---
tool: Crawlinx SEO audit
source: https://crawlinx.com
domain: lincklin.cn
report_url: "https://crawlinx.com/audit/lincklin.cn"
crawl_id: 20260712T122856-0307f359
health_score: 66
generated_at: 2026-07-12T18:21:30Z
total_pages: 50
html_pages: 49
indexable_pages: 49
errors: 0
warnings: 65
notices: 209
---

# SEO audit: lincklin.cn

**Health score: 66/100** — needs attention

> Audited by [Crawlinx](https://crawlinx.com) — the free technical-SEO crawler. Full report: https://crawlinx.com/audit/lincklin.cn

## Overview

| Metric | Value |
| --- | --- |
| Pages crawled | 50 |
| HTML pages | 49 |
| Indexable | 49 |
| Non-indexable | 0 |
| Orphan pages | 0 |
| Broken internal links | 0 |
| Avg response time | 1362 ms |
| Errors / Warnings / Notices | 0 / 65 / 209 |
| Status codes | 50×200 |

## Categories

| Category | Score | Errors | Warnings | Notices |
| --- | ---: | ---: | ---: | ---: |
| Structured Data | 81 | 0 | 31 | 1 |
| Headings | 90 | 0 | 16 | 0 |
| Content | 92 | 0 | 13 | 11 |
| Duplicate | 98 | 0 | 3 | 0 |
| Canonical | 99 | 0 | 1 | 0 |
| Meta Description | 99 | 0 | 1 | 22 |
| Agent Readiness | 100 | 0 | 0 | 25 |
| Images | 100 | 0 | 0 | 22 |
| Performance | 100 | 0 | 0 | 1 |
| Security | 100 | 0 | 0 | 49 |
| Sitemap | 100 | 0 | 0 | 1 |
| Social | 100 | 0 | 0 | 50 |
| URL | 100 | 0 | 0 | 27 |

## Issues

### Warnings (65)

#### Missing canonical tag

- Rule: `canonical.missing` · Category: Canonical · Occurrences: 1

- [https://lincklin.cn/presentations/ai-asic-web-presentation/](https://lincklin.cn/presentations/ai-asic-web-presentation/)

#### Thin content (low word count)

- Rule: `content.thin` · Category: Content · Occurrences: 13

- [https://lincklin.cn/about/](https://lincklin.cn/about/) — 107 words
- [https://lincklin.cn/blog/ai-coding-asic-presentation/](https://lincklin.cn/blog/ai-coding-asic-presentation/) — 141 words
- [https://lincklin.cn/blog/category/ai-accelerator/](https://lincklin.cn/blog/category/ai-accelerator/) — 78 words
- [https://lincklin.cn/blog/category/eda/](https://lincklin.cn/blog/category/eda/) — 101 words
- [https://lincklin.cn/blog/category/fpga/](https://lincklin.cn/blog/category/fpga/) — 79 words
- [https://lincklin.cn/blog/category/ic-design/](https://lincklin.cn/blog/category/ic-design/) — 76 words
- [https://lincklin.cn/blog/category/risc-v/](https://lincklin.cn/blog/category/risc-v/) — 100 words
- [https://lincklin.cn/links/](https://lincklin.cn/links/) — 103 words
- [https://lincklin.cn/presentations/ai-asic-web-presentation/](https://lincklin.cn/presentations/ai-asic-web-presentation/) — 0 words
- [https://lincklin.cn/projects/ecg-snn-accelerator/](https://lincklin.cn/projects/ecg-snn-accelerator/) — 182 words
- [https://lincklin.cn/research/](https://lincklin.cn/research/) — 168 words
- [https://lincklin.cn/search/](https://lincklin.cn/search/) — 58 words
- [https://lincklin.cn/tags/ai-agent/](https://lincklin.cn/tags/ai-agent/) — 84 words

