What's the clockrate of the EIB ring bus in CELL?

[j^aws]

Probably missed it somewhere in the mountain of docs, but is the EIB ring bus clocked at the same speed as the CELL? Can't seem to find confirmation on this...

Ta!

 

[nAo]

Probably missed it somewhere in the mountain of docs, but is the EIB ring bus clocked at the same speed as the CELL? Can't seem to find confirmation on this...

AFAIK it's clocked at half the main clock

 

[j^aws]

Probably missed it somewhere in the mountain of docs, but is the EIB ring bus clocked at the same speed as the CELL? Can't seem to find confirmation on this...

AFAIK it's clocked at half the main clock

You don't have a reference/link to that info? Thanks again...

 

[nAo]

You don't have a reference/link to that info? Thanks again...

Each ring is 16 bytes wide and runs at
half of the core clock frequency, so some IBM graphics call it
an 8-byte interface (running at full core frequency)

http://www-306.ibm.com/chips/techlib/techlib.nsf/techdocs/D9439D04EA9B080B87256FC00075CC2D (page 7)

 

[j^aws]

You don't have a reference/link to that info? Thanks again...

Each ring is 16 bytes wide and runs at
half of the core clock frequency, so some IBM graphics call it
an 8-byte interface (running at full core frequency)

http://www-306.ibm.com/chips/techlib/techlib.nsf/techdocs/D9439D04EA9B080B87256FC00075CC2D (page 7)

Ah, thought I'd read it before!

That would be with 4 rings, 64 Bytes/cycle. I guess that is a sustained figure while the claimed 96 Bytes/cycle would be a burst rate below,

http://www.research.ibm.com/cell/