Throughput Scalability on General Inference Workload
50,000RPS
SwiftEmbed
Updated 1mo ago
Evaluation Results
| Method | Links | ||
|---|---|---|---|
2025.10 | 50,000 | 10,000 | |
2025.10 | 15,000 | 2,000 | |
2025.10 | 12,000 | 2,000 | |
2025.10 | 10,000 | 10,000 | |
2025.10 | 8,500 | 1,000 | |
2025.10 | 8,000 | 2,000 | |
2025.10 | 4,200 | 800 | |
2025.10 | 4,000 | 1,000 | |
2025.10 | 2,500 | 500 | |
2025.10 | 2,100 | 800 | |
2025.10 | 2,000 | 10,000 | |
2025.10 | 1,200 | 500 | |
2025.10 | 850 | 1,000 | |
2025.10 | 420 | 800 | |
2025.10 | 250 | 500 |