►
From YouTube: 20220104 FPGA standup
Description
Open Research Institute FPGA Stand-up meeting for 4 Jan 2022.
D
Yep
sure
so
nothing
for
past
couple
of
weeks
I
was
on
vacation
yeah
enjoyed.
My
parents
are
here
from
india,
so
it's
going
well.
Yeah
now
plan
is
as
discussed
earlier.
I
will
get
back
to
my
development
for
end-to-end
workflows.
I
left
it
somewhere
in
the
middle,
where
I
was
testing
thomas's
code,
so
yeah.
So
otto
and
me
were
both
approaching
the
same
problem
from
parallel
path.
So
we'll
continue
on
that.
D
B
B
A
A
B
B
So
what
I've
tried
to
do
is
get
to
the
point
where
we
are
able
to
just
drop
in
vhdl
or
or
verilog,
so
any
sort
of
hdl
that
people
want
to
do
that
they
they
should
be
able
to
just
drop
it
in,
and
so
I'm
used
to
it
being
a
little
bit
more
straightforward
than
it
is
now
so.
The
past
couple
weeks,
I've
been
exploring
all
of
the
different
paths
that
analog
devices
has
for
the
9371,
and
that
includes
like
things
like
the
the
tes
which.
A
B
The
transceiver
in
experimental
or
evaluation
software-
and
this
generates
it
looks
promising
because
it
generates
a
lot
of
c
files
for
you
to
use
and
and
that
you're
supposed
to
be
able
to
go
to
the
sdk
and
vivado
and
and
and
move
forward.
Well,
this
turned
out
to
probably
not
be
the
best
path
forward.
B
It
looks
like
it's
great
and
it
allows
you
to
evaluate
your
waveforms,
but
iio
api
and
the
iio
path
seems
to
be
the
best
and
the
reason
why
I
didn't
do
that
starting
out
first
is
because
two
weeks
ago
I
simply
couldn't
get
the
live
io
to
to
compile
and
to
work
on
choco
cat,
which
now
looking
back,
probably
should
have
pushed
a
little
harder
on
that.
So
what
we've
we've
done
me
and
paul
and
a
few
other
people
have
have
gotten
a
whole
lot
better
about.
B
So
I
think
our
assumption
was
that
you
all
would
be
doing
a
lot
of
this
work,
that
you
would
just
somehow
know
how
to
use
the
equipment
and
that
it
does
not
appear
to
be
the
best
path
forward.
I
think
that
we
need
to
provide
a
lot
more
service
and
support
and
be
a
lot
more
competent
at
how
it
works.
So
that's
what
we
spent
the
break
doing
and
we're
going
to
continue
to
to
to
move
forward,
to
get
to
the
point
where
it's
much
easier
to
drop
in.
B
So
I
think,
there's
a
there's,
also
a
variety
of
different
approaches.
For
example,
the
paul
probably
could
speak
a
lot
more
about
the
nfs
being
added
to
an
sd
card,
so
the
sd
card
image
that
that
they
assume
that
you
will
be
using
for
for
this
sort
of
work.
It
is
sort
of
assumes
that
you're
physically
there
and
you
have
access
to
the
sd
card.
B
But
then
also
we
have
some
advice
about
nbd,
which
is
another
method
of
of
booting
that
that
may
work.
Okay.
So
so
that's
it.
For
my
retrospective,
for
for
me
over
the
next
week,
is
just
simply
make
this
easy
for
everyone
to
use
and
to
really
dig
in
a
lot
harder
on
on
that.
So
that's
what
I'm
going
to
be
using
there's
a
variety
of
blockers
and
there's
that
we
are.
We
have
all
the
resources
that
we
need.
B
We
have
the
equipment
that
we
need
and
we
think
we
have
access
to
everything
at
analog
devices
and
xilinx
that
we
need
in
order
to
solve
the
problems.
I
think
it's
just
a
question
of
us
getting
competent
more
quickly
in
order
to
help
everyone
here.
So
I'm
going
to
put
it
back
to
paul,
to
correct
anything
or
expand
on
anything
that
I've
said.
A
No,
I
believe,
if
you
got
everything
right,
I'm
sorry.
I
forgot
to
to
bring
that
up
myself.
The
nbd
and
nfs
are,
as
far
as
I
can
tell
interchangeable
for
our
purposes.
They
both
have
the
same
characteristics.
You
have
to
set
up
a
server
on
your
your
vm
on
the
mach
on
the
linux
machine,
and
you
have
to
enable
the
kernel
that
you're
booting
to
use
it.
