►
From YouTube: CHIPS Alliance TSC Meeting - 2021-09-22
Description
No description was provided for this meeting.
If this is YOUR meeting, an easy way to fix this is to add a description to your video, wherever mtngs.io found it (probably YouTube).
A
Have
reached
quorum-
and
I
see
michael's
here
too
so
I'd
like
to
call
the
september
meeting
of
the
technical
steering
committee
to
order.
Let's
see
if
I
can
share
my
screen.
Oh
man
just
installed
zoom
on
a
new
laptop,
so
it's
asking
me
for
a
bunch
of
permissions.
A
C
B
You
know
I'd
rather
not
because
I've
got
guys
working
in
my
house
today,
so
I
just
called
in
for
a
little
bit
so
not
convenient.
Sorry.
A
Okay,
so
definitely
have
quorum
yeah.
I
was
just
copying
some
items
from
the
previous
time,
so
I
think
one
thing
that
we
failed
to
do
was
like
pick
a
new
meeting
time
and
we
adjusted
it
this
time,
which
seemed
to
work
better
for
some
people.
C
A
A
All
right,
the
next
item
I
wanted
to
discuss
was
project
graduation
status
and
applications.
The
one
on
my
plate
that
I
wanted
to
do
was
the
sandbox
application
for
the
diplomacy
project.
A
I
sent
out
an
email
about
that
which
I
think
got
three
approvals
which
is
not
sufficient
to
pass,
but
nobody
said
they
wouldn't
approve
it.
So
I
assume
people
just
forgot
about
it,
so
maybe
we
can
just
do
it
in
person
here,
since
we
have
quorum
and
record
that
and
then
add
it
to
github.
A
All
right,
I
I'm
going
to
take
that
as
an
approval.
Then.
Yes,
sorry,
I
was
muted.
Yes,
I
wanted
to
express.
A
A
D
I
have
a
question
about
there's
a
component
called
antler,
but
I'm
using
two
of
the
subsystems.
I
mean
sherlock
and
usgm,
but
that
component
needs
a
fork
for
probably
a
very
long
duration
and
today,
it's
all
still
in
a
in
a
private
organization
which
is
mine.
D
So
this
component,
which
is
a
general
purpose,
parcel
generator
a
fork,
is
at
the
right
place
to
to
put
it
into
the
cheaper
ions.
It's
kind
of
strange
today
that
there
are
two
components
of
chipotle,
and
that
depends
on
something
which
is
hosted
in
a
in
a
private
github,
and
I
would
want
this
to
maybe
come
to
cheap
alliance.
D
But
it's
not
a
it's,
not
a
component,
which
is
literally
it's
a
parser
generator
for
any
kind
of
language.
It's
not
very
specific
to
chips
of
or
what
we
do.
C
F
The
so
the
context
is
a
little
bit.
Antler
generates
code.
That
is
somewhat
slow
and
alarm
has
some
improvements,
and
there
are
some
other
people
that
contribute
improvements
that
make
it
better.
But
it's
right
now
not
in
the
state
where
it
would
be
actually
easy
to
be
mergeable
upstream,
where
we
can
say:
okay
upstream,
take
this
change
partially
because
they
are
opposed
to
this
particular
kind
of
thing.
And
so
I
guess
it
will
require
a
little
bit
of
work
more
towards
upstream
to
get
it
in
there.
F
So,
in
the
meantime,
ellen
had
forked
it
and
his
own
github,
essentially
and
use
that
as
a
sub
module
inside
sherlock
and
yeah
from
what
I
understand,
he
would
like
to
not
necessarily
host
it
there,
but
also
then
trips
alliance
in
particular,
because
it's
not
only
him
contributing
to
it,
but
other
people
contributing
to
sherlock
actually
would
like
to
contribute
there.
And
so
it
would
be
like
a
bit
too
weird
of
a
dependency.
C
C
That's
a
perfectly
workable
situation.
I
think
we
already
have
it
like.
We
have
forks
of
very
later
and
whatnot
in
chips
alliance,
and
it
doesn't
mean
that
chip
science
now
takes
over
the
project.
