<azonenberg>
i mirrored it in case the original disappears
<whitequark>
404
<azonenberg>
oops
<azonenberg>
00
<azonenberg>
not 000
<pie__>
i can blink leds. now what.
<azonenberg>
metal gate SiGe/SOI/copper interconnect CPLDs on the 150 down to 90nm nodes
<azonenberg>
To launch in 2002+
<azonenberg>
Slide 4, BladeRunner and StarFighter
<whitequark>
my eyes
<azonenberg>
interestingly enough CR-II is not on that list so i suspect one of those later became CR-II
<azonenberg>
It's possible BladeRunner is CoolRunner and StarFighter was the axed "2D grid of PLAs" chip
<azonenberg>
Slide 6 actually hints at that
<azonenberg>
But BladeRunner-II and StarFighter-II never hit the market (150 nm)
<azonenberg>
and one of the two 180nm devices got axed, the other became CR-II
<azonenberg>
In slide 8 you can see they were planning a web based fitter and ide, maybe with server side fitting?
rohitksingh_work has quit [Read error: Connection reset by peer]
rohitksingh_work has joined ##openfpga
rohitksingh_work has quit [Read error: Connection reset by peer]
rohitksingh_work has joined ##openfpga
rohitksingh_wor1 has joined ##openfpga
rohitksingh_work has quit [Ping timeout: 240 seconds]
pie__ has quit [Ping timeout: 240 seconds]
sgstair has quit [Ping timeout: 260 seconds]
sgstair has joined ##openfpga
azonenberg has quit [Ping timeout: 276 seconds]
azonenberg has joined ##openfpga
q3k has joined ##openfpga
sgstair has quit [Ping timeout: 240 seconds]
scrts has quit [Ping timeout: 248 seconds]
eduardo_ has joined ##openfpga
eduardo__ has quit [Ping timeout: 263 seconds]
sgstair has joined ##openfpga
Bike has joined ##openfpga
scrts has joined ##openfpga
scrts has quit [Ping timeout: 260 seconds]
pie__ has joined ##openfpga
pie__ has quit [Ping timeout: 248 seconds]
sgstair has quit [Ping timeout: 248 seconds]
Bike is now known as Bicyclidine
rohitksingh_wor1 has quit [Read error: Connection reset by peer]
pie__ has joined ##openfpga
soylentyellow has quit [Ping timeout: 240 seconds]
pie__ has quit [Ping timeout: 256 seconds]
X-Scale has quit [Read error: Connection reset by peer]
rohitksingh has joined ##openfpga
pie_ has joined ##openfpga
pie_ has quit [Ping timeout: 256 seconds]
pie_ has joined ##openfpga
thallia has quit [Ping timeout: 264 seconds]
thallia has joined ##openfpga
jhol has quit [Quit: Coyote finally caught me]
genii has joined ##openfpga
rohitksingh has quit [Quit: Leaving.]
rohitksingh has joined ##openfpga
noobineer has joined ##openfpga
pie_ has quit [Ping timeout: 268 seconds]
noobineer has quit [Ping timeout: 240 seconds]
rohitksingh has quit [Quit: Leaving.]
cblam has quit []
cblam has joined ##openfpga
pie_ has joined ##openfpga
rohitksingh has joined ##openfpga
bitd has joined ##openfpga
rohitksingh has quit [Quit: Leaving.]
rohitksingh has joined ##openfpga
digshadow has quit [Ping timeout: 264 seconds]
rohitksingh has quit [Read error: Connection reset by peer]
rohitksingh has joined ##openfpga
digshadow has joined ##openfpga
<awygle>
re: AVR Rust from yesterday
<awygle>
In the case of AVR the main reason is that some LLVM codegen bugs prevent you from building core for AVR. These bugs are related to 128-bit integers and formatting floats.
cr1901 has joined ##openfpga
<balrog>
awygle: oof...
<balrog>
gcc has a bug (or rather lack of implementation) of long double that causes MAME not to build on ppc64 anymore
<rqou>
ugh ppc long double
<balrog>
since MAME has started using user-defined literals
<balrog>
and those *must* be long double for float types
<rqou>
if you link with musl, you can force things to use ieee long double rather than ibm long double
<mithro>
And potentially *super* powerful for certain applications
<awygle>
yeah it's clever. very flexible, you're less likely to get choked up by lack of routing resources while still having tons of free LUTs
<mithro>
It's disappointing they only do 800 Mbps LVDS data rate
<awygle>
this first generation's I/O situation is somewhat unfortunate yes. or at least it doesn't map well to any of my use cases.
<mithro>
awygle: Depends on price and availability too...