C
B
B
B
E
Okay,
so
really
I
don't
know
because
I
started
watching
the
last
youtube
video
and
in
which
there
was
stated
that
the
nfs
approach
was
a
little
bit
tricky
to
implement,
and
then
I
tried
to
engage
discussion
with
the
with
team
members,
but
I
I
had
no
luck
in
this,
so
I
tried
myself
to
find
an
alternative
and
I
found
experimented
decent
into
my
pc
and
it
seems
to
to
work
fine,
and
so
nobody
asked
me
to
go
this
way.
I
decided
my
my
head.
E
This
is
also
very
useful
for
classroom.
I
have
two
children
and
maybe
having
the
network
phase
system
not
complicated
as
nfs,
but
instead
of
as
a
raw
device,
so
much
lower
level
is
useful
knowledge
also
another
another
topic.
So
if
there
is
a
specific,
if
there
are
no
problem
on
to
nfs,
it's
it's
really
there
and
yeah.
E
I
I
will
be
happy,
then,
to
understand
if
there
is
some
other
stuff
to
to
try.
Otherwise
I
can
go
on
to
nbd
it's
just
for
my
personal
knowledge.
First
of
all,
so
out
of
curiosity,
I
tried
this
this
way,
that's
all
from
yeah
as
a
starting
point.
So
I
didn't
know
what
to
do.
I
decided
to
do
one
thing,
but
if
you
guys
have
a
process
indication,
maybe
I
can
follow
it.
B
Cool
okay
yeah,
I
think,
probably
paul's
the
the
best
one
to
comment
on
on
on
mbd
rather
than
me,
but
I
genuinely
appreciate
your
contribution
and
and
helping
out
your
suggestions
and,
and
so
yeah
like
like
well
I'll,
just
I'll
defer
to
paul
to
make
any
comments
about
mbd
versus
nfs
or
the
the
progress
forward.
I
I
think
that,
having
both
both
options
is
fantastic
and
and
really
appreciate
the
work.
A
Yeah,
as
I
said,
I
believe
that
nfs
and
nbd
are
going
to
be
pretty
much
interchangeable
for
our
purposes
unless
nbd
is
always
built
into
the
kernel,
and
I
don't
think
that's
the
case.
I
did
a
little
15
second
research
into
this
and
found
that
it's
also
a
installable
module.
So
it's
the
same
as
nfs
in
terms
of
our
complexity,
even
though
the
the
implementation
complexity
that
which
we
hopefully
don't
have
to
worry
about
is
is
lower,
so
hold
hold.
F
F
For
that
sorry,
I
I
have
not
completely
update
my
my
my
knowledge
of
all
the
the
projects
right
now,
so
I
just
get
in
with
this
door,
but
I
tried
to
to
understand
better
all
the
all
the
participants
so
right
now
I
begin
to
work
on
the
design
and
I
go
to
vivado
2021
and
now
it
seems
to
that
the
rtr
dvd
s2
encoder
is
in
the
design.
I
need
to
connect
all
that
and
maybe
have
to
understand
some
something
about
clerk
etc.
F
F
F
B
Oh,
thank
you
no
welcome,
and
we
definitely
appreciate
your
experience
and
and
and
all
the
wonderful
things
that
you're
sharing
no
thank
you.
I
look
forward
to
helping
you
achieve
all
of
the
goals.
Is
there
anything
at
the
moment?
That's
blocking
you
or
the
the
resources
that
you
need
other
than
information
and
and
cooperation.
F
B
Yeah
now
that's
a
good
question.
Anybody
want
to
give
some
comments
or
or
pointers
yeah.
What
we're
working
towards
is
an
end-to-end
demo,
so
we're
trying
very
hard
to
get
a
complete
design
on
our
platform
and
it
feels
like
we're
close,
but
there's
all
sorts
of
stuff
that
we
still
need
to
do
so,
I'm
going
to
be
quiet
and
let
other
people
answer.
C
F
F
C
F
B
Yeah,
I
think
that's
true
for
all
of
us,
so
we're
we're
all
very
motivated
to
make
this
work
and
and
work
very
well.
You
know,
but
we
each
have
a
lot
of
learning
to
do
so
and
helping
each
other
and
sharing
and
being
of
mutual
aid
is,
is
the
only
real
way
that
we'll
be
able
to
accomplish
this.
So
all
right,
anshul,
yeah,
see
ya.
Manchel
has
to
go,
and
let's
see,
I
think,
has
everyone
had
a
chance
to
speak.
I
think
that's
the
case.
B
E
E
B
Yeah,
that
was
a
very
strange
experience,
so
the
the
that
was
probably
what
what
convinced
us
that
maybe
the
the
using
the
transceiver
evaluation
software.
This
is
software,
that's
from
analog
devices,
so
they
want
you
to
use
a
tes
which
is
an
application
and
it
you
you
go
in.