#### Near-duplicate page content

- Rule: `duplicate.near` · Category: Duplicate · Occurrences: 2

- [https://lincklin.cn/blog/](https://lincklin.cn/blog/) — near-duplicate of 1 page(s): https://lincklin.cn/blog/category/gpu-architecture/
- [https://lincklin.cn/blog/category/gpu-architecture/](https://lincklin.cn/blog/category/gpu-architecture/) — near-duplicate of 1 page(s): https://lincklin.cn/blog/

#### Homepage duplicated across protocol/host variants (no redirect)

- Rule: `duplicate.protocol_host` · Category: Duplicate · Occurrences: 1

- [https://lincklin.cn/](https://lincklin.cn/) — https://www.lincklin.cn/ serves the same homepage content with a 200 and no redirect — Google sees a duplicate. Pick one canonical host/scheme and 301-redirect the others.

#### Missing <h1>

- Rule: `h1.missing` · Category: Headings · Occurrences: 16

- [https://lincklin.cn/about/](https://lincklin.cn/about/)
- [https://lincklin.cn/archive/](https://lincklin.cn/archive/)
- [https://lincklin.cn/blog/](https://lincklin.cn/blog/)
- [https://lincklin.cn/blog/category/ai-accelerator/](https://lincklin.cn/blog/category/ai-accelerator/)
- [https://lincklin.cn/blog/category/eda/](https://lincklin.cn/blog/category/eda/)
- [https://lincklin.cn/blog/category/fpga/](https://lincklin.cn/blog/category/fpga/)
- [https://lincklin.cn/blog/category/gpu-architecture/](https://lincklin.cn/blog/category/gpu-architecture/)
- [https://lincklin.cn/blog/category/ic-design/](https://lincklin.cn/blog/category/ic-design/)
- [https://lincklin.cn/blog/category/risc-v/](https://lincklin.cn/blog/category/risc-v/)
- [https://lincklin.cn/links/](https://lincklin.cn/links/)
- [https://lincklin.cn/presentations/ai-asic-web-presentation/](https://lincklin.cn/presentations/ai-asic-web-presentation/)
- [https://lincklin.cn/projects/](https://lincklin.cn/projects/)
- [https://lincklin.cn/research/](https://lincklin.cn/research/)
- [https://lincklin.cn/search/](https://lincklin.cn/search/)
- [https://lincklin.cn/tags/](https://lincklin.cn/tags/)
- [https://lincklin.cn/tags/ai-agent/](https://lincklin.cn/tags/ai-agent/)

#### Missing meta description

- Rule: `meta_desc.missing` · Category: Meta Description · Occurrences: 1

- [https://lincklin.cn/presentations/ai-asic-web-presentation/](https://lincklin.cn/presentations/ai-asic-web-presentation/)

