BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:Declarative query processing in imperative managed runtimes
DTSTART:20141203T161500
DTEND:20141203T171500
DTSTAMP:20260406T185409Z
UID:2a7d870fccba7564912d616cf9b2e035665eba666174fb142eec980f
CATEGORIES:Conferences - Seminars
DESCRIPTION:Stratis Viglas\, University of Edinburgh\nThe falling price of
  main memory has led to the development and growth of in-memory databases.
   At the same time\, new advances in memory technology\, like persistent 
 memory\, make it possible to have a truly universal storage model\, access
 ed directly through the programming language in the context of a fully man
 aged runtime.  This environment is further enhanced by language-integrate
 d query\, which has picked up significant traction and has emerged as a ge
 neric\, safe method of combining programming languages with databases with
  considerable software engineering benefits.  Our perspective on language
 -integrated query is that it combines the runtime of a programming languag
 e with that of a database system.  This leads to the question of how to t
 ightly integrate these two runtimes.  Our proposal is to apply just-in-ti
 me code generation and compilation techniques that have recently been deve
 loped for general query processing.  The idea is that instead of compilin
 g queries to query plans\, which are then interpreted\, the system generat
 es customized native code that is then compiled and executed by the query 
 engine.  At the same time\, we must enable the runtime to take advantage 
 of advances in main memory technology and\, primarily\, persistent memory.
   Persistent memory is byte-addressable\, but exhibits asymmetric I/O: wr
 ites are typically one order of magnitude more expensive than reads.  Byt
 e addressability combined with I/O asymmetry render the performance profil
 e of persistent memory unique. Thus\, it becomes imperative to find new wa
 ys to seamlessly incorporate it into data processing in managed runtimes.
   We do so in the context of fundamental query processing operations and 
 introduce the notion of write-limited algorithms that effectively minimize
  the I/O cost.  We give a high-level API that enables the system to dynam
 ically optimize the workflow of the algorithms\; or\, alternatively\, allo
 ws the developer to tune the write profile of the algorithms. This dynamic
  adaptation fits in well with the notion of just-in-time compilation.  
   We present the results of our work in integrating database and programm
 ing language runtimes through code generation and extensive just-in-time a
 daptation.  Our techniques deliver significant performance improvements o
 ver non-integrated solutions.  Our work  makes important first steps tow
 ards a future where data processing applications will commonly run on mach
 ines that can store their entire datasets in-memory\, and will be written 
 in a single programming language employing higher-level APIs and  languag
 e-integrated query to provide  transparent and highly efficient querying.
LOCATION:BC 410 https://plan.epfl.ch/?room==BC%20410
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR