- use smaller regions (increased region count) so there will be less idle time for the workers who completed their tasks
- get rid of some configuration related stuff during initialization which might save a few tens of milliseconds hopefully
- update temperature value parsing instruction order to get benefit of ILP better (hopefully)
* decrease instruction level parallelism
it turns out doing 2 things was too much. perf annotate showed spilling.
* more trickery with latency hiding
* work-stealing, lookp tables, credits
* do not assume gender
* Disable The GC
Cuts off sometimes up to 1 seconds
of runtime on my machine.
* Remove Confusing Byte-Order Parameter
Bytes have no Byte-Order ;)
* Provide More Memory to Run the 10K set
* Fix Comparison Function
* Justin's implementation
* Rename justin to Judekeyser
* Back to previous implementation of vectors
* Reading names as sequences of integers
* Fixing tests
* Scale down the number of NIO workers
---------
Co-authored-by: Justin Dekeyser <justin.dekeyser@Justins-MacBook-Pro.local>
instead of writing result line by line, implemented random.choices for randomisation of multiple stations and writing large batche ot the disk, also instead of "round" just using :.1f which is probably quicker on a large scale, because it's not a mathematical function
* added code
* Fixed pointers bugs
* removed my own benchmark
* added comment on how I handle hash collisions
* executed mwvn clean verify
* made scripts executable & fixed rounding issues
* Fixed way of dealing with hash collisions
* changed method name sameNameBytes to isSameNameBytes
* changes script from sh to bash
* fixed chunking bug
* Fixed bug in chunking when file size is too small
* added Runtime.getRuntime().availableProcessors
* added improvemnts on string copying, calculation of next index of Map in case on collision & improved string comparing
* Some clean up, fine tuning, removing non-supported options, added credit
section and additional comments.
* Put license header year back to 2023 to pass checks.
* Remove static linking (as it requires some more setup on the target
machine).
- split big regions into shared smaller tasks, so the workers complete their own tasks can pick up from the remaining instead of leaving its core idle
- reduce number of executed instructions in the hot path
/**
* Solution based on thomaswue solution, commit:
* commit d0a28599c2
* Author: Thomas Wuerthinger
* Date: Sun Jan 21 20:13:48 2024 +0100
*
* The goal here was to try to improve the runtime of his 10k
* solution of: 00:04.516
*
* With Thomas latest changes, his time is probably much better
* already, and maybe even 1st place for the 10k too.
* See: https://github.com/gunnarmorling/1brc/pull/606
*
* But as I was already coding something, I'll submit just to
* see if it will be faster than his *previous* 10k time of
* 00:04.516
*
* Changes:
* It's a similar idea of my previous solution, that if you split
* the chunks evenly, some threads might finish much faster and
* stay idle, so:
* 1) Create more chunks than threads, so the ones that finish first
* can do something;
* 2) Decrease chunk sizes as we get closer to the end of the file.
