bump ic-stable-structures

[github-actions[bot]]

Note Diffing the performance result against the published result from main branch. Unchanged benchmarks are omitted.

Map

	binary_size	generate 1m	max mem	batch_get 50	batch_put 50	batch_remove 50	upgrade
hashmap	160_221	6_984_044_999	61_987_852	288_670	5_536_856_410	310_195	9_128_784_003
triemap	163_474	11_463_655_150	74_216_172	222_926	549_435	540_205	13_075_158_546
rbtree	158_149	5_979_229_900	57_996_060	88_905	268_573	278_352	5_771_880_608
splay	159_956	11_568_250_103	53_995_996	552_014	581_765	810_321	3_722_474_749
btree	187_897	8_224_242_789	31_104_012	277_542	384_171	429_041	2_517_941_583
zhenya_hashmap	160_509	2_201_622_562	22_773_100	48_627	61_839	70_872	2_695_448_620
btreemap_rs	478_816 ($\textcolor{red}{0.25\%}$)	1_651_590_463	27_590_656	66_862	112_477	76_234	2_660_975_735 ($\textcolor{green}{-0.00\%}$)
imrc_hashmap_rs	482_941 ($\textcolor{red}{0.66\%}$)	2_392_906_831	244_973_568	32_763	163_245	98_394	5_191_575_314 ($\textcolor{green}{-0.00\%}$)
hashmap_rs	469_257 ($\textcolor{red}{0.27\%}$)	403_296_648	73_138_176	16_851	21_680	20_263	1_144_828_112 ($\textcolor{red}{0.00\%}$)

Priority queue

	binary_size	heapify 1m	max mem	pop_min 50	put 50	pop_min 50.1	upgrade
heap	147_638	4_684_519_403	29_995_956	511_505	186_471	487_225	2_655_609_909
heap_rs	462_981 ($\textcolor{green}{-0.19\%}$)	121_602_221	18_284_544	51_661	18_245	51_802	440_739_960 ($\textcolor{green}{-0.00\%}$)

Growable array

	binary_size	generate 5k	max mem	batch_get 500	batch_put 500	batch_remove 500	upgrade
buffer	151_004	2_082_623	65_644	73_092	671_517	127_592	2_474_639
vector	152_551	1_588_260	24_580	105_191	149_932	148_094	3_844_445
vec_rs	460_777 ($\textcolor{red}{0.24\%}$)	265_683	1_310_720	13_014	25_363	21_247	2_743_803 ($\textcolor{green}{-0.00\%}$)

Stable structures

	binary_size	generate 50k	max mem	batch_get 50	batch_put 50	batch_remove 50	upgrade
btreemap_rs	478_816 ($\textcolor{red}{0.25\%}$)	70_026_986	2_555_904	57_181	86_494	75_309	113_837_919 ($\textcolor{green}{-0.00\%}$)
btreemap_stable_rs	488_874 ($\textcolor{red}{2.13\%}$)	4_608_383_725 ($\textcolor{red}{9.09\%}$)	2_031_616 ($\textcolor{green}{-22.50\%}$)	2_778_442 ($\textcolor{red}{9.87\%}$)	5_070_568 ($\textcolor{red}{10.10\%}$)	8_637_788 ($\textcolor{red}{10.49\%}$)	653_330 ($\textcolor{green}{-0.00\%}$)
heap_rs	462_981 ($\textcolor{green}{-0.19\%}$)	6_139_838	2_293_760	44_362	18_477	44_345	23_149_344 ($\textcolor{green}{-0.00\%}$)
heap_stable_rs	450_949 ($\textcolor{green}{-0.02\%}$)	278_284_330 ($\textcolor{green}{-0.41\%}$)	458_752	2_329_886 ($\textcolor{green}{-0.72\%}$)	240_249 ($\textcolor{green}{-0.38\%}$)	2_312_267 ($\textcolor{green}{-0.73\%}$)	653_433
vec_rs	460_777 ($\textcolor{red}{0.24\%}$)	2_866_886	2_228_224	13_014	14_113	13_710	21_249_880 ($\textcolor{green}{-0.00\%}$)
vec_stable_rs	448_899 ($\textcolor{red}{0.64\%}$)	65_186_207 ($\textcolor{green}{-0.00\%}$)	458_752	59_189 ($\textcolor{red}{0.33\%}$)	77_384 ($\textcolor{green}{-0.00\%}$)	79_580 ($\textcolor{red}{0.25\%}$)	653_447

