7c1824d58c
cross: list: document why we use the flat impl
2022-08-15 18:46:44 -07:00
dcbef02765
cross: backfill test for list Appendable::append
2022-08-15 16:52:03 -07:00
a3f2b0b33f
cross: list: more tests for fold operation
2022-08-15 03:39:22 -07:00
663d657969
cross: list: implement a Fold operation
2022-08-15 03:33:11 -07:00
1a86fb5ca3
cross: list: fold MaybeMeta and Meta into one trait
2022-08-15 02:32:47 -07:00
22051a39f8
cross: list: s/decl_/impl_/: these are implementations, not declarations
2022-08-14 21:04:52 -07:00
35d0e6a96d
cross: list: add an Appendable trait.
...
this isn't tested... hope it works!
2022-08-14 21:03:49 -07:00
6e9b4465cb
impl IntoList for all 100 list types
...
didn't end up using the Chomp stuff.
2022-08-14 21:01:49 -07:00
2225f98ec8
cross: list: implement IntoList on top of ChompTuple
...
this should be easier for macro generation of the impl.
2022-08-14 19:52:37 -07:00
19893157fa
port: legacy sim accessors test to spirv
2022-08-14 19:16:09 -07:00
f61c0aeb00
spirv_backend: document why we don't support f64
2022-08-14 19:15:58 -07:00
1e994a4feb
app: multi_core_inverter: more experiments (clock decay length, drive current, clock length)
2022-08-14 16:24:58 -07:00
f1143ec365
app: multi_core_inverter: don't save meas.csv
...
this boosts perf by like 5x
2022-08-13 16:20:24 -07:00
ee93c22f4a
app: multi_core_inverter: perf: move the stimulus Gating to outside the CurlStimulus
...
the region.contains() logic is much more expensive than the time bounds
check.
this gets an easy 50% perf boost to the ENTIRE simulation
2022-08-13 15:00:56 -07:00
434dc2cbd5
app: multi_core_inverter: de-virtualize the stimuli
...
this gets like a 5% perf gain.
there are obviously cleaner ways to do it (e.g. use a ListVisitor),
but this is only a proof of concept.
given the low perf difference, i'll likely revert this or explore other
options.
2022-08-13 03:51:52 -07:00
bbb8b2b9ae
driver: better APIs around list-based stimuli
2022-08-13 03:51:01 -07:00
df2ccac1d7
cross: list: implement Indexable up to P99
...
was previously P4
2022-08-13 03:50:11 -07:00
858e787c19
driver: allow preserving the Stimuli as a concrete List
2022-08-12 18:03:10 -07:00
40449c4165
stim: implement AbstractStimulus for any List of stimuli
...
note that this doesn't handle zero-length lists (yet).
2022-08-12 17:08:53 -07:00
c2f8acaf24
cross: add a List visitor/walker type
2022-08-12 16:17:26 -07:00
468f5c5225
cross: rename 'VariantHandler' -> 'Visitor'
2022-08-12 15:20:36 -07:00
a74e7fa9a0
Driver: parameterize over the Stimulus type
2022-08-12 14:47:45 -07:00
e141047bec
Driver: simplify the StimuliAdapter
...
it was basically dead code.
2022-08-12 14:42:56 -07:00
31fd83eb34
apps: multi_core_inverter: setup for 4ns clock phases
...
double what they were before. gives more time for things to settle.
2022-08-12 14:42:17 -07:00
658b8616d7
Torus: correct the cross_section_normal method -- again
...
i should add integration tests
2022-08-12 02:23:53 -07:00
084c5bc342
Region: remove Clone from the trait, and also parameterize everything
...
i didn't necessarily *want* to parameterize it all,
but it turned out to be easier to do that than to force all users to
workaround the lack of Clone.
2022-08-12 01:42:19 -07:00
d5fbb4e9b2
Region: remove the Serialization requirement
2022-08-12 00:57:01 -07:00
090b1ca09a
BUGFIX: Torus: don't normalize the cross section normal
...
this would have led to incorrectly scaled current measurements
(but not incorrect current generation). we were likely severely
over-estimating the current.
2022-08-12 00:41:25 -07:00
ae1eb861be
instrument the stimulus evaluation in our sim
...
... stimulus evaluation accounts for like 80% of the execution time 🤦
2022-08-11 22:57:43 -07:00
09bc7492ed
expose diagnostics into the Sim, and capture stimuli evaluation
...
this isn't publicly exposed yet.
2022-08-11 22:43:07 -07:00
e7cc78a947
diagnostics: split into their own file
2022-08-11 22:31:05 -07:00
d379a7b0ee
app: multi_core_inverter: try a related experiment where S0 is initialized to logic low
2022-08-11 22:24:19 -07:00
aa8f474f52
driver: Diagnostics: clean up the impl a bit
2022-08-11 19:04:12 -07:00
4a33912164
driver: abstract the render time measurements behind a Diagnostics api
2022-08-11 18:58:22 -07:00
f7b72a72be
driver: abstract the step diagnostics measurements
2022-08-11 18:41:41 -07:00
a413a4d391
driver: move last_diag_time out of the Diagnostics object
2022-08-11 18:38:36 -07:00
0c9f04981a
driver: relegate more diagnostics formatting to the Diagnostics impl
2022-08-11 18:36:35 -07:00
6f1e1557b3
driver: diagnostics: track the actual number of frames stepped
...
this allows fps-related diagnostics to be meaningful after
serialization / restarts.
2022-08-11 18:29:58 -07:00
e85d38d415
driver: split the Diagnostics out into their own object
...
more diagnostic refactoring to come
2022-08-11 18:27:30 -07:00
831cbfa76c
app: multi_core_inverter: tune the state serializations
...
less frequent (for less disk space), and also save state
in a recoverable manner
2022-08-11 18:22:55 -07:00
1928ad71cd
serializer renderer: gracefully handle the case where we run out of disk space
...
we might still run out of space when writing CSVs and other outputs...
but that's at least less likely,
as long as we serialize the rendering.
2022-08-11 18:21:46 -07:00
c83a44299f
app: multi-core-inverter: fix S4 drive signal specification
...
there was a spuriuous high -> low transition
2022-08-11 15:24:26 -07:00
e23ab9efd7
app: multi_core_inverter: try a 5-stage inverter (each stage inverts)
...
we're diverging from the blog pretty far now.
but it turns out that, because of the inversion in Maxwell's
$\nabla x E = -dB/dT$ equation, the trivial wiring actually leads to
natural inverters.
2022-08-11 03:01:08 -07:00
652621e47a
app: multi_core_inverter: more precise clock management
...
try to control the edges when the clock is release to prevent ringing.
2022-08-10 16:39:56 -07:00
59a4419130
app: multi_core_inverter: more detailed drive cycle
2022-08-10 15:47:28 -07:00
2f91418095
post: add doc-comments for these tools
2022-08-10 14:28:20 -07:00
46a53a4dde
app: multi_core_inverter: fix up the drive sequence
...
see the code comment for explanation.
2022-08-10 01:43:36 -07:00
3998d72d02
app: multi_core_inverter: drive all four cores for four clock cycles
2022-08-10 01:35:42 -07:00
4fe8be8951
when writing Measurements to a CSV, format them in a machine-readable manner
...
i haven't tested the ones which contains commas -- hopefully the CSV
encoder deals with these :-)
2022-08-10 01:34:37 -07:00
8a3a64face
meas: correctly render SI prefixes for negative numbers
...
the previous implementation treated negative numbers as effectively
having unknown magnitude, rendering them without any adjustment.
2022-08-10 01:17:49 -07:00
