Hi,
I am unable to come across any material that critically examines the reasons I must go in for Microblaze or Powerpc. Is processor obsolesence the only reason why an FPGA designer would go in for Microblaze? Could the present PowerPC architecture could become obsolete?
Regards, Sandy
No. To be honest, that's probably the last reason.
More chance of MicroBlaze becoming obsolete first to be honest.
If you can't think of a reason why you must choose one over the other, then it probably doesn't matter.
Cheers, Jon
Jon,
It is true that some soft processors have become obsolete (Nios is now Nois II). So you have reason to doubt that another one would be successful.
However, both MicroBlaze and PicoBlaze are doing very well, azs measured by the thousands of downloads. Each download may be just one person "kicking the tires" or it may be one download that gets placed in tens of thousands of boards, we have no way to know.
We have decided that whatever we support, Power PC, MicroBlaze, or PicoBlaze, it can not change such that old code doesn't work! That is the death of any processor. Rather, all old machines may evolve, but they must evolve in a way that allows old code to move with the changes (similar to Intel and the x86).
A processor needs to exist, it needs an eco-system (compilers, linkers, loaders, debuggers, etc), it needs to meet the needs (speed, power, cost), and it needs to last the intended length of the company that wishes to use it (it is a printer every 6 months? or a new network switch every three years? or a new spacecraft once a decade?).
Many questions. Look up the answers,
Austin
I agree - unless you are in a super long-lifetime industry, obsolescence is probably a long way down the list.
I think it's the old favourites price and performance/features. If you really need a PPC (or MGTs, or lots of DCMs, or...), then you have to be prepared to pay for it. Things may change with V5, but the price disparity of a V4FX vs an S3/S3E is enormous. If you can do the job with a MicroBlaze in a Spartan3/3E you'll save a fortune.
You can clock the PPCs faster, so in terms of raw benchmarking MIPs lies^H^H^H numbers it will always look superior. But, remember there's an entire FPGA sitting next to it - the performance critical stuff should be in hardware anyway!
Earlier this year I did some back of the envelope estimates of replacing a mass-produced PPC single board computer with a V4FX-based solution. In the 1K-10K qty we could buy the entire SBC with memory and interfaces for about the same price as just the V4FX25 device on its own. The need for FPGA fabric wasn't really there, so we didn't proceed.
John
The 2 options are good at different things. The way I look at is if your task is computationally intensive, go for the PPC because you can clock it at hundreds of MHz and run out of on-macro cache. If the task is basically an I/O processor and you're plugging OPB peripherals, go with the microblaze, its faster for that kind of stuff.
snipped-for-privacy@gmail.com wrote:
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.