Statistics

• binary_size: 0.39% [0.04%, 0.74%]
• max_mem: -22.50%
• cycles: 1.80% [0.29%, 3.32%]

SHA-2

	binary_size	SHA-256	SHA-512	account_id	neuron_id
Motoko	173_220	247_480_401	228_033_044	30_017	20_760
Rust	478_975 ($\textcolor{red}{0.22\%}$)	82_511_945 ($\textcolor{green}{-0.00\%}$)	56_525_949 ($\textcolor{green}{-0.00\%}$)	42_419 ($\textcolor{green}{-0.03\%}$)	44_417 ($\textcolor{green}{-0.03\%}$)

Certified map

	binary_size	generate 10k	max mem	inc	witness	upgrade
Motoko	206_484	4_390_019_361	3_430_044	519_711	327_767	225_153_243
Rust	503_666 ($\textcolor{red}{0.62\%}$)	6_199_155_471	2_228_224	983_538	288_311 ($\textcolor{green}{-0.01\%}$)	5_816_677_999 ($\textcolor{green}{-0.00\%}$)

Statistics

• binary_size: 0.42% [-0.84%, 1.67%]
• max_mem: no change
• cycles: -0.01% [-0.02%, 0.00%]

Basic DAO

	binary_size	init	transfer_token	submit_proposal	vote_proposal	upgrade
Motoko	236_862	497_618	16_244 ($\textcolor{green}{-0.56\%}$)	12_672 ($\textcolor{red}{0.02\%}$)	14_136 ($\textcolor{red}{0.01\%}$)	128_910 ($\textcolor{green}{-0.06\%}$)
Rust	782_736 ($\textcolor{red}{0.41\%}$)	547_986 ($\textcolor{green}{-0.02\%}$)	86_609 ($\textcolor{red}{0.01\%}$)	105_996 ($\textcolor{red}{0.03\%}$)	117_902 ($\textcolor{green}{-0.00\%}$)	1_624_429 ($\textcolor{green}{-0.01\%}$)

DIP721 NFT

	binary_size	init	mint_token	transfer_token	upgrade
Motoko	195_127	472_267	22_357	4_729	71_602
Rust	804_410 ($\textcolor{red}{0.59\%}$)	217_220 ($\textcolor{green}{-0.02\%}$)	325_588 ($\textcolor{green}{-0.04\%}$)	78_144 ($\textcolor{green}{-0.00\%}$)	1_797_069 ($\textcolor{green}{-0.05\%}$)

Statistics

• binary_size: 0.50% [-0.07%, 1.06%]
• max_mem: no change
• cycles: -0.05% [-0.13%, 0.02%]

Heartbeat

	binary_size	heartbeat
Motoko	123_696	3_758 ($\textcolor{green}{-49.21\%}$)
Rust	23_839	469

Timer

	binary_size	setTimer	cancelTimer
Motoko	129_966	15_227	1_684
Rust	423_634 ($\textcolor{red}{0.27\%}$)	43_472 ($\textcolor{red}{0.00\%}$)	7_569

Statistics

• binary_size: 0.27%
• max_mem: no change
• cycles: 0.00%

Publisher & Subscriber

	pub_binary_size	sub_binary_size	subscribe_caller	subscribe_callee	publish_caller	publish_callee
Motoko	144_769	131_630	14_651	8_456	10_539	3_669
Rust	458_430 ($\textcolor{red}{0.27\%}$)	511_900 ($\textcolor{red}{0.22\%}$)	51_628 ($\textcolor{red}{0.01\%}$)	34_416 ($\textcolor{red}{0.01\%}$)	74_394 ($\textcolor{green}{-0.00\%}$)	44_008 ($\textcolor{green}{-0.01\%}$)