It's
more
like
the
fork
is
from
chip's
alliance,
because
chipsets
wants
to
have
its
own
kind
of
version
that
it
depends
on
because
perhaps
somewhere
a
random
place
in
the
internet
I
mean
alan's
account
is
not
a
random
place
in
the
internet,
of
course,
but
chips
alliance
makes
more
sense,
of
course.
So
I
absolutely
don't
see
a
problem
to
do
that.
C
A
We
should
maybe
discuss
what
the
criteria
would
be
for
it
like
yeah.
I
think,
there's
something,
there's
some
difference
between
hosting
a
dependency
that
you
want
to
make
changes
to
and
saying
that
like.
Actually,
this
is
down
like
a
chips
alliance
project
right,
I
feel
like
it's
once
you
make
like
a
sandbox
application
for
it
and
say
that
this
is
like
something
that
chips
alliance
is
vetting.
I
think
that's
when,
like
stuff,
like
the
licensing
terms,
begin
to
become
you
know,
an
issue
that
needs
to
be
resolved.
C
Yep
I
mean-
perhaps
one
thing
to
be
noted
is
that
we
might
not
want
to
fork
stuff-
that's
not
on
open
source
licenses,
for
example,
just
so
that
we
don't
like,
send
the
wrong
message
and
stuff
like
that.
But
that's
kind
of
I
think
secondary
and
given
that
we
only
already
have
the
precedent
of
say
very
later,
which
is
gpl
right.
I
don't
think
we
should
have
any
limitations
regarding
like
open
source
projects,
us
dependencies
or
projects
of
interest
that
we
just
mirror.
C
Let's
say
I
would
call
it
like
this
just
so
that
we
don't
call
like,
because
if
we
call
it
fork,
then
it
means
like
forking
is
often
considered
like
a
hostile
activity,
and
we
don't
mean
that
kind
of
working.
C
Might
go
into
our
github
and
see
some
really
confusing
things
and
get
discouraged,
so
that
would
be
my
kind
of
tentative
intuitive
cut-off
point
where
I'd
say:
okay,
let's,
let's,
let's
consider
this
now,
but
as
as
long
as
kind
of
open
source
dependencies
that
people
really
need
and
actually
are
using
in
solid
chip
science
projects.
I
think
we
should
be
pretty
relaxed.
A
Yeah
that
makes
sense
to
me
all
right.
Well
speaking,
of
chips,
alliance
policies,
there
was
this
issue
from
last
time
about
cla
enforcement,
and
I
guess
that
was
henner's
action
item
to
take
care
of.
Did
that
did
that
get
resolved.
It's
kind
of
like
a
project
by
project
thing
right,
so
was
that
just
a
matter
of
getting
it
turned
on.
F
For
free
yeah,
I've
not
worked
on
it.
Last
time
I
was
well.
I
that's.
I
need
to
make
sure
that
I
can
get
cla
approved
because
I'm
in
an
or
like
I'm
in
the
google
organization,
so
I
actually
can't
like
check
the
box
myself,
so
someone
else
has
to
check
it,
and
so
I
put
I
yeah.
I
did
a
fire
and
forget
there
and
I
didn't
know
if
that
actually
get
reserved.
I
have
to
double
check,
so
nothing
happened
since
then.
A
Yeah,
I
think
this
is
something
that
everybody
has
to
kind
of
work
out
within
whatever
their
own
organization
is.
It
is
nice
that
once
you
get
it,
if
your
organization
is
willing
to
approve
it
for
all
contributors,
then
it
just
sort
of
applies
automatically
to
any
further
contributions
from
your
org.
That's
that's
been
helpful
for
us.
A
All
right
next
topic
remaining
from
last
time,
was
talking
about
working
group
structure.
I
don't
know
if
we've
made
any
progress
or
reached
any
particular
decisions
in
the.
In
the
meantime,
we.
C
C
Of
course
tim
has
made
a
bunch
of
comments,
but
like
that's
not
unexpected,
if
anyone
knows
tim
so,
but
apart
from
that,
we
still
need
actual
text.
You
know
yeah
yeah.
How
are
the
groups
sorted
is
actually
a
valid
question.
