Overview Foils September95
	Some Remarks on the Future of HPF --
	Metacomputing and the World Wide Web
		Icase HPF Workshop
		August 23,1995

		Geoffrey Fox
		NPAC
		Syracuse University
		111 College Place
		Syracuse NY 13244-4100
	Abstract of HPF Futures Presentation
		This talk discusses the impact of World Wide Web and large scale distributed (meta) computing on the future of HPF.
		It builds on concepts in Web Vision Presentation with an overall pervasive WebWindows environment
		The Boston-NPAC Proposal for WebWork which notes the critical importance of the excellent software engineering environment that can be built on top of WebWindows
		Discussion of Java which allows proper client-server implementations on the Web and the trade-off between Interpreted and Compiled Environments
			Implications of this for parallel C++ and Fortran
		Domain Specific Problem Solving Environments and the relation to WebScript
		VRML as an example of a universal 3D data structure
	Guidelines for HPF Futures
		We need to do more than make HPF available on PC/Workstation networks and shared memory multiprocessors
		We need to make HPF (and similar HPCC technologies) the languages that typical(many) PC/WS user will adopt because it supports the distributed computing opportunities of the World Wide Web
		This implies that we need to build HPF on top of common Web or distributed computing standards.
			HPF brings to the Web world, the accumulated experience of synchronization and parallel algorithms from HPCC.
			 This includes collective communication (multi-cast) etc.
	Interpreters versus Compilers -- I?
		We need to use compilers on tightly coupled systems such as MPP's (shared and distributed memory)
		But for metacomputing, the hardware intrinsically has latencies that suggests increased flexibility of interpreters is more appropriate
			This implies a hybrid compiler-interpreter environment
		Maybe frontends should be built with interpreters such as object-oriented PERL5 so easier to link with Web. 
		Note that interpreted environment will have best software engineering support and so suggestion is  -- taking SP2 as example:
			debug on SP2 set up with compute-enhanced Web Server on each node with say MPI running on top of HTTP message passing protocol
			execute debugged code on conventional SP2 with high performance compiled environment
	Interpreters versus Compilers -- Domain Specific Environments
		This implies that we should allow hybrid model not just for task (interpreted) versus data parallelism(compiled)
			Rather should support full data parallelism in interpreter
			NPAC demonstrated a protype HPF interprter at SC93
		Current Web Interpreters include Java TCL and PERL(5) which are optimized for different application domains
			For instance PERL is optimized for documents/files and Java for browsers
		This leads to WebScript Concept of interoperable interpreters optimized for different domains
			WebHPL (High Performance Language) is then script optimized for computing which links compiled HPL modules on tightly coupled MPP's
			This naturally suggests that we can link domain specific systems (e.g. partial differential equation toolkit) to HPF future and WebHPL
	Java and HPF Futures
		Java is a C++ subset which interestingly does not have pointers as these are unsafe in necessary secure metacomputing environment.
		Thus Java has removed the part of C++ which is hardest to parallelize
		Java may not "survive" but if it doesnt something better will! Thus it makes sense to study and experiment with it
		Natural first step is to use Java to build the interpreted "shell" which we called HPFCL for HPF coordination Language.
			This is task parallel script linking HPF modules
		Java is partially compiled as you take basic Java high-level code and compile down to a universal Java machine language. This is very similar to concepts in ANDF (Architecture Neutral Distribution Format) but with a different goal
			Java's model supports universal heterogenous clients linked together in metacomputing 
	VRML and HPF Futures
		VRML -- Virtual Reality Model Language -- is an object oriented database built as a subset of the SGI Inventor System 
		VRML can be considered as another script optimized for graphics but not many interesting processing (compute) capabilities are in current standard
		VRML can be considered as an example of a universal data structure allowing exchange of 3D objects over the Web.
			These objects could either be tanks in a videogame or parts of an aircraft used in large scale simulation
		Thus useful to consider data parallel VRML and building CC++ or HPF(Fortran90) modules to support VRML
		HPCC community should join with the Web to ensure that standards such as VRML can be implemented efficiently either in parallel (maybe a niche) but also in a distributed network (similar issues where HPCC can contribute and clearly very important)