Statistics

• binary_size: 0.24% [0.09%, 0.39%]

• max_mem: no change

• cycles: 0.00% [-0.01%, 0.01%]

  Overall Statistics

• binary_size: 0.38% [0.18%, 0.59%]

• max_mem: -22.50%

• cycles: 0.83% [0.11%, 1.55%]

[github-actions[bot]]

Note The flamegraph link only works after you merge. Unchanged benchmarks are omitted.

Collection libraries

Measure different collection libraries written in both Motoko and Rust. The library names with _rs suffix are written in Rust; the rest are written in Motoko. The _stable and _stable_rs suffix represents that the library directly writes the state to stable memory using Region in Motoko and ic-stable-stuctures in Rust.

We use the same random number generator with fixed seed to ensure that all collections contain the same elements, and the queries are exactly the same. Below we explain the measurements of each column in the table:

• generate 1m. Insert 1m Nat64 integers into the collection. For Motoko collections, it usually triggers the GC; the rest of the column are not likely to trigger GC.
• max mem. For Motoko, it reports rts_max_heap_size after generate call; For Rust, it reports the Wasm's memory page * 32Kb.
• batch_get 50. Find 50 elements from the collection.
• batch_put 50. Insert 50 elements to the collection.
• batch_remove 50. Remove 50 elements from the collection.
• upgrade. Upgrade the canister with the same Wasm module. For non-stable benchmarks, the map state is persisted by serializing and deserializing states into stable memory. For stable benchmarks, the upgrade takes no cycles, as the state is already in the stable memory.

💎 Takeaways

• The platform only charges for instruction count. Data structures which make use of caching and locality have no impact on the cost.
• We have a limit on the maximal cycles per round. This means asymptotic behavior doesn't matter much. We care more about the performance up to a fixed N. In the extreme cases, you may see an $O(10000 n\log n)$ algorithm hitting the limit, while an $O(n^2)$ algorithm runs just fine.
• Amortized algorithms/GC may need to be more eager to avoid hitting the cycle limit on a particular round.
• Rust costs more cycles to process complicated Candid data, but it is more efficient in performing core computations.

Note

• The Candid interface of the benchmark is minimal, therefore the serialization cost is negligible in this measurement.
• Due to the instrumentation overhead and cycle limit, we cannot profile computations with very large collections.
• The upgrade column uses Candid for serializing stable data. In Rust, you may get better cycle cost by using a different serialization format. Another slowdown in Rust is that ic-stable-structures tends to be slower than the region memory in Motoko.
• Different library has different ways for persisting data during upgrades, there are mainly three categories:
  • Use stable variable directly in Motoko: zhenya_hashmap, btree, vector
  • Expose and serialize external state (share/unshare in Motoko, candid::Encode in Rust): rbtree, heap, btreemap_rs, hashmap_rs, heap_rs, vector_rs
  • Use pre/post-upgrade hooks to convert data into an array: hashmap, splay, triemap, buffer, imrc_hashmap_rs
• The stable benchmarks are much more expensive than their non-stable counterpart, because the stable memory API is much more expensive. The benefit is that they get fast upgrade. The upgrade still needs to parse the metadata when initializing the upgraded Wasm module.
• hashmap uses amortized data structure. When the initial capacity is reached, it has to copy the whole array, thus the cost of batch_put 50 is much higher than other data structures.
• btree comes from mops.one/stableheapbtreemap.
• zhenya_hashmap comes from mops.one/map.
• vector comes from mops.one/vector. Compare with buffer, put has better worst case time and space complexity ($O(\sqrt{n})$ vs $O(n)$); get has a slightly larger constant overhead.
• hashmap_rs uses the fxhash crate, which is the same as std::collections::HashMap, but with a deterministic hasher. This ensures reproducible result.
• imrc_hashmap_rs uses the im-rc crate, which is the immutable version hashmap in Rust.