I
would
probably
go
and
and
reorder
it
a
little
bit
because
because
why
not
alphabetical
orders
seems
to
have
more
sense,
not
no
order
whatsoever.
C
Oh
yeah,
one
x,
10,
is
actually
not
on
the
list
right
now.
Okay,
so
I
just
moved
analog
workgroup
and
it
seems
like
that
fixed
everything.
E
C
It's
you
know
two,
two
or
three
lines
of
text
really
about
what
the
group
is
supposed
to
do
and
I
don't
think
it
needs
to
follow
any
super
formal
guidelines.
I
think
look
at
the
descriptions
for
tools
and
risk
idv
and
system
relog,
and
you
know
all
the
ones
that
do
have
a
description
and
we
just
need
to
make
sure
the
other
ones
have
two
I
mean.
Theoretically,
we
can
push
srini,
for
example
like
okay.
C
C
And
same
goes
for
rocket
action,
so
we
can
cheer
henry
as
he
types
and
then
we
can
also
talk
about
the
the
other
things.
One
of
the
things
that
definitely
needs
discussions
course
because
apparently
course
is
like
non-existent.
C
It's
it's
not
meeting
anymore.
It's
not
doing
much
stuff,
because
swerve
is
just
out
there,
it's
kind
of
being
used,
but
it's
not
extremely.
You
know
popular
with
the
crowds
and
they
just
stopped
meeting
just
like
they
were
just
the
same
people
meeting
and
nobody
kind
of
joining
them
externally.
C
So
the
question
is
kind
of
what
do
we
do
with
you
know
like
defunct
work
groups
I
mean
course
is
like
an
important
part
of
the
ecosystem,
so
we
don't
want
it
to
kind
of
die
on
the
one
hand,
but
like
I'm
wondering
what
to
do
and
to
avoid
like
fake
groups
that,
like
don't
exist
and
people
like
reach
out
to
us
and
say
I
wanna
like
do
something
in
the
chorus
group
and
then
we
have
to
tell
them.
Oh
yeah,
you
know
what
like
this
doesn't
really
exist.
B
We
can
certainly
remove
it
from
any
public
literature.
I
guess
is
one
way
to
put
it
and
then,
if
we
get
folks
interested
in
that,
we
can
certainly
say
that's
available
right.
That
could
be
true
for
any
topic
to
me,
but
I
certainly
shouldn't
I
should
certainly
remove
from
any
slide
deck.
If
you
will
that,
I
show
relative
to
potential
membership.
C
C
Rockets,
which
is
kind
of
cores
and
other
stuff,
of
course
we
have
like-
and
there
is
a
lot
of
interesting
course
out
there
in
the
office
domain
well,.
B
You
know
we
can
just
you
know
I
can
just
I
can
just
leave
it
in
the
quote:
unquote,
advertising
slides
and
you
know,
and
if
somebody
specifically
asked
about
that,
you
know
just
kind
of
dance
around
the
topic
a
bit
and
see
how
you
know
just
trying
to
get
people
interested.
So
that's
fine
too.
I
mean
on
the
positive
side.
Is
it
shows
a
broad
interest
of
what
chips
is
trying
to
do
right
and,
as
you
correctly
know,
that
is
definitely
an
area
of
interest.
C
Because
we
can
delete
it
from
here,
we
can
not
list
it
in
the
tst
repository
and
we
don't
do
it
anyway,
because
course
just
didn't
have
a
description.
Okay,
it's
on
the
list
right,
so
we
need
to
kind
of
remove
it,
which
is
kind
of
visible,
but
perhaps
tough
luck.
You
know
perhaps
there's
nothing
to
be
done
about
it,
and
then
you
know
there
are
cores
like
black
pirate.
Let's
say
right
that
exists
out
there
and
there
are
people
working
on
them
actively
and
washington,
I
think,
is
in
the
process
of
joining.
C
They
will
at
some
point,
join
so
they're
kind
of
one
party
that
could
potentially
be
interested,
and
then
you
know
who
knows
if
amd
joins
and
isn't
kind
of
interested
in
picking
up
course,
alibaba
involved
in
activity
around
you
know,
building
courses
as
well
to
be
honest,
so
kind
of
perhaps
it's
a
bit
kind
of
a
matter
of
animation
like
just
getting
people
to
to
start
meeting,
but
of
course
we
can't
force
them,
so
we
could
just
start
hiding
it
from
our
public-facing
marketing
at
least.