#### Rich-result schema missing required properties

- Rule: `schema.invalid` · Category: Structured Data · Occurrences: 31

- [https://lincklin.cn/blog/ai-coding-asic-presentation/](https://lincklin.cn/blog/ai-coding-asic-presentation/) — Article missing: image
- [https://lincklin.cn/blog/algorithm-to-rtl-flow/](https://lincklin.cn/blog/algorithm-to-rtl-flow/) — Article missing: image
- [https://lincklin.cn/blog/digital-ic-toolchain/](https://lincklin.cn/blog/digital-ic-toolchain/) — Article missing: image
- [https://lincklin.cn/blog/fpga-nn-dataflow/](https://lincklin.cn/blog/fpga-nn-dataflow/) — Article missing: image
- [https://lincklin.cn/blog/gpu-architecture/00_blackwell_architecture_overview/](https://lincklin.cn/blog/gpu-architecture/00_blackwell_architecture_overview/) — Article missing: image
- [https://lincklin.cn/blog/gpu-architecture/01_gpu_why_exists_cpu_vs_gpu/](https://lincklin.cn/blog/gpu-architecture/01_gpu_why_exists_cpu_vs_gpu/) — Article missing: image
- [https://lincklin.cn/blog/gpu-architecture/02_gpu_program_execution/](https://lincklin.cn/blog/gpu-architecture/02_gpu_program_execution/) — Article missing: image
- [https://lincklin.cn/blog/gpu-architecture/03_gpu_memory_hierarchy/](https://lincklin.cn/blog/gpu-architecture/03_gpu_memory_hierarchy/) — Article missing: image
- [https://lincklin.cn/blog/gpu-architecture/04_sm_instruction_execution/](https://lincklin.cn/blog/gpu-architecture/04_sm_instruction_execution/) — Article missing: image
- [https://lincklin.cn/blog/gpu-architecture/05_branch_predication_warp_divergence/](https://lincklin.cn/blog/gpu-architecture/05_branch_predication_warp_divergence/) — Article missing: image
- [https://lincklin.cn/blog/gpu-architecture/06_data_hazards_sync/](https://lincklin.cn/blog/gpu-architecture/06_data_hazards_sync/) — Article missing: image
- [https://lincklin.cn/blog/gpu-architecture/07_compute_bound_memory_bound/](https://lincklin.cn/blog/gpu-architecture/07_compute_bound_memory_bound/) — Article missing: image
- [https://lincklin.cn/blog/gpu-architecture/08_matrix_multiplication/](https://lincklin.cn/blog/gpu-architecture/08_matrix_multiplication/) — Article missing: image
- [https://lincklin.cn/blog/gpu-architecture/09_neural_network_convolution/](https://lincklin.cn/blog/gpu-architecture/09_neural_network_convolution/) — Article missing: image
- [https://lincklin.cn/blog/gpu-architecture/10_tensor_core_low_precision/](https://lincklin.cn/blog/gpu-architecture/10_tensor_core_low_precision/) — Article missing: image
- [https://lincklin.cn/blog/gpu-architecture/11_cache_memory_transactions/](https://lincklin.cn/blog/gpu-architecture/11_cache_memory_transactions/) — Article missing: image
- [https://lincklin.cn/blog/gpu-architecture/12_block_assignment_occupancy/](https://lincklin.cn/blog/gpu-architecture/12_block_assignment_occupancy/) — Article missing: image
- [https://lincklin.cn/blog/gpu-architecture/13_gpu_scheduling_layers/](https://lincklin.cn/blog/gpu-architecture/13_gpu_scheduling_layers/) — Article missing: image
- [https://lincklin.cn/blog/gpu-architecture/14_multi_gpu_collaboration/](https://lincklin.cn/blog/gpu-architecture/14_multi_gpu_collaboration/) — Article missing: image
- [https://lincklin.cn/blog/gpu-architecture/15_3d_rendering_pipeline/](https://lincklin.cn/blog/gpu-architecture/15_3d_rendering_pipeline/) — Article missing: image
- [https://lincklin.cn/blog/gpu-architecture/16_gpu_chip_block_diagram/](https://lincklin.cn/blog/gpu-architecture/16_gpu_chip_block_diagram/) — Article missing: image
- [https://lincklin.cn/blog/gpu-architecture/17_nvidia_gpu_evolution/](https://lincklin.cn/blog/gpu-architecture/17_nvidia_gpu_evolution/) — Article missing: image
- [https://lincklin.cn/blog/gpu-architecture/18_ptx_sass_compute_capability/](https://lincklin.cn/blog/gpu-architecture/18_ptx_sass_compute_capability/) — Article missing: image
- [https://lincklin.cn/blog/gpu-architecture/19_cuda_compiler_optimization/](https://lincklin.cn/blog/gpu-architecture/19_cuda_compiler_optimization/) — Article missing: image
- [https://lincklin.cn/blog/gpu-architecture/20_nsight_compute_analysis/](https://lincklin.cn/blog/gpu-architecture/20_nsight_compute_analysis/) — Article missing: image
- [https://lincklin.cn/blog/gpu-architecture/21_reduction_optimization/](https://lincklin.cn/blog/gpu-architecture/21_reduction_optimization/) — Article missing: image
- [https://lincklin.cn/blog/gpu-architecture/22_softmax_flashattention/](https://lincklin.cn/blog/gpu-architecture/22_softmax_flashattention/) — Article missing: image
- [https://lincklin.cn/blog/gpu-architecture/23_layernorm_rmsnorm/](https://lincklin.cn/blog/gpu-architecture/23_layernorm_rmsnorm/) — Article missing: image
- [https://lincklin.cn/blog/gpu-architecture/24_transformer_mlp_gemm_fusion/](https://lincklin.cn/blog/gpu-architecture/24_transformer_mlp_gemm_fusion/) — Article missing: image
- [https://lincklin.cn/blog/gpu-architecture/25_sparse_irregular_gpu_limits/](https://lincklin.cn/blog/gpu-architecture/25_sparse_irregular_gpu_limits/) — Article missing: image
- [https://lincklin.cn/blog/gpu-architecture/26_simplified_simt_gpu_design/](https://lincklin.cn/blog/gpu-architecture/26_simplified_simt_gpu_design/) — Article missing: image

