Speed Test on Fez Hydra

[hagster]

The OP said he wanted more RAM. I love the little cerb40 but it’s a bit skinny on the memory front.

Bang for buck I think the G400 is pretty compabable, but obviously if you don’t need that much bang don’t waste yer buck. The obvious cheaper alternative is to try and eek out enough performance on the G120, perhaps with a bit of RLP sauce in the critical sections.

[Simon_from_Vilnius]

@godefroi

Cerb40 is frikin’ cheap too, smaller, and not saddled with the Gadgeteer sockets. Plus it performs much better. I’d bet it uses significantly less current, too, but I’ve never measured. Now that Cerb40 is premium, it has feature parity with the other premium modules. Just better all around, I’d say.

Have you tried using SD card AND N18 display AND networking stack AND some other thingies? I did, and it cannot handle it.

[Afshin]

@ Simon -

1571ms on Raptor, FW 4.3.5.

Thanks for the update. As others mentioned, now it becomes a performance/cost trade-off.

Did you try native application or Gadgeteer application on Raptor?

[Simon_from_Vilnius]

@Afshin

@ Simon -

1571ms on Raptor, FW 4.3.5.

Thanks for the update. As others mentioned, now it becomes a performance/cost trade-off.

Did you try native application or Gadgeteer application on Raptor?

By saying “native” I assume you mean “plain NETMF”. Yes, it was NETMF.

There’s another option to consider — RLP. With G400 (and G120), you can implement performance-critical section on bare metal using RLP. Requires some work, but I think even G120 would beat Cerberus/Netduino easily…

[Afshin]

@ Simon - Yes, I meant plain NetMF.

I haven’t worked around RLP yet. But, I will start it from the following URL.
https://www.ghielectronics.com/docs/50/rlp

I will examine the integration of some RLP-written methods into my code.

But, I assume the RLP approach is not a general approach, as I think it aims GHI products with ARM processors.

[Simon_from_Vilnius]

@Afshin

@ Simon - Yes, I meant plain NetMF.

I haven’t worked around RLP yet. But, I will start it from the following URL.
https://www.ghielectronics.com/docs/50/rlp

I will examine the integration of some RLP-written methods into my code.

But, I assume the RLP approach is not a general approach, as I think it aims GHI products with ARM processors.

Yes, RLP works only on GHI offering (and Hydra is running only a subset of it). So, if your goal is to overtake COM port completely, then you can only do this on EMX/G120/G400/Cerberus, since Hydra cannot use RLP to handle hardware interrupts. Well, at least officially. In practice, though, bare-metal coding is all about pointers to memory, so, if it’s possible to load custom functions to RAM (and it is possible with Hydra), then you can definitely hack interrupt controller’s registers to replace GHI handler pointers with yours. But to do that, you’ll have to dig deep into SAM9x35 manual, and you might also break something in GHI firmware. Depending on what your salary is, it might be cheaper to just go with G400 or G120.