It
says
a
graphical
user
interface
and
you
configure
the
the
ad9371.
B
B
What
they
intend
for
you
to
use
is
2019.1
and
and
then
you're
supposed
to
build
your
application.
That
also
addresses
the
9371
on
the
zc1
we're
using
the
zc706,
and
so
when
you
go
to
build
it,
there's
all
these
missing
files,
there's
missing
header
files,
there's
four
of
them
and
you
go
to
the
forum
and
they
just
say:
oh
just
implement
it
yourself,
and
so
I've
never
really
had
this
experience
before,
where
a
ide
and
a
a
tool
tool
chain
provided
by
a
company
just
you
had
to
then
implement
spy.
B
For
this
way
forward,
there's
also
another
way
forward
that
we
experimented
with,
which
was
the
no
os
so
no
operating
system
at
all
method
of
dealing
with
the
9371
and
as
somebody
that
that
grew
up
and
mostly
programmed
in
assembly.
I
thought
this
has
to
be
better
well,
it
wasn't
either
so
using
the
no
os
approach
for
the
9371
means
that
you're
implementing
absolutely
everything
from
scratch,
and
I
think
that
that's
probably
too
much
to
implement
from
scratch.
B
B
That's
why
I
was
asking
about
the
missing
files
just
to
see
if
anybody
else
had
already
done
this,
if
we
already
had
a
set
of
the
headers
and
implementations
in
the
project,
and
the
answer
looks
to
be
no
and
that
we
should
not
be
following
that
that
way
forward.
So
that's
that's.
That's
the
background
on
on
that
particular
investigation.
So
iio
appears
to
be
the
right
way.
E
E
E
So,
in
my
experience
it
and
one
can
create
a
general
method
without
having
to
deal
with
every
kind
of
message
separately,
and
so
the
knowledge
of
the
message
is
left
to
the
to
the
host
connecting
to
the
to
the
board,
and
in
this
way,
then
we
can
somewhat
abstract
and
that
can
have
just
like
a
script
that
read
and
write
in
memory,
because
normally
with
the
software,
we
don't
we
don't
decode
data.
We
are
on
the
control
plane
not
onto
the
data
plane.
E
So
so
one
can
even
think
to
this
to
his
approach,
but
yeah
would
would
be
more
software
than
than
than
so
without
using
linux
and
so
nbd
even
nfs,
maybe
can
be.
Five
system
are
really
not
so
huge,
so
maybe
can
be
loaded
in
ram
it's
requested.
So
also
the
problem
is
completely
skipped
and
other
problems.
We
are
working
on.
B
F
B
Okay,
no,
this
is
a
it.
May
it
cheers
my
heart,
so
I
very
good
outstanding.
Thank
you.
Everyone,
I'll
edit
and
post
the
the
video
so
that
folks,
that
weren't
able
to
be
here
can
can
catch
up.
It's
wonderful
to
be
back
and
it
looks
like
we
have
plenty
to
do
everyone.
Please
ask
questions
and
post
your
work
on
slack
and
we
will
move
forward
over
january
to
to
getting
a
good
and
end
demo.
B
B
You
know
and
that's
that's
the
goal,
and
it's
a
it's
a
ambitious
one,
because
a
lot
of
the
setup
and
and
grappling
with
things
like
dma
access
and
understanding
how
how
this
system
on
chip
works
I've
taken
longer
than
than
we
anticipated,
but
we
do
have
a
lot
of
the
components
working.
So
it's
a
question
of
correctly
integrating
and
learning
how
to
use
the
tool
learning,
how
to
use
the
tools
and
the
resources.
So
so
we're
we're
looking
to
work
hard
through
the
rest
of
january
to
get
end-to-end
demo.
E
Thank
you.
Thank
you,
michelle.
I
think
in
this
purpose,
even
the
emulator
can
help,
because
if
we
are
studying,
then
the
dma,
as
far
as
I
know,
it's
it's
emulated
almost
accurately,
because
the
software
is
work,
as
is
obviously
there
can
be
a
bug
in
into
the
port,
exciting
sporting
also
to
work.
I
I
add
devices
to
qm,
also
I'm
I'm
more
confident
in
working
with
the
with
this
emulator,
and
I
think
it's
very
very,
but
obviously
if
the
problem
is
at
the
fpga
level
here
can
help
in
this.
B
Right,
yeah,
no,
thank
you
all
right.
We
will
close
the
meeting
and
see
you
on
slack
and
we'll
be
back
next
week.
So
that's
yeah!
That's
that's!
Our
next
stand
up.
The
tradition
for
for
agile
management
is
not
is
to
have
this
sort
of
meeting
every
day,
but
for
for
this
sort
of
open
source
work
every
week
is
what
we've
been
doing.