### Notices (209)

#### Heading levels skip a level

- Rule: `agent.heading_skip` · Category: Agent Readiness · Occurrences: 24

- [https://lincklin.cn/blog/gpu-architecture/02_gpu_program_execution/](https://lincklin.cn/blog/gpu-architecture/02_gpu_program_execution/) — h4 follows h2 — skips h3
- [https://lincklin.cn/blog/gpu-architecture/03_gpu_memory_hierarchy/](https://lincklin.cn/blog/gpu-architecture/03_gpu_memory_hierarchy/) — h3 follows h1 — skips h2
- [https://lincklin.cn/blog/gpu-architecture/04_sm_instruction_execution/](https://lincklin.cn/blog/gpu-architecture/04_sm_instruction_execution/) — h3 follows h1 — skips h2
- [https://lincklin.cn/blog/gpu-architecture/05_branch_predication_warp_divergence/](https://lincklin.cn/blog/gpu-architecture/05_branch_predication_warp_divergence/) — h3 follows h1 — skips h2
- [https://lincklin.cn/blog/gpu-architecture/06_data_hazards_sync/](https://lincklin.cn/blog/gpu-architecture/06_data_hazards_sync/) — h3 follows h1 — skips h2
- [https://lincklin.cn/blog/gpu-architecture/07_compute_bound_memory_bound/](https://lincklin.cn/blog/gpu-architecture/07_compute_bound_memory_bound/) — h3 follows h1 — skips h2
- [https://lincklin.cn/blog/gpu-architecture/08_matrix_multiplication/](https://lincklin.cn/blog/gpu-architecture/08_matrix_multiplication/) — h3 follows h1 — skips h2
- [https://lincklin.cn/blog/gpu-architecture/09_neural_network_convolution/](https://lincklin.cn/blog/gpu-architecture/09_neural_network_convolution/) — h3 follows h1 — skips h2
- [https://lincklin.cn/blog/gpu-architecture/10_tensor_core_low_precision/](https://lincklin.cn/blog/gpu-architecture/10_tensor_core_low_precision/) — h3 follows h1 — skips h2
- [https://lincklin.cn/blog/gpu-architecture/11_cache_memory_transactions/](https://lincklin.cn/blog/gpu-architecture/11_cache_memory_transactions/) — h3 follows h1 — skips h2
- [https://lincklin.cn/blog/gpu-architecture/12_block_assignment_occupancy/](https://lincklin.cn/blog/gpu-architecture/12_block_assignment_occupancy/) — h3 follows h1 — skips h2
- [https://lincklin.cn/blog/gpu-architecture/13_gpu_scheduling_layers/](https://lincklin.cn/blog/gpu-architecture/13_gpu_scheduling_layers/) — h3 follows h1 — skips h2
- [https://lincklin.cn/blog/gpu-architecture/14_multi_gpu_collaboration/](https://lincklin.cn/blog/gpu-architecture/14_multi_gpu_collaboration/) — h3 follows h1 — skips h2
- [https://lincklin.cn/blog/gpu-architecture/15_3d_rendering_pipeline/](https://lincklin.cn/blog/gpu-architecture/15_3d_rendering_pipeline/) — h3 follows h1 — skips h2