*/
* CalculateAverage_pdrakatos
* Rename to be valid with rules
* CalculateAverage_pdrakatos
* Rename to be valid with rules
* Changes on scripts execution
* Fixing bugs causing scripts not to be executed
* Changes on prepare make it compatible
* Fixing passing all tests
* Increase direct memory allocation buffer
* Fixing memory problem causes heap space exception
* Fresh solution to optimize performance of the execution
* Solution without unsafe
* Solution without unsafe
* Solution without unsafe, remove the usage of bytebuffer, passes the create_measurements3 test
* bug fix for 10k test, update also the CreateMeasurements3.java to use '\n' as newline instead of the os value (if it runs on windows it uses crlf and "breaks" the file format )
* new version that should perform way better than the previous one
* removed prepare script for giovannicuccu
* removed some comments
---------
Co-authored-by: Giovanni Cuccu <gcuccu@imolainformatica.it>
* improve speed, thanks to the following improvements:
- loop unrolling and eleminating extra calculations
- eleminating instance level variable access
- quicker equals check, checking long by long chunks instead of bytes
- update GraalVM version to the latest
* faster equals check
* fix equals bug in 10K, more optimizations on equals and calculate hash parts
* New solution optimized for Linux/AMD hardware
* Optimize solution, try to fix 10K bug on native
* Optimize solution, move records to a local field
* test timing
* revert back accidentally pushed code
---------
Co-authored-by: Yavuz Tas <yavuz.tas@ing.com>
* calculate_average_mtopolnik
* short hash (just first 8 bytes of name)
* Remove unneeded checks
* Remove archiving classes
* 2x larger hashtable
* Add "set" to setters
* Simplify parsing temperature, remove newline search
* Reduce the size of the name slot
* Store name length and use to detect collision
* Reduce memory loads in parseTemperature
* Use short for min/max
* Extract constant for semicolon
* Fix script header
* Explicit bash shell in shebang
* Inline usage of broadcast semicolon
* Try vectorization
* Remove vectorization
* Go Unsafe
* Use SWAR temperature parsing by merykitty
* Inline some things
* Remove commented-out MemorySegment usage
* Inline namesMem.asSlice() invocation
* Try out JVM JIT flags
* Implement strcmp
* Remove unused instance variables
* Optimize hashing
* Put station name into hashtable
* Reorder method
* Remove usage of MemorySegment.getUtf8String
Replace with UNSAFE.copyMemory() and new String()
* Fix hashing bug
* Remove outdated comments
* Fix informative constants
* Use broadcastByte() more
* Improve method naming
* More hashing
* Revert more hashing
* Add commented-out code to hash 16 bytes
* Slight cleanup
* Align hashtable at cacheline boundary
* Add Graal Native image
* Revert Graal Native image
This reverts commit d916a42326d89bd1a841bbbecfae185adb8679d7.
* Simplify shell script (no SDK selection)
* Move a constant, zero out hashtable on start
* Better name comparison
* Add prepare_mtopolnik.sh
* Cleaner idiom in name comparison
* AND instead of MOD for hashtable indexing
* Improve word masking code
* Fix formatting
* Reduce memory loads
* Remove endianness checks
* Avoid hash == 0 problem
* Fix subtle bug
* MergeSort of parellel results
* Touch up perf
* Touch up perf
* Remove -Xmx256m
* Extract result printing method
* Print allocation details on OOME
* Single mmap
* Use global allocation arena
* Add commented-out Xmx64m XXMaxDirectMemorySize=1g
* withinSafeZone
* Update cursor earlier
* Better assert
* Fix bug in addrOfSemicolonSafe
* Move declaration lower
* Simplify code
* Add rounding error test case
* Fix DANGER_ZONE_LEN
* Deoptimize parseTemperatureSimple()
* Inline parseTemperatureAndAdvanceCursor()
* Skip masking until the last load
* Conditionally fetch name words
* Cleanup
* Use native image
* Use supbrocess
* Simpler code
* Cleanup
* Avoid extra condition on hot path
* added code
* Fixed pointers bugs
* removed my own benchmark
* added comment on how I handle hash collisions
* executed mwvn clean verify
* made scripts executable & fixed rounding issues
* Fixed way of dealing with hash collisions
* changed method name sameNameBytes to isSameNameBytes
* changes script from sh to bash
* fixed chunking bug
* Fixed bug in chunking when file size is too small
* added Runtime.getRuntime().availableProcessors
* Solution without unsafe
* Solution without unsafe
* Solution without unsafe, remove the usage of bytebuffer, passes the create_measurements3 test
* bug fix for 10k test, update also the CreateMeasurements3.java to use '\n' as newline instead of the os value (if it runs on windows it uses crlf and "breaks" the file format )
---------
Co-authored-by: Giovanni Cuccu <gcuccu@imolainformatica.it>
- faster merge by ignoring empty entries in the map
- enable CDS for faster startup (added `prepare_serkan-ozal.sh` to generate CDS archive in advance)
- some tweaks with JVM options
- optimized result printing
* Read file with multiple virtual threads and process chunks of file data in parallel.
