wu-arch/vortex
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

performance refactoring - rebalanced stream buffers accross the device to enforce output buffering rule at compoments boudaries, finally resolved block ram R/W collusion discrepencies,

Browse Source
This commit is contained in:
Blaise Tine
2020-12-19 02:45:06 -08:00
parent 29cd2f5dff
commit 4bbd7bf408
76 changed files with 1313 additions and 1098 deletions
Download Patch File Download Diff File Expand all files Collapse all files

5
hw/opae/README
View File

@@ -66,7 +66,7 @@ make ase
./run_ase.sh build_ase_1c ../../benchmarks/opencl/vecadd/vecadd
# modify "vsim_run.tcl" to dump VCD trace
vcd file vortex.vcd
vcd file trace.vcd
vcd add -r /*/Vortex/hw/rtl/*
run -all
@@ -104,8 +104,11 @@ lsof +D build_ase_1c
make -C pipeline clean && make -C pipeline > pipeline/build.log 2>&1 &
make -C cache clean && make -C cache > cache/build.log 2>&1 &
make -C core clean && make -C core > core/build.log 2>&1 &
make -C core8 clean && make -C core8 > core8/build.log 2>&1 &
make -C vortex clean && make -C vortex > vortex/build.log 2>&1 &
make -C top clean && make -C top > top/build.log 2>&1 &
make -C top1 clean && make -C top1 > top1/build.log 2>&1 &
make -C top8 clean && make -C top8 > top8/build.log 2>&1 &
# How to calculate the maximum operating frequency?
200 Mhz -> period = 1/200x10^6 = 5ns