Silicon Investor (SI) -- The First Internet Community

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Microcap & Penny Stocks : Patriot Scientific - PTSC -- Ignore unavailable to you. Want to Upgrade?

To: sdr who wrote (5073)	7/4/1998 10:28:00 AM
From: bob	Respond to of 8581

ALL, Here's an interesting post from USENET: -------------------------------------------------------------------------------- Subject: Re: How About a Java Co-Processor? From: msimon@tefbbs.com Date: 1998/06/10 Message-ID: <357de147.4304472@news.megsinet.net> Newsgroups: comp.lang.java.advocacy [More Headers] [Subscribe to comp.lang.java.advocacy] Check out the Patriot Scientific PSC1000 also called the SHboom. They sell their development board for $499 with Java. And yes they get Pentium 100 like raw performace for 1/100th the power. Of course doing Java compared to the Pentium it SCREAMs. Think of Java notebooks once the software base is there. Designed by Chuck Moore of the first? two stack virtual machine fame. Harris Semiconductor makes the RTX2000 which is a two stack machine designed by Chuck Moore suitable for Java. No Java yet available as far as I know. This chip is on a lot of satelites and other spacefaring objects. You might also want to check out what Chuck is doing at iTV Corp. Then there is Jeff Fox's F21 - not quite ready for production - Jeff is working with Chuck Moore. Or VHDL cores from John Ribble of Sandpipers/Changing Tiger. John is a College Professor - did a core for some people I worked for. Small & reasonably fast. John of course worked with Chuck Moore. Or the MuP21 (predicessor of the iTV machine and F21) from Dr. Ting of Offete enterprises. Dr Ting also has VHDL cores. And of course Dr. Ting works with Chuck Moore. There is a lot of activity in this field. Its just not too well known. Simon - BTW I have seen Chuck Moore in the halls of Novix but I have not actually spoken to him. ---------------------------------------------------------------------------------------------------------------------------- Andrew Carpenter <kagato@auslink.net> wrote: >Kevin Swan wrote: >> >> Why spend $200 on a CPU optimized for Java when for the same price >> you can get an AMD K6 CPU optimized for everything? The reason >> we're not seeing these chips is becaues they're not practical. To >> really blast through Java code would require: >> 1. Lots of memory - Why not let other apps use it too? Put >> it on your motherboard! >> 2. A fast CPU - This means cash, and if you're going to spend >> cash on a processor, why not upgrade the CPU in your whole >> system? CPUs are cheap because there is competition. >> AMD/Cyrix/Intel/Centaur, they all want your business, and >> are slashing prices. Javachips would only come from one >> vendor. >> 3. Access to hardware - Remember, Java programs are programs >> like any other. They'll need to use the system bus, and >> will be competing with other peripherals. This would cost >> you any of the speed benefit you might have gained by >> having a cheap Java VM hardware chip. >> >> A good Java VM with a JIT compiler is already faster than any >> low-end Java chip. People forked over the cash for 386 boards >> for their ATs because they would have needed a board upgrade >> too. Plus, hardware was more expensive back then. Nowadays, >> it would be cheaper just to get a new CPU, or even a new >> CPU and motherboard. > >I don't know what figures you're basing this off... you're saying going >to hardware is slower than writing software? Why are 3D accelerator >cards selling so well then? > >First off you will be seeing the chips, in embedded systems and >probably NCs. They are in development by a number of companies (but >don't ask me to name them, please... :)). What was being questioned here >was, if the chips become available, would "Java-Blaster" cards be >produced? > >To address your points: >1 - why couldn't it use the system memory? It would no doubt use onboard >memory for cache etc too, but what prevents access to system memory? >2 - the java chips I was reading about were planned as very low power, >and I think they ran at a surprisingly low cycle rate. Speed was due to >many other factors. >3 - I still can't work out this point. Yes, hardware access would be >needed. The "VM" would instead be a tiny library that loaded the classes >to the daughterboard. But in what possible way does this give you a huge >performance hit? > >I don't know if the daughterboard idea will catch on anyway... but I >think that within the year you'll be seeing hardware running java, >native. > >Andrew >