* Updated logic to bucket every chunk of aggs into a vector and merge them into a TreeMap for printing.
* Virtual Thread / File Channels Impl.
* Renamed files with GHUsername.
* Added statement to get vals before updating.
* Added executable permission to the files.
* Latest snapshot (#1)
preparing initial version
* Improved performance to 20seconds (-9seconds from the previous version) (#2)
improved performance a bit
* Improved performance to 14 seconds (-6 seconds) (#3)
improved performance to 14 seconds
* sync branches (#4)
* initial commit
* some refactoring of methods
* some fixes for partitioning
* some fixes for partitioning
* fixed hacky getcode for utf8 bytes
* simplified getcode for partitioning
* temp solution with syncing
* temp solution with syncing
* new stream processing
* new stream processing
* some improvements
* cleaned stuff
* run configuration
* round buffer for the stream to pages
* not using compute since it's slower than straightforward get/put. using own byte array equals.
* using parallel gc
* avoid copying bytes when creating a station object
* formatting
* Copy less arrays. Improved performance to 12.7 seconds (-2 seconds) (#5)
* initial commit
* some refactoring of methods
* some fixes for partitioning
* some fixes for partitioning
* fixed hacky getcode for utf8 bytes
* simplified getcode for partitioning
* temp solution with syncing
* temp solution with syncing
* new stream processing
* new stream processing
* some improvements
* cleaned stuff
* run configuration
* round buffer for the stream to pages
* not using compute since it's slower than straightforward get/put. using own byte array equals.
* using parallel gc
* avoid copying bytes when creating a station object
* formatting
* some tuning to increase performance
* some tuning to increase performance
* avoid copying data; fast hashCode with slightly more collisions
* avoid copying data; fast hashCode with slightly more collisions
* cleanup (#6)
* tidy up
* Initial submission for jonathan_aotearoa
* Fixing typos
* Adding hyphens to prepare and calculate shell scripts so that they're aligned with my GitHub username.
* Making chunk processing more robust in attempt to fix the cause of the build error.
* Fixing typo.
* Fixed the handling of files less than 8 bytes in length.
* Additional assertion, comment improvements.
* Refactoring to improve testability. Additional assertion and comments.
* Updating collision checking to include checking if the station name is equal.
* Minor refactoring to make param ordering consistent.
* Adding a custom toString method for the results map.
* Fixing collision checking bug
* Fixing rounding bug.
* Fixing collision bug.
---------
Co-authored-by: jonathan <jonathan@example.com>
* CalculateAverage_pdrakatos
* Rename to be valid with rules
* CalculateAverage_pdrakatos
* Rename to be valid with rules
* Changes on scripts execution
* Fixing bugs causing scripts not to be executed
* Changes on prepare make it compatible
* Fixing passing all tests
* Increase direct memory allocation buffer
* Fixing memory problem causes heap space exception
* Initial impl
* Fix bad file descriptor error in the `calculate_average_serkan-ozal.sh`
* Disable Epsilon GC and rely on default GC. Because apparently, JIT and Epsilon GC don't play well together in the eval machine for short lived Vector API's `ByteVector` objects
* Take care of byte order before processing key length with bit shift operators
* Fix key equality check for long keys
/**
* Solution based on thomaswue solution, commit:
* commit d0a28599c2
* Author: Thomas Wuerthinger <thomas.wuerthinger@oracle.com>
* Date: Sun Jan 21 20:13:48 2024 +0100
*
* Changes:
* 1) Use LinkedBlockingQueue to store partial results, that
* will then be merged into the final map later.
* As different chunks finish at different times, this allows
* to process them as they finish, instead of joining the
* threads sequentially.
* This change seems more useful for the 10k dataset, as the
* runtime difference of each chunk is greater.
* 2) Use only 4 threads if the file is >= 14GB.
* This showed much better results on my local test, but I only
* run with 200 million rows (because of limited RAM), and I have
* no idea how it will perform on the 1brc HW.