C
B
A
Mute
kind
of
seems
like
it's
a
function
of
meetings
really
like
the
work
group
can
notionally
exist.
If
it
doesn't
have
a
meeting
time,
then
there's
nothing
to
participate
in,
but
that,
like
these,
can
be
somewhat
decoupled
from
each
other.
Like
a
working
group
can
exist
without
a
meeting
time.
It
still
has
a
mit.
It
still
has
a
mailing
list,
presumably
right.
C
The
more
I
think
about
it,
the
more
it's
the
challenge
of
in
most
areas,
except
for
tools,
but
in
most
areas
projects
really
don't
have
a
lot
of
overlap
in
terms
of
like
collaboration.
C
C
So
until
we
have
like
a
huge
community
around
one
specific
implementation,
it's
very
hard
to
form
a
work
group
because,
like
random,
like
two
people,
maintaining
this
core
and
two
people
making
that
core,
they
don't
necessarily
have
much
to
talk
about,
and
so
I
mean
they
might,
of
course,
have
common
problems
and
share
experiences.
And
so
on.
So
I'm
not
saying
it's
completely
impossible
that
such
a
work
group
would
form,
but
it's
certainly
difficult
to
like
animate
an
ecosystem.
C
A
Yeah
and
it's
mostly
about
uncourse,
stuff
or
adding
accelerators
to
it
like
there's,
actually
not
very
much
development
going
on
on
the
actual
core
itself.
Well,
that's
not
true.
The
the
hypervisor
extension
got
added
recently,
but
historically
there
hasn't
been
that
much
like
discussion
of
the
like
core
internals
itself,
so
much
as
like.
Well,
what
can
I
add
to
this?
Additionally.
C
So,
in
a
sense,
we'll
fade
we'll
face
similar
problems
with
interconnects,
and
anything
else
is
that
it's
it's
a
bunch
of
competing
in
a
sense
standards
and
that's
fine,
like
it's.
It's
okay,
to
have
it
like
that,
but
it's
just
difficult
to
get
those
people
to
meet
and
talk
to
each
other
and
you
know
maintain
a
work
group.
So
the
same.
C
And
I
mean
perhaps
it's
inevitable
and
we
can't
do
anything
about
it.
The
question
is:
how
do
we
advertise
like
what
chips
does?
Perhaps
we
should
call
it
like
topical
areas,
you
know
and
or
streams
or
something
I
don't
know,
because
the
word
groups
do
imply
like
there's
a
word
group.
People
are
meeting
and
doing
stuff
together.
B
What
I
mean
od,
sorry
yeah-
I
mean
I
was
just
watching
some
videos
on
odsa
earlier
this
morning
and
that
certainly
is
a
term
that
they
use,
which
is
streams
I
mean.
I
do
agree
that
you
know
the
notion
of
cores.
You
know,
of
course,
work
groups
of
fuel
to
be
able
to
share
things.
I
mean
that
could
be
like
an
isa,
for
example,
or
different
implementation,
approaches
or
different
methodologies,
but
I
don't
know
how
much
that
really
does
get
shared,
and
you
could
argue
that
some
of
that
really
should
belong
in
more
specific
work.
B
Groups
like
an
analog
work
group,
for
example,
could
share
different
methodology,
implementation
methodology,
challenges
and
peneon
presented
some
of
that
yesterday
at
a
conference,
and
I've
asked
to
see
if
they
could
maybe
come
to
the
analog
work
group
but
anyway,
so
it
is
kind
of
a
hard
problem.
C
There's
one
thought
that
we
haven't
exercised
that
I
think
is
very
worthwhile
is
like
work
with
risk
five
on
this
and
I'm
sure
that
the
risk
five
foundation
or
international
now
might
be
interested
in
kind
of
discussing.
You
know
how
to
funnel
people
into
chips
alliance
who
want
to
talk
about
various.
