Benchmark on Intel Core i5 13400F and NVIDIA GeForce RTX 4070

The following benchmarks were conducted using an Intel Core i5 13400F and NVIDIA GeForce RTX 4070. The index size is represented in kilobytes, while all other metrics are measured in milliseconds (ms).

These benchmarks illustrate the performance trade-offs among the different index types. While HybridTrieBigramInvertedIndex is faster, it comes at the cost of higher indexing time and reduced search accuracy. Meanwhile, GPULinearIndex provides a substantial speed boost while maintaining accuracy, making it a viable option for environments with GPU support.

For optimal performance, select the index type that best fits your application’s needs.

Exact Search Time (ms) (English)

Text Size	Linear	GPU	Bigram
538	0.44	0.44	0.06
829	0.66	0.68	0.06
1712	1.28	1.31	0.07
3001	2.22	2.25	0.08
3748	2.75	2.71	0.09

---

Fuzzy Search Time (ms) (English)

Text Size	Linear	GPU	Bigram
538	10.74	4.62	0.16
829	16.54	4.67	0.21
1712	34.05	5.21	0.28
3001	59.28	4.84	0.41
3748	74.10	6.17	0.42

---

Fuzzy Search Time (ms) (Japanese)

Text Size	Linear	GPU	Bigram
476	3.26	4.16	0.10
789	5.16	4.58	0.10
1305	8.64	4.79	0.15
2394	15.36	5.45	0.18
3020	20.07	4.24	0.19

---

Indexing Time (ms) (English)

Text Size	Linear	GPU	Bigram
538	46.68	42.76	106.06
829	67.92	65.52	165.12
1712	138.18	132.00	368.06
3001	230.44	226.46	646.16
3748	287.44	286.18	821.26

---

Index Size (Gzipped, kbyte) (English)

Text Size	Linear	GPU	Bigram
538	191	191	91
829	294	294	131
1712	607	607	247
3001	1058	1058	411
3748	1324	1324	506

---

Index Size (Gzipped, kbyte) (Japanese)

Text Size	Linear	GPU	Bigram
476	165	165	136
789	272	272	229
1305	452	452	366
2394	837	837	644
3020	1053	1053	813