SI
SI
discoversearch

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
PTSC 0.589+7.6%Jan 14 4:00 PM EST

 Public ReplyPrvt ReplyMark as Last ReadFilePrevious 10Next 10PreviousNext  
To: Jon Tara who wrote (4099)12/29/1997 10:55:00 PM
From: Benedict Arnold  Read Replies (2) of 8581
 
Jon, get over it.

The people who are going to buy the chip aren't some dumb schmucks who wouldn't know an ALU from and FPU. They will know the difference.

As for the stockholders, I being one, don't care just how it executes Java, just so that it does, executes correctly, and does it at low cost. Hell, if this were the ARM and it coincidentially "ran" Java as well, I'd be just as pleased.

As to the numbers, 38% of the bytecodes map directly (same or fewer bytes of opcodes) and those accounts for 79% of the bytecodes executed. So, only 21% of the time is a longer instruction sequence being executed to do the equivalent operation as in the JVM. If almost 80% of the bytecodes translatioins executed perform equivalent native instructions on the processor, that's pretty native.

I know stack machines very well, and there are NO other stack machines that even come close. What makes the PSC1000 unique is the opcodes are generally the same size and the local register stack maps to the JVM local variables. Many other operations would be satisfied by any stack machine, but these two factors are key. The code expansion of ony 20% also distinguishes the PSC1000 from any other MPU by at least a factor of 10 (typical expansion is >200%).

The AMD and CYRIX 80x86 clones translate 80x86 opcodes into internal Rops or micro-ops and execute those. While that is a hardware function, granted, that makes the process *completely* invisible, they are still considered to "execute" 80x86 code natively. It really is a matter of semantics of the term "run" and what percentage constitutes "native" (albeit translated). Nobody disputes that. I'm sure that SoftPC users on the Mac don't care that its a translation either, just so that the programs run.

I guarantee you that if you call the company or get any of their literature that they are quite up front about the fact that a bytecode to opcode translation occurs. Taking one sentence from a "sound bite" in a press release and determining that a company is not truthful is both naive and extreme.

Go back to sleep.

Benedict Arnold
Report TOU ViolationShare This Post
 Public ReplyPrvt ReplyMark as Last ReadFilePrevious 10Next 10PreviousNext