Constantin Fürst
|
fdc72df6de
|
add results for cpu performance and analysis thereof to chapter 3
|
11 months ago |
Constantin Fürst
|
9cd6a41205
|
add benchmark evaluations for software and new evaluations in pdf format
|
11 months ago |
Constantin Fürst
|
d73b1cea68
|
add brute sw path results
|
11 months ago |
Constantin Fürst
|
f2059d4d47
|
use 12 threads for each brute benchmark
|
11 months ago |
Constantin Fürst
|
439ec97c8e
|
remove smart sw copy, add brute sw copy bench
|
11 months ago |
Constantin Fürst
|
24f96247fc
|
re-run cpu peak perf for allnodes with software path correctly set
|
11 months ago |
Constantin Fürst
|
e9090fb482
|
redo benchmarks for allnodes cpu due to mistake in modifying the json files
|
11 months ago |
Constantin Fürst
|
317a74d164
|
benchmark software peak performance for smart and brute force from node 0 for thesis
|
11 months ago |
Constantin Fürst
|
3d83dbba80
|
redo all possible figures in svalable pdf and include these, provide longer image captions and apply some misc recommendations from andre
|
11 months ago |
Constantin Fürst
|
ba9dae8bde
|
export plots from the benchmarks as pdf for scalability and integration with thesis
|
11 months ago |
Constantin Fürst
|
56805a6ad3
|
review of everything written today, some rewording and mostly adding todos to anything that sticks out as less than optimal
|
11 months ago |
Constantin Fürst
|
b8ce6b3add
|
shorten one statement in the waitoncompletion nsd to make it slimmer and therefore easier to read
|
11 months ago |
Constantin Fürst
|
ed019a672e
|
remove inner-loop timings from the benchmark pseudocode structo, these are not evaluated and pollute the diagram
|
11 months ago |
Constantin Fürst
|
8af4f46d0b
|
add unit to y-axis-label for peakthroughput plots
|
11 months ago |
Constantin Fürst
|
594a2e62cf
|
update modified references to chapter 3 which have turned from sections into subsections
|
11 months ago |
Constantin Fürst
|
ada6d8811a
|
continue writing chapter 3 microbenchmarks
|
11 months ago |
Constantin Fürst
|
718ce39693
|
plot the throughput benchmark for source node 0 and destination nodess {8,11,12,15} only and as a bar plot for use in thesis
|
11 months ago |
Constantin Fürst
|
1ec7d438b2
|
rework the multithread benchmark plot to be more compact for use in the thesis
|
11 months ago |
Constantin Fürst
|
ac3288169a
|
continue writing chapter 3, microbenchmarks
|
11 months ago |
Constantin Fürst
|
3e42225852
|
add a more compact version of the benchmark method that only shows single and batch, not multi submission
|
11 months ago |
Constantin Fürst
|
d626c263ff
|
re-configure the submission method benchmark for displaying in the thesis
|
11 months ago |
Constantin Fürst
|
15af959232
|
rewrite the setup section of chapter 2 to contain less but larger (word count) steps
|
11 months ago |
Constantin Fürst
|
96f08fce81
|
add figure visualizing public interface of cachedata and cache to chapter 4 design
|
11 months ago |
Constantin Fürst
|
4e74a79f99
|
remvoe list of tables from the thesis as we do not yet have any
|
11 months ago |
Constantin Fürst
|
d0a439808d
|
update bachelor.pdf with the changes from the last 3 commits
|
11 months ago |
Constantin Fürst
|
de2525ca7c
|
remove unused commands from the preamble
|
11 months ago |
Constantin Fürst
|
711b29e74c
|
finish sections on system setup and dml interface for chapter 2
|
11 months ago |
Constantin Fürst
|
ba9d18d895
|
handle cleardoublepage in bachelor.tex and not the chapter files themselves
|
11 months ago |
Constantin Fürst
|
719e9fa198
|
write the section on programming interface of chapter 2
|
11 months ago |
Constantin Fürst
|
97acc7403c
|
rework chapter 2 with suggestions after review by andre
|
11 months ago |
Constantin Fürst
|
ae28c5a771
|
reset figure numbering at beginning of actual text to get figures starting from 1 and not 2 which was previously the case
|
11 months ago |
Constantin Fürst
|
76e7f192b1
|
add source file for the benchmark structure diagramm
|
11 months ago |
Constantin Fürst
|
2eb2bcdfe9
|
start writing chapter 03
|
11 months ago |
Constantin Fürst
|
c8d4d5c844
|
slightly change the sequence diagram for threading scenario
|
11 months ago |
Constantin Fürst
|
b218280a25
|
add another bullet point to futurework in chapter 7
|
11 months ago |
Constantin Fürst
|
cd78e4730d
|
add a paragraph about hbm in the state chapter
|
11 months ago |
Constantin Fürst
|
5d5a3a6649
|
add a first batch of graphics to implementation chapter
|
11 months ago |
Constantin Fürst
|
7be56e6419
|
fix more spelling mistakes from previously written parts of state and designt, add the generated pdf file to git too
|
11 months ago |
Constantin Fürst
|
e551e6b56c
|
fix spelling mistakes and such in the recent additions
|
11 months ago |
Constantin Fürst
|
27cd0f07d1
|
continue writing the implementation chapter, which is almost complete now
|
11 months ago |
Constantin Fürst
|
bbdb2d9588
|
prevent another type of deadlock that could arise from destroying a CacheData instance that never had tasks assigned to it due to waiting infinitely on it
|
11 months ago |
Constantin Fürst
|
e9785f3f9e
|
prevent deadlocks by setting the cache pointer to the source address in case of errors
|
11 months ago |
Constantin Fürst
|
fa23c6c50a
|
improve the handler distribution and waiting on task completion for cacher
|
11 months ago |
Constantin Fürst
|
41cf89f7a6
|
continue writing the implementation chapter
|
11 months ago |
Constantin Fürst
|
627ab7eec6
|
continue writing the implementation chapter
|
11 months ago |
Constantin Fürst
|
7e48c96828
|
distribute multiple per-node locks for the cache state and allocate them on the respective numa node for fastest possible access
|
11 months ago |
Constantin Fürst
|
cc452fb1a4
|
Merge branch 'dev-caching' into writing
|
11 months ago |
Constantin Fürst
|
1a3cb6dada
|
prettify credit in the readme
|
11 months ago |
Constantin Fürst
|
0ad0e4af04
|
remove the manual build script and add numa and cpu asignment to the execution script
|
11 months ago |
Constantin Fürst
|
6cc49daf89
|
remove all unused files and benchmark methods, adapt the MAX-Benchmark to use the cacher, remove manually-set numa configuration and replace it with dynamically adapting to the configured affinity, add two more template-options to the worker that control whether a is cached as well and whether scanb waits on the caching
|
11 months ago |