I am wishing there were a way to use the G parts on my main-board, and the L parts on my other boards, and still use X-links.
I would also probably use a dedicated switch if it existed. As it is the L1-128 can serve that function.
Regards,
Mike
Wishing for a G-link to L-link bridge
-
- Active Member
- Posts: 51
- Joined: Fri Jan 29, 2010 4:36 pm
-
- XCore Expert
- Posts: 956
- Joined: Fri Dec 11, 2009 3:53 am
- Location: Sweden, Eskilstuna
Intresting question.
As I understood it the G and L Xlinks isn't compatible, but on the other hand on the X64 board a L1 chip is used as an XTAG2 to program the G chips over XLINK.
Can someone explain how the L series and G series can be mixed and what cannot be done, as well !?
As I understood it the G and L Xlinks isn't compatible, but on the other hand on the X64 board a L1 chip is used as an XTAG2 to program the G chips over XLINK.
Can someone explain how the L series and G series can be mixed and what cannot be done, as well !?
Probably not the most confused programmer anymore on the XCORE forum.
-
Verified
- XCore Legend
- Posts: 1163
- Joined: Thu Dec 10, 2009 9:20 pm
- Location: Bristol, UK
This is false. Its uses JTAG (essentially a XTAG2 on the board)lilltroll wrote:
on the other hand on the X64 board a L1 chip is used as an XTAG2 to program the G chips over XLINK.
Technical Director @ XMOS. Opinions expressed are my own
-
- XCore Addict
- Posts: 165
- Joined: Wed Feb 10, 2010 2:32 pm
I'm afraid that there are a number of important differences between the XS1-L links and the XS1-G links, which means that you can not reasonably bridge between the two types.lilltroll wrote:Can someone explain how the L series and G series can be mixed and what cannot be done, as well !?
A critical difference is the start up procedure: reset tokens have slightly different meanings for the XS1-L and XS1-G links, and the 'hello' token is introduced in the XS1-L link. This difference makes them incompatible.
Other differences include the introduction of 'credit 32' token in the XS1-L link, and important routing changes which allow non hypercube configurations in the XS1-L.
-
- XCore Expert
- Posts: 754
- Joined: Thu Dec 10, 2009 6:56 pm
It is feasible to implement a bridge using an FPGA or CPLD or an XMOS chip for the matter.
In the latter case you can forget running the links on full speed though as you cannot use the xlink hardware but have to use the ports.
In the latter case you can forget running the links on full speed though as you cannot use the xlink hardware but have to use the ports.