C
C
If
we
wanted
to
kind
of
reboot
a
course
group,
it
might
be
a
place
where,
if
they
told
people
hey,
look,
there's
a
coursework
group
and
chips
alliance
and
you
can
join
it,
then
perhaps
people
would
come
and
you
know,
discuss
their
implementations
of
risk
class.
C
One
thing
that
we
can
do
is
like
a
call
for
contributors
kind
of
thing.
You
know
where
we
just
say
like
hey:
we
have
swerve
but
we're
open
to
having
you
know
more
people
discussing
their
implementations
and
reach
us
out.
C
If
you
want
to
join
the
group,
and
that
would
potentially
be
some
kind
of
mechanism
which
would
be
welcoming
at
one
time
but
wouldn't
kind
of
state
the
untrue
which
is
like
making
it
seem
like,
there's
a
course
worker,
that's
meeting
where
in
actuality
it
isn't,
and
I
was-
I
was
kind
of
completely
convinced
it
was
meeting
by
the
way.
So
even
I
was
fooled,
let
alone
someone
who
has
nothing
to
do
with
chips.
You
know
they
look
at
the
website.
They'll
see
a
something
on
the
course
and
they'll
just
assume.
C
We
have
a
group
which
we
don't
right.
So
perhaps
this
kind
of
wording
where
we
say
hey,
we
have
a
bunch
of
cars
or
looking
to
kind
of
grow.
This
group
here
say:
if
you
have
interest
just
reach
out
to
us
and
then
we
can
also
go
to
risk
five
and
say:
hey
we're.
Looking
for
contributors
in
space,
you
know
we
could
kind
of
enter
this
work
group
reboot
mode
in
a
sense.
B
I
mean
I
certainly
I
think
you
know
some
discussion
on.
You
know.
Risk
five
and
different
implementation
approaches
would
be
an
interesting
topic.
I'm
sure
verification
is
another
topic,
that's
of
interest
as
well,
and
I
don't
know
if
all
this
five
chords
are
synthesized
or
some
people
do
some
type
of
structured
custom
approach
obviously
depends
upon
performance
objectives
for
the
course
right,
but
certainly
there
are
a
myriad
of
ways
of
implementing
such
a
function.
C
C
I
mean
verification,
and
you
know,
insulation
and
stuff,
like
that.
It's
kind
of
in
the
course
context
is
quite
important.
I
mean
those
things
exist,
also
in
context
of
other
encore
ip
of
course,
but
we
could
wrap
such
topics
under
under
course.
In
a
sense,
if
we
thought
about
that
that
way,
you
know.
C
F
E
E
E
F
Things
that
are
relevant
in
the
open
source
world
would
probably
go
beyond
like
an
actual
like
cpu
core
kind
of
thing,
but
I
mean
I'm
thinking
of
what
we
need
to
have
at
some
point
is
some
sort
of
whatever
lip
c
equivalent
of
of
for
for
hardware,
where
there's
like
five
was
defined
and
usb
defined
and
whatever,
and
where
a
good
high
quality
ip
is
developed
that
can
be
used
in
all
kinds
of
chips.
F
In
general,
something
that
is
making
sure
that
we
that
that
we
can
proliferate
open
source,
the
open
source
idea
in
the
hardware
world
and
having
something
as
a
nucleation
point
might
be
good.
Now.
The
question
is:
is
this
right
place
or
not,
or
is
this
the
right
time
already?
I
don't
know,
but
we
should
keep
something
like
that
in
mind.
With
that
in
mind,
I
mean
when
we
choose
a
name
right,
if
we
say
oh
yeah
risk
five,
we
are
already
tying
us
down
to
risk
five.
F
F
If
but
something
something
that
represents
kind
of
open
source
projects
that
can
lead
us
into
a
world
where,
whatever
you
put
together
your
system
on
the
chip
and
whatever
you
have
your
risk,
five
core
and
you
put
your
usbc
part
in
and
the
whatever
the
the
whatever
other
hardware
periphery
that
that
you
need-
and
that
is
like
high
quality
well
tested
industry
grade
well
at
this
point,
would
be
called
ip
but
yeah.
F
E
Challenge
so
usually
in
conferences,
like
icc,
there's
like
a
session
called
process,
cores
and
processors,
and
you
know
I
feel
like
that-
that's
initially
that's.