*/
* fix test rounding, pass 10K station names
* improved integer conversion, delayed string creation.
* new algorithm hash, use ConcurrentHashMap
* fix rounding test
* added the length of the string in the hash initialization.
* fix hash code collisions
* cleanup prepare script
* native image options
* fix quardaric probing (no change to perf)
* mask to get the last chunk of the name
* extract hash functions
* tweak the probing loop (-100ms)
* fiddle with native image options
* Reorder conditions in hope it makes branch predictor happier
* extracted constant
* Improve hash function
* remove limit on number of cores
* fix calculation of boundaries between chunks
* fix IOOBE
---------
Co-authored-by: Jason Nochlin <hundredwatt@users.noreply.github.com>
* Contribution by albertoventurini
* Shave off a couple of hundreds of milliseconds, by making an assumption on temperature readings
* Parse reading without loop, inspired by other solutions
* Use all cores
* Small improvements, only allocate 247 positions instead of 256
---------
Co-authored-by: Alberto Venturini <alberto.venturini@accso.de>
* Update with Rounding Bugfix
* Simplification of Merging Results
* More Plain Java Code for Value Storage
* Improve Performance by Stupid Hash
Drop around 3 seconds on my machine by
simplifying the hash to be ridicules stupid,
but faster.
* Fix outdated comment
* Dmitry challenge
* Dmitry submit 2.
Use MemorySegment of FileChannle and Unsafe
to read bytes from disk. 4 seconds speedup in local test
from 20s to 16s.
* tonivade improved not using HashMap
* use java 21.0.2
* same hash same station
* remove unused parameter in sameSation
* use length too
* refactor parallelization
* use parallel GC
* refactor
* refactor
1. Use Unsafe
2. Fit hashtable in L2 cache.
3. If we can find a good hash function, it can fit in L1 cache even.
4. Improve temperature parsing by using a lookup table
* Go implementation by AlexanderYastrebov
This is a proof-of-concept to demonstrate non-java submission.
It requires Docker with BuildKit plugin to build and export binary.
Updates
* #67
* #253
* Use collision-free id lookup
* Use number of buckets greater than max number of keys
* Init Push
* organize imports
* Add OpenMap
* Best outcome yet
* Create prepare script and calculate script for native image, also add comments on calculation
* Remove extra hashing, and need for the set array
* Commit formatting changes from build
* Remove unneeded device information
* Make shell scripts executable, add hash collision double check for equality
* Add hash collision double check for equality
* Skip multithreading for small files to improve small file performance
* final comit
changing using mappedbytebuffer
changes before using unsafe address
using unsafe
* using graalvm,correct unsafe mem implementation
---------
Co-authored-by: Karthikeyans <karthikeyan.sn@zohocorp.com>
* Inline parsing name and station to avoid constantly updating the offset field (-100ms)
* Remove Worker class, inline the logic into lambda
* Accumulate results in an int matrix instead of using result row (-50ms)
* Use native image
* Deploy v2 for parkertimmins
Main changes:
- fix hash which masked incorrectly
- do station equality check in simd
- make station array length multiple of 32
- search for newline rather than semicolon
* Fix bug - entries were being skipped between batches
At the boundary between two batches, the first batch would stop after
crossing a limit with a padding of 200 characters applied. The next
batch should then start looking for the first full entry after the
padding. This padding logic had been removed when starting a batch. For
this reason, entries starting in the 200 character padding between
batches were skipped.
* fast-path for keys<16 bytes
* fix off by one error
the mask is wrong for he 2nd word when len == 16
* less chunks per thread
seems like compact code wins. on my test box anyway.
* Some clean up, small-scale tuning, and reduce complexity when handling longer names.
* Do actual work in worker subprocess. Main process returns immediately
and OS clean up of the mmap continues in the subprocess.
* Update minor Graal version after CPU release.
* Turn GC back to epsilon GC (although it does not seem to make a
difference).
* Minor tuning for another +1%.