<mithro>
If they released documentation of their bitstream format it would be interesting and surprising :-P
<azonenberg_work>
awygle: hard wired LVDS, interesting
<azonenberg_work>
i.e. dedicated vs switchable
sgstair has quit [Ping timeout: 246 seconds]
<azonenberg_work>
lol thaaat would be nce
<azonenberg_work>
fat chance though
<awygle>
yeah lol
<awygle>
still it's nice in the abstract to see new players
<awygle>
even if this is only half a new player
<rqou>
whee, my "favorite" thing ever is to forget to power on a peripheral on the stm32
<rqou>
i guess this is why linux gained a clock tree framework
<azonenberg_work>
Also these guys are SMALL
<azonenberg_work>
20-ball WLCSP option available for the smallest parts
<azonenberg_work>
thats like ice40 package density
<azonenberg_work>
if not smaller
<awygle>
yeah but then the big ones get up to 484-ball BGAs
<azonenberg_work>
they have a 3x3mm 49-ball 0.4mm pitch FBGA
<azonenberg_work>
they go up to 676
<azonenberg_work>
... at 0 .6mm pitch
<azonenberg_work>
That's 676 balls in less board area than a xilinx ft256
<awygle>
oo i didn't check the pitch
<azonenberg_work>
sounds fuuuuun to route
<awygle>
RIP PCB costs tho
<azonenberg_work>
i bet the transceivers are dedicated to pcie too
<azonenberg_work>
no general purpose
<azonenberg_work>
i.e. not usable for anything besides pcie
<awygle>
yeah i expect so
sgstair has joined ##openfpga
<rqou>
azonenberg_work: where's my xc2c384/512? :P
digshadow has quit [Ping timeout: 240 seconds]
<azonenberg_work>
rqou: Coming, eventually
<azonenberg_work>
you wanted me to try getting data off them first
<azonenberg_work>
i've been busy :p
<azonenberg_work>
only reason i had time to do those chips is it was a research project for $dayjob
<rqou>
but you just did that "unknown radio?"
<rqou>
ah
<azonenberg_work>
not client hardware hence why i can post it
<rqou>
$dayjob actually pays you to do work in your own lab?
<rqou>
seems a bit sketchy
<azonenberg_work>
this was at the ioa lab
<rqou>
oh ok
<azonenberg_work>
davis did the decap, i did the imaging and analysis
<awygle>
rqou: where's my ice40lp384/640? :p
<rqou>
"Coming, eventually"
<azonenberg_work>
Coming as soon as we get fusion power
<azonenberg_work>
In 10-15 years
<azonenberg_work>
:D
* awygle
goes on a thirty-minute rant about reactor types and polywells
* awygle
's cats give him That Look again
<rqou>
ugh, every time i try to use an adafruit/sparkfun product i find that they always manage to make them suck
<qu1j0t3>
oh?
<rqou>
they always do weird shit like cramming lipo charger chips everywhere
<qu1j0t3>
what happened? i have not bought an adafruit board but i've been tempted a few times. always wondered what the catch was
<rqou>
or using text rather than binary protocols
<rqou>
and adafruit especially loves their shitty discrete level shifters
<azonenberg>
lol
<azonenberg>
eew
<azonenberg>
can you say "slow rise times"?
<rqou>
that's usually not a problem since they tend to put them on i2c
<rqou>
but the circuit they tend to use _only_ works if the "external" side is higher than the "internal" side
<rqou>
and they don't have any bypass jumpers or anything because that would be too easy
<azonenberg_work>
Lol
<azonenberg_work>
eeew
<rqou>
hmm, some of the newer boards are better
<rqou>
they expose enough signals that you can make their level shifter thingy "go away"
digshadow has joined ##openfpga
<rqou>
also adafruit especially likes to give everything a silly cute name so that it's even more confusing what they're talking about
<rqou>
one of the worst ones i found in a quick search is the "Huzzah! Adafruit.io Internet of Things Feather ESP8266"
<rqou>
brilliant /s
<rqou>
it has an esp8266, and i have no idea what the need the rest of that name for
m_w has quit [Ping timeout: 276 seconds]
bitd has quit [Quit: Leaving]
<awygle>
rqou: iirc "feather" is the form factor
<awygle>
(the rest is SEO)
<rqou>
yeah, still a silly name
* azonenberg
hates SEO with a burning passion
<rqou>
azonenberg: "These 10 IoT devices will make you join a botnet! Number 6 will amaze you!"
<cr1901>
level shifter thingy? Also, is there a way to get 5v out from 3.3v in w/ fast rise times?
<rqou>
solution: stop using 5V
<cr1901>
Here's my solution to your solution: Bite me :).
<cr1901>
If I'm using 5v I'm prob interfacing to something vintage, so rise times aren't an issue. But still it "would be nice" to see how fast it can actually go
<awygle>
i mean, probably. do you care about power consumption?
Bicyclidine is now known as Bike
<awygle>
fast NPN to ground, small resistor to 5V good enough?
<awygle>
you could probably drive that into a high-impedance low-capacitance input to a better drive stage that was purely 5V, too
<awygle>
actually you can just drive into a TTL input stage
m_w has joined ##openfpga
<kc8apf>
large LED matrix panels still use 5V SPI up to ~20MHz