I
thought
that
was
the
the
initial
goal
like
to
have
a
processors
and
cores
session
and
then
inside
the
session.
You'll
have
different.
You
know,
implementations
right
now,
I'm
not
I'm
not
an
expert
in
this
file
but
like
if
I
want
to
use
a
low
power
core,
I'm
not
sure
which
one
to
use.
C
I
mean
it
kind
of
raised
an
important
point
of
encore
ip,
of
course,
so
we
don't
have
really
a
group
for
it.
I
mean
rockets
the
closest
because
it
covers
more
than
cores,
but
we
don't
have
like
you
know:
high
quality
upcore
ip.
They
could
just
plug
and
play
into
different
things,
and
the
reason
is
that
we
are
working
on
having
this.
As
you
know,
especially
the
dr
controller,
that's
where
we're
probably
the
closest,
but
with
other
things,
it's
still,
you
know
sometimes
has
to
pass
before
those
things
actually
exist.
C
So
spinning
up
a
group
that
just
meets
and
not
their
heads
is
probably
not
what
we
want
to
do
at
this
point,
because
we
we're
seeing
that
already
what
exists
is
already
something
that
we
have
to
manage
and
figure
out
and
name
and
wrap
together
and
so
on.
So,
but
it's
a
valid
remark
like
we.
There
should
be
an
encore
working
group
at
some
point.
C
I
just
don't
know
if
now's
the
time,
because
we
need
to
have
like
specific
companies
and
universities
kind
of
interested.
I
mean
we
could
always.
C
We
could
always
send
around
a
you
know
an
email
saying
you
know
what
kind
of
work
groups
would
you
be
interested
in
having
in
gypsy
lives
right
and
just
do
a
survey,
and
perhaps
people
answer
some
interesting
things
that
we
didn't
think
about.
We
just,
of
course,
have
to
give
them
some
options
and
then
perhaps
also
leave
some
space,
for
you
know
extra
answers,
and
this
is.
F
I
guess
finding
a
name
is
a
good
name
or
whatever
can
be
something
that
could
provide
the
right,
nucleation
point
for
someone
who's
like.
Oh,
I
have
this
thing.
I
don't
really
know
where
it
would
fit,
but
then,
if
they
see
oh
there's
course
and
on
course,
and
causes
kind
of
a
good
name
and
then.
F
F
Mean
it's
in
a
way
I
mean
in
the
software
world
would
be
called
a
standard
library
right
of
things.
That's
what
I
mean
after
we
sorted
out
all
the
little
things
of
actually
making
asics
and
fpgas
work
with
open
source
tools
really
well.
F
That
should
be
one
of
the
next
things,
because,
ideally,
I
would
like
to
have
a
software,
random
software
engineer
or
random,
whatever
engineer
be
able
to
put
together
their
thing
by
doing
regular
software
techniques,
and
that
is
typically,
that
typically
involves
finding
a
bunch
of
open
source
libraries
that
are
interesting
for
whatever
you
want
to
do
and
then
add
whatever
you
need
to
do
for
your
particular
use
case,
and
in
that
sense
it's
sort
of
a
standard
library.
F
But
that
sounds
super
generic
at
this
point
and
I
guess
requires
a
little
bit
more
brainstorming
to
come
up
with
something
that
is
also
catchy
where
people
can
latch
on.
To
I
mean
there's
the
site,
I
don't
know
opencourse.org
or
whatever,
and
they
have
more
than
just
course.
They
have
all
kinds
of
there's
all
kinds
of
things.
I
don't
know
how
well
maintained
that
that
stuff
is
but
it's
something
where
people
could
see.
F
Oh
yeah,
but
there's
the
chips,
alliance,
standard
library,
kind
of
thing
at
some
point
it
can
refer
to
it.
Let's.
F
In
that
way,
yes,
because
there's
no,
because
there's
no
in
that
sense,
there's
no
standard
library
that
comes
with.
I
don't
know,
system,
verilog
or
vhdl
like
this,
but
there's
there's
a
set
of
if
there
was
a
set
of
standard
things
where
you
don't
have
to
think
about
or
yeah.