Map

	binary_size	generate 1m	max mem	batch_get 50	batch_put 50	batch_remove 50	upgrade
hashmap	160_221	6_984_044_999	61_987_852	288_670	5_536_856_410	310_195	9_128_784_003
triemap	163_474	11_463_655_150	74_216_172	222_926	549_435	540_205	13_075_158_546
rbtree	158_149	5_979_229_900	57_996_060	88_905	268_573	278_352	5_771_880_608
splay	159_956	11_568_250_103	53_995_996	552_014	581_765	810_321	3_722_474_749
btree	187_897	8_224_242_789	31_104_012	277_542	384_171	429_041	2_517_941_583
zhenya_hashmap	160_509	2_201_622_562	22_773_100	48_627	61_839	70_872	2_695_448_620
btreemap_rs	478_816	1_651_590_463	27_590_656	66_862	112_477	76_234	2_660_975_735
imrc_hashmap_rs	482_941	2_392_906_831	244_973_568	32_763	163_245	98_394	5_191_575_314
hashmap_rs	469_257	403_296_648	73_138_176	16_851	21_680	20_263	1_144_828_112

Priority queue

	binary_size	heapify 1m	max mem	pop_min 50	put 50	pop_min 50	upgrade
heap	147_638	4_684_519_403	29_995_956	511_505	186_471	487_225	2_655_609_909
heap_rs	462_981	121_602_221	18_284_544	51_661	18_245	51_802	440_739_960

Growable array

	binary_size	generate 5k	max mem	batch_get 500	batch_put 500	batch_remove 500	upgrade
buffer	151_004	2_082_623	65_644	73_092	671_517	127_592	2_474_639
vector	152_551	1_588_260	24_580	105_191	149_932	148_094	3_844_445
vec_rs	460_777	265_683	1_310_720	13_014	25_363	21_247	2_743_803

Stable structures

	binary_size	generate 50k	max mem	batch_get 50	batch_put 50	batch_remove 50	upgrade
btreemap_rs	478_816	70_026_986	2_555_904	57_181	86_494	75_309	113_837_919
btreemap_stable_rs	488_874	4_608_383_725	2_031_616	2_778_442	5_070_568	8_637_788	653_330
heap_rs	462_981	6_139_838	2_293_760	44_362	18_477	44_345	23_149_344
heap_stable_rs	450_949	278_284_330	458_752	2_329_886	240_249	2_312_267	653_433
vec_rs	460_777	2_866_886	2_228_224	13_014	14_113	13_710	21_249_880
vec_stable_rs	448_899	65_186_207	458_752	59_189	77_384	79_580	653_447

Environment

• dfx 0.15.1
• Motoko compiler 0.10.0 (source a3ywvw0a-p5a03qy6-vscbl9j8-qxszbxa6)
• rustc 1.73.0 (cc66ad468 2023-10-03)
• ic-repl 0.5.1
• ic-wasm 0.6.0

  Cryptographic libraries

Measure different cryptographic libraries written in both Motoko and Rust.

• SHA-2 benchmarks
  • SHA-256/SHA-512. Compute the hash of a 1M Wasm binary.
  • account_id. Compute the ledger account id from principal, based on SHA-224.
  • neuron_id. Compute the NNS neuron id from principal, based on SHA-256.
• Certified map. Merkle Tree for storing key-value pairs and generate witness according to the IC Interface Specification.
  • generate 10k. Insert 10k 7-character word as both key and value into the certified map.
  • max mem. For Motoko, it reports rts_max_heap_size after generate call; For Rust, it reports the Wasm's memory page * 32Kb.
  • inc. Increment a counter and insert the counter value into the map.
  • witness. Generate the root hash and a witness for the counter.
  • upgrade. Upgrade the canister with the same Wasm. In Motoko, we use stable variable. In Rust, we convert the tree to a vector before serialization.

