4 ms·
With Java being a sort of common API for lots of programmers to reach the hardware, I have always wondered if we would ever saw hardware that would directly imp
by BenoitP 11y ago
With Java being a sort of common API for lots of programmers to reach the hardware, I have always wondered if we would ever saw hardware that would directly implement the JVM bytecode, or even JavaSE.
Apparently Altera has something called JVXtreme[1], which claim x55 peak and x15 sustained performance increase.
From the pdf: "JVXtreme accelerates the actual execution of Java by executing 87 of the most commonly used Java byte codes in hardware."
[1] https://www.altera.com/en_US/pdfs/products/ip/ampp/documents/jvxtreme-java_acc.pdf https://www.altera.com/en_US/pdfs/products/ip/ampp/documents...
- solarexplorer 11y agoPerformance increase over a JVM _interpreter_. They admit this is slower than compiled code. It would be interesting to see data on the complete Cost/Performance/Power/Memory tradeoff relative to a state-of-the-art JIT compiler.
- tjdetwiler 11y agoARM has something similar: http://www.arm.com/products/processors/technologies/jazelle.php http://www.arm.com/products/processors/technologies/jazelle....
- makomk 11y agoHad. I believe all modern ARM cores have totally dropped support for executing Java bytecode natively in favour of traditional JIT compilers.
- new299 11y agoThere were lots of people doing and trying to do this earlier in Java's development: http://en.wikipedia.org/wiki/Java_processor http://en.wikipedia.org/wiki/Java_processor http://en.wikipedia.org/wiki/Jazelle http://en.wikipedia.org/wiki/Jazelle I've not heard so much about those projects lately.
- kevinnk 11y agoIt turns out that while Java byte code is very nice as a compiler target, it's horrible as an instruction set. Theres been quite a few experiments with hardware accelerated java byte code, and all of them turn out to be slower then just targeting a simpler instruction set. (Something that's been played out before in CISC vs RISC)