I
just
use
that
fifo
and
it
has
all
these
parameters.
I
need-
or
I
just
use
that
that
usb
core
or
that
ddr
interface
and.
F
Essentially,
the
kind
of
things
that
vendors
these
days
provide,
but
they
are
also
like
super
proprietary
and
makes
you
that
makes
you
kind
of
locked
into
a
particular
vendor,
and
so
we
want
to
have
that
thing.
That
is
not
vendor
locked
in,
but
it's
independent.
I
don't
know
what
what
do
vendors
call
their
library
kind
of
things,
maybe.
B
We
call
it,
they
call
it
an
ip
catalog
and
you
are
correct.
You
know
whether
the
cadence
or
synopsis-
and
there
are
others
that
that
you
are
locked
in
effectively
and
it
is
expensive.
F
B
I
p
catalog,
I
see
pliers,
so
you
know
that's
why
one
idea
I
had
for
the
group
was
calling
it.
You
know
some
type
of
ip
group
right
but
effectively
to
your
point
that
it
promotes
and
develops
a
collection
of
ip.
If
you
will
that
a
hardware
developer
or
a
software
developer
doesn't
really
matter
but
could
come
in
and
pick
that
up
and
it's
a
trusted
source
right.
That's
one
of
the
reasons
why
you
do
license
it
from
cadence
or
synopsis.
Is
that
it's
a
trusted
source
for
the
ip.
F
Okay,
so
yeah,
we
should
avoid
the
word
ip
wherever
possible,
because
I
think
it's
a
super
bad
choice
of
of
anything
that
has
to
do
with
open
source.
I
mean
it
has
a
super
bad
drive.
F
F
So
but
catalog
but
catalog
seems
to
be
so.
If
it's
commonly
called
catalog,
maybe
we
can
call
it
a
standard
catalog
or
something
so
so
that
that
people
can
see.
Oh
catalog
is
alright.
This
is
if
catalog
is
kind
of
the
name
for
library.
In
the
the
hardware
world
or
more
accepted
name
in
the
hardware
world.
What
library
would
be
in
the
software
world,
we
could
probably
find
names
around
the
catalog
wording.
C
By
the
way,
one
thing
that
kind
of
for
the
encore
stuff,
one
thing
that
people
tend
to
say
is
also
io.
Quite
simply
so
we
find
if
we
find
a
name
for
chorus
if
we
wanted
the
corresponding
name
for,
for
the
other
part
like
io,
can
be
a
a
word
we
could
use.
C
I
like
the
discussions
around
like
catalog
and
so
on.
I
think
it's
kind
of
getting
somewhere.
I
also
think
ip
is
a
bad
word.
F
C
Yeah
periphery
might
also
be
kind
of
thought
of
us
yeah.
It's
it's
a
hard
topic
and
I
don't
think
everyone
has
cracked
it's
entirely
like
it's
just
some
customary
things.
People
say
they
don't
even
always
make
sense
like
the
ip,
but
soc
components
is
another
word
that
came
to
my
mind,
that
is
kind
of
fairly
short
but
accurately
captures
most
like
sure.
Sometimes
people
don't
want
to
build
an
soc,
but
in
practice,
like
99
of
people
do
even
if
you're
building
like
a
stupid,
dumb
chip
that
does
nothing.
B
I
mean,
I
think
in
some
ways
what
I
consider
this:
it's
like
we're
trying
to
create
a
well-stocked
pantry.
If
you
will
I
like
to
cook
right,
so
you
know
it's
like.
We
have
a
lot
of
different.
We
want
to
have
a
lot
of
different
ingredients
that
a
developer
could
come
in
and
say
you
know
I
need
a
bit
of
this.
I
need
a
bit
of
that.
It's
not
necessarily
a
core,
it
could
be
a
periphery,
but
it
could
be
some
other
type
of
helper
function.
F
Now
the
question
is:
is
there
already
something
like
that
existing
is?
Is
there
already
a
set
of
things
that
people
usually
turn
to
when
they're
like?
Oh,
I
need
whatever
hp,
abc
kind
of
no,
not
abc
open
course.
