hw: Add Reduction capabilities #143
Draft
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Lore0599
force-pushed
the
feature/reduction
branch
9 times, most recently
from
6c02065 to
777e645
Compare
Lore0599
force-pushed
the
feature/reduction
branch
2 times, most recently
from
d420a31 to
27f24dc
Compare
…rrier operation * hw: Extend package with configuration for parallel and offload reduction * hw: Introduce LSB-And operation in ```floo_reduction_arbiter.sv``` * hw: Add parallel reduction support to the ```floo_router.sv``` * hw: Add all parameters relevant for the parallel and offload reduction to the ```floo_nw_router.sv``` / ```floo_router.sv``` without changing the port-list * hw: Add support to the ```floo_nw_chimney.sv``` (merged from commit hash: 4a7f9a1 by raroth)
* hw: add sequential reduction controller to FlooNoC * hw: add generic offload port to all routers * hw: added small parameterizable ALU to FlooNoC which can be used to run simple Integer operation (Add / Mul / Min / Max) * misc: reduction controller supports multiple configuration depending on area / performence requirements. * dataflow: any offload collective flit is brunch off at the input of the router and forwarded to the reduction logic. The final reduced flit will be merged into the normal datastream with an additional slave port on the reduction arbiter / wormhole arbiter. (merged from commit hash: 4a7f9a1 by raroth)
* When it is dedected that the flit will leave the reduction controller after the reduction then the controller starts with the next reduction even if it is not finished yet.
When an endpoint initiates a wide multicast DMA transfer from another endpoint to itself (and possibly other endpoints), the following deadlock occurs. The DMA issues an AR, causing a read burst to come back from the router to the initiator endpoint. When the first read beats arrive, the DMA issues a write burst. This write burst loops back to the initiator endpoint, and may take control of the physical link. As writes lock the link for the entire burst, but the burst cannot complete as it needs the stalling read beats to feed the write burst (a DMA requirement), there is a deadlock. Note, this can happen also without multicast, so long as the system uses loopback.
Lore0599
force-pushed
the
feature/reduction
branch
4 times, most recently
from
f58137f to
8d67524
Compare
Lore0599
force-pushed
the
feature/reduction
branch
from
8d67524 to
ff20b89
Compare
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Add support for reduction logic