SHA-2

	binary_size	SHA-256	SHA-512	account_id	neuron_id
Motoko	173_220	247_480_401	228_033_044	30_017	20_760
Rust	478_975	82_511_945	56_525_949	42_419	44_417

Certified map

	binary_size	generate 10k	max mem	inc	witness	upgrade
Motoko	206_484	4_390_019_361	3_430_044	519_711	327_767	225_153_243
Rust	503_666	6_199_155_471	2_228_224	983_538	288_311	5_816_677_999

Environment

• dfx 0.15.1
• Motoko compiler 0.10.0 (source a3ywvw0a-p5a03qy6-vscbl9j8-qxszbxa6)
• rustc 1.73.0 (cc66ad468 2023-10-03)
• ic-repl 0.5.1
• ic-wasm 0.6.0

  Sample Dapps

Measure the performance of some typical dapps:

• Basic DAO, with heartbeat disabled to make profiling easier. We have a separate benchmark to measure heartbeat performance.
• DIP721 NFT

Note

• The cost difference is mainly due to the Candid serialization cost.
• Motoko statically compiles/specializes the serialization code for each method, whereas in Rust, we use serde to dynamically deserialize data based on data on the wire.
• We could improve the performance on the Rust side by using parser combinators. But it is a challenge to maintain the ergonomics provided by serde.
• For real-world applications, we tend to send small data for each endpoint, which makes the Candid overhead in Rust tolerable.

Basic DAO

	binary_size	init	transfer_token	submit_proposal	vote_proposal	upgrade
Motoko	236_862	497_618	16_244	12_672	14_136	128_910
Rust	782_736	547_986	86_609	105_996	117_902	1_624_429

DIP721 NFT

	binary_size	init	mint_token	transfer_token	upgrade
Motoko	195_127	472_267	22_357	4_729	71_602
Rust	804_410	217_220	325_588	78_144	1_797_069

Environment

• dfx 0.15.1
• Motoko compiler 0.10.0 (source a3ywvw0a-p5a03qy6-vscbl9j8-qxszbxa6)
• rustc 1.73.0 (cc66ad468 2023-10-03)
• ic-repl 0.5.1
• ic-wasm 0.6.0

  Heartbeat / Timer

Measure the cost of empty heartbeat and timer job.

• setTimer measures both the setTimer(0) method and the execution of empty job.
• It is not easy to reliably capture the above events in one flamegraph, as the implementation detail of the replica can affect how we measure this. Typically, a correct flamegraph contains both setTimer and canister_global_timer function. If it's not there, we may need to adjust the script.

Heartbeat

	binary_size	heartbeat
Motoko	123_696	3_758
Rust	23_839	469

Timer

	binary_size	setTimer	cancelTimer
Motoko	129_966	15_227	1_684
Rust	423_634	43_472	7_569

Environment

• dfx 0.15.1
• Motoko compiler 0.10.0 (source a3ywvw0a-p5a03qy6-vscbl9j8-qxszbxa6)
• rustc 1.73.0 (cc66ad468 2023-10-03)
• ic-repl 0.5.1
• ic-wasm 0.6.0

  Publisher & Subscriber

Measure the cost of inter-canister calls from the Publisher & Subscriber example.

	pub_binary_size	sub_binary_size	subscribe_caller	subscribe_callee	publish_caller	publish_callee
Motoko	144_769	131_630	14_651	8_456	10_539	3_669
Rust	458_430	511_900	51_628	34_416	74_394	44_008

Environment

• dfx 0.15.1
• Motoko compiler 0.10.0 (source a3ywvw0a-p5a03qy6-vscbl9j8-qxszbxa6)
• rustc 1.73.0 (cc66ad468 2023-10-03)
• ic-repl 0.5.1
• ic-wasm 0.6.0