any other data warehouse.
Clickstream analysis on hundreds of millions of rows with SQream usingTableau visualizer - Insights in under 2 seconds.
Organizations today produce exponentially more data than they did just a few years ago, but their databases weren’t built to handle these new volumes. As a result, reporting takes way too long, and some complex analytics simply cannot be done. The Era of Massive Data is upon us, and a new approach is required to overcome the limitations of traditional CPU-based data stores.
SQream 大規模多重平行 GPU 關聯式資料庫
- Ad-Hoc 查詢
- 加速 ETL 及資料準備
- 整合或取代暨有 Data Warehouse 或 Hadoop 環境