C
Is
the
closest
one,
but
it's
very
very
bad
and
messy,
and
I
wouldn't
recommend
it
for
anything
serious.
Then
there
is
a
then
there's
libra
cores,
which
tries
to
be
a
reboot
of
open
course,
but
kind
of
didn't
succeed
in
that.
In
the
sense
it
really
needs
a
lot
of
investment
and
a
lot
of
people
to
to
pull
it
off,
and
you
know,
liver
cars
was
driven.
Fossey
fossey
is
a
bunch
of
great
people,
but
like
they
couldn't
just
in
their
spare
time
the
little
resources
they
had.
They
couldn't
really
build.
C
C
Then
I
mean
there
is
just
things
like
lidax,
you
know
that's
the
correct
custom
for
fpgas,
but
there
isn't
that
dream
and
that
vision
has
been
thrown
around
many
many
times
and
nobody
has
pulled
over.
F
Okay-
and
that
also
means
okay,
but
that
also
means
at
this
point.
We
should
ask
ourselves:
okay,
who
are
we
to
suggest
something
like
that?
And
we
should
see
how
things
kind
of
gradually
evolve
in
the
in
the
world
and
then,
as
we
do
with
trips
alliance
right
now,
pull
in
potential
very
like
candidates
that
are
that
that
provide
a
good
way
of
doing
things
and
then
being
the
nucleation
point
for
project
potentially
adds
add
more
stuff
to
it.
So.
F
By
the
way,
I
guess
at
this
point
keep
it
in
mind
as
in
like
have
the
open,
catalog
or
whatever.
We
call
it
at
some
point,
but
I
guess
at
this
point
there's
probably
not
too
much
where
you
could
say:
oh
okay!
Well,
let's
just
start
a
little
thing
where
we
pull
in
a
various
things,
I'm
I
don't
know
how
how.
C
Well,
in
any
case,
I
think
we
should
try
to
focus
on
cleaning
up.
You
know
the
mess
that
we
have
right
now
in
a
sense,
so
kind.
Of
course,
I
think
a
good
action
point
is
reaching
out
to
risk
five
and
saying
hey.
We
wanted
to
kind
of
encourage
people
to
join.
You
know
the
course
working
group
in
chips,
because
we
want
to
encourage
these
discussions
and
we
think
we
have
a
good.
C
We
would
be
a
good
home
for
this,
and
perhaps
they
disagree
and
they
say
no,
no
risk
5
needs
to
do
this,
but
I
assume
they
will
say.
Yes,
absolutely
it's
a
great
idea
and
it's
a
further
point
of
collaboration
to
organizations.
So
in
the
latter
case,
they'll
probably
help
us
kind
of
pull
our
ranks
together
and
find
an
actual
group
that
will
meet
if
we
want
to
maintain.
You
know
cars
as
a
as
an
entity.
B
C
Implementation
most
likely,
but
but
yeah
exactly
stuff,
like
verification
or
like
just
common
problems
that
people
run
into
while
kind
of
doing
course
right.
It's
okay!
I
think
it's
a
worthwhile
group
to
have,
but
we
should
at
least
try
to
save
it
and
then
like,
if
literally
nobody,
steps
up,
then
of
course
we'll
we'll
just
have
to
kind
of
wind
it
down
right.
B
So
let
me
I'll
draft
him
a
note,
and
I
will
copy
you
michael
and
henry
on
this,
and
you
know
see
what
what
mark
has
to
say.
Sure.
Okay,.
A
C
No
thank
you
henry
for
the
descriptions.
Yep
no
problem,
useful
tbd
is
for
meetings
for
rocket
right.
C
A
But
I'll
I'll
either
update
you
or
the
document
when
that
happens.
Hopefully,
hopefully
we'll
have
one
this
week
and
I
will
send
out
a
survey
about
a
possible
new
time
for
this
meeting
to
the
mailing
list.
C
C
I
think
that's
what
I
meant
it
doesn't
have
to
be
ideal
for
cars.
I'll
just
write
up
the
short
text
say
that
we're
looking
for
people
to
collaborate
around
course-
and
I
think
that'll
do
for
now
until
we
get
more
input
from
risk,
5.