Hacking asie's DSTwo software

[asiekierka]

.

 

[Recorderdude]

Here I'll post my DSTwo apps.

First of all comes a benchmark for the DSTwo (a port of Linpack).
Notice that as the DSTwo uses software floating-point, floating-point operations WILL be slow. (I will try to port a benchmark which does not use floating-point later).

http://64pixels.org/linpack.zip

Your URL produces a 404.

Perhaps you could upload it to GBATEMP's filetrip instead?

Also, you probably already know, but BassAceGold did one of these too awhile ago.

http://gbatemp.net/t253485-system-tester-plugin

 

[BassAceGold]

My system tester was mainly designed to measure the limitation of transferring data through slot 1 where this benchmark will measure computing power.

 

[asiekierka]

.

 

[asiekierka]

.

 

[pcmantinker]

Here I'll post my DSTwo apps.

First of all comes a benchmark for the DSTwo (a port of Linpack).
Notice that as the DSTwo uses software floating-point, floating-point operations WILL be slow. (I will try to port a benchmark which does not use floating-point later).

http://64pixels.org/linpacktwo.zip

My apologies for reviving such an old thread, but doesn't the MIPS have an FPU? I just took a computer architecture class based on the MIPS and learned that there is a floating point unit in MIPS CPUs. I know that the MIPS has instructions to handle floating point numbers on the hardware side so I was just curious whether the SDK implemented these instructions yet. Is the MIPS instruction set for this CPU the same as what one would run in a simulator such as MARS or SPIM? Forgive me if I'm wrong about the instructions, but I was just curious about how this relates to what I learned about the MIPS in my classes this semester.

 

[asiekierka]

.

 

[Boriar]

All links appears to be broken. Please, can you upload they to filetrip or at least check the links?

 

[pcmantinker]

Here I'll post my DSTwo apps.

First of all comes a benchmark for the DSTwo (a port of Linpack).
Notice that as the DSTwo uses software floating-point, floating-point operations WILL be slow. (I will try to port a benchmark which does not use floating-point later).

http://64pixels.org/linpacktwo.zip

My apologies for reviving such an old thread, but doesn't the MIPS have an FPU? I just took a computer architecture class based on the MIPS and learned that there is a floating point unit in MIPS CPUs. I know that the MIPS has instructions to handle floating point numbers on the hardware side so I was just curious whether the SDK implemented these instructions yet. Is the MIPS instruction set for this CPU the same as what one would run in a simulator such as MARS or SPIM? Forgive me if I'm wrong about the instructions, but I was just curious about how this relates to what I learned about the MIPS in my classes this semester.

You're wrong, there are tons of kinds and revisions of MIPS CPUs. This is the JZ4740 which has extra video processing instructions and possibly no floats.

Thanks for the clarification. I wasn't sure which MIPS CPU this was. I'll have to read more about the chip to learn about its specific instruction set and architecture.

EDIT: Here is a list of documents for the JZ4740 for anyone interested: http://www.amebasystems.com/downloads/hard...4720/Jz4740-PM/

Sadly, it looks like there isn't hardware floating point support. It's a shame since software floating point emulation is much slower.