The Delft-Java engine: An introduction

Proceedings 25th EUROMICRO Conference. Informatics: Theory and Practice for the New Millennium

1999

Section: Related Workmentioning

confidence: 89%

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Delft-Java dynamic translation

Glossner

Proceedings 25th EUROMICRO Conference. Informatics: Theory and Practice for the New Millennium

1999

“…Sun states that this provides up to 5x performance improvement over JIT compilers [ 121. Another approach to hardware acceleration is dynamic instruction translation. To the best of our knowledge, the DELFT-JAVA processor [3] is the only processor to incorporate this feature. In hardware assisted dynamic translation, JAVA Virtual Machine instructions are translated onthe-fly into the DELFT-JAVA instruction set.…”

Section: Hardware Approaches To Bytecode Executionmentioning

confidence: 99%

“…After being fetched, instructions are dynamically translated into the DELFT-JAVA instruction set. This is accomplished by making a circular buffer out of the register file [3]. Because the instructions are stored in cache memory as JAVA instructions, branching and method invocation code produced by JAVA compilers will execute properly on the DELFT-JAVA architecture.…”

Section: Operationmentioning

confidence: 99%

Delft-Java link translation buffer

Glossner¹,

Proceedings. 24th EUROMICRO Conference (Cat. No.98EX204)

We describe the hardware support in the DELFT-JAVA processor which enables e8cient dynamic linking of JAVA programs. The proposed mechanism supports method invocation of dynamically linked classes through the use of a Link Translation Buffer (LTB). Since our Instruction Set Architecture directly supports dynamically linked method invocation, the Link Translation Buffer is architecturally transparent to the executing program. The operation of the LTB is described and preliminary performance results are reported. Method invocation differences between the C+ + programming language and the JAVA programming language are outlined. Preliminary performance results for the Link Translation Buffer suggest that program performance may improve from 1 . l x to 1 . 5~ when a suitable LTB is used to cache frequently utilized methods.

Java signal processing: FFTs with bytecodes

Glossner

Thilo

Concurrency: Pract. Exper.

1998

Self Cite

This paper investigates the possibility of using Java as a language for digital signal processing. We compare the performance of the fast Fourier transform using Java interpreters, compilers, and native execution. To characterize the Java language as a platform for signal processing, we have implemented a traditional FFT algorithm in both C and Java and compared their relative performances. Additionally, we have developed a Tensor algebra FFT library in both Matlab and Java. Each of the Tensor libraries has been coded to exploit the characteristics of the source language. Our results indicate that the latest Sun Solaris 2.6 Java platform can provide performance within 20% to 60% of optimized C code on short FFT computations. On longer FFT computations, Java is about a factor of 2 to 3 times less efficient than optimized C code. We expect this gap to narrow with better compiler technology and direct execution on Java processors such as the Delft‐Java multithreaded processor. © 1998 John Wiley & Sons, Ltd.