- [https://lincklin.cn/blog/gpu-architecture/16_gpu_chip_block_diagram/](https://lincklin.cn/blog/gpu-architecture/16_gpu_chip_block_diagram/) — h3 follows h1 — skips h2
- [https://lincklin.cn/blog/gpu-architecture/17_nvidia_gpu_evolution/](https://lincklin.cn/blog/gpu-architecture/17_nvidia_gpu_evolution/) — h3 follows h1 — skips h2
- [https://lincklin.cn/blog/gpu-architecture/18_ptx_sass_compute_capability/](https://lincklin.cn/blog/gpu-architecture/18_ptx_sass_compute_capability/) — h3 follows h1 — skips h2
- [https://lincklin.cn/blog/gpu-architecture/19_cuda_compiler_optimization/](https://lincklin.cn/blog/gpu-architecture/19_cuda_compiler_optimization/) — h3 follows h1 — skips h2
- [https://lincklin.cn/blog/gpu-architecture/20_nsight_compute_analysis/](https://lincklin.cn/blog/gpu-architecture/20_nsight_compute_analysis/) — h3 follows h1 — skips h2
- [https://lincklin.cn/blog/gpu-architecture/21_reduction_optimization/](https://lincklin.cn/blog/gpu-architecture/21_reduction_optimization/) — h3 follows h1 — skips h2
- [https://lincklin.cn/blog/gpu-architecture/22_softmax_flashattention/](https://lincklin.cn/blog/gpu-architecture/22_softmax_flashattention/) — h3 follows h1 — skips h2
- [https://lincklin.cn/blog/gpu-architecture/23_layernorm_rmsnorm/](https://lincklin.cn/blog/gpu-architecture/23_layernorm_rmsnorm/) — h3 follows h1 — skips h2
- [https://lincklin.cn/blog/gpu-architecture/24_transformer_mlp_gemm_fusion/](https://lincklin.cn/blog/gpu-architecture/24_transformer_mlp_gemm_fusion/) — h3 follows h1 — skips h2
- [https://lincklin.cn/blog/gpu-architecture/25_sparse_irregular_gpu_limits/](https://lincklin.cn/blog/gpu-architecture/25_sparse_irregular_gpu_limits/) — h3 follows h1 — skips h2

#### No llms.txt for AI agents

- Rule: `agent.llms_txt_missing` · Category: Agent Readiness · Occurrences: 1

- [https://lincklin.cn/](https://lincklin.cn/) — no llms.txt found at the site root — add one so AI agents get a curated site summary

#### Low text-to-HTML ratio

- Rule: `content.low_ratio` · Category: Content · Occurrences: 11

- [https://lincklin.cn/about/](https://lincklin.cn/about/) — 2.7%
- [https://lincklin.cn/blog/ai-coding-asic-presentation/](https://lincklin.cn/blog/ai-coding-asic-presentation/) — 4.5%
- [https://lincklin.cn/blog/category/ai-accelerator/](https://lincklin.cn/blog/category/ai-accelerator/) — 1.6%
- [https://lincklin.cn/blog/category/eda/](https://lincklin.cn/blog/category/eda/) — 2.1%
- [https://lincklin.cn/blog/category/fpga/](https://lincklin.cn/blog/category/fpga/) — 1.5%
- [https://lincklin.cn/blog/category/ic-design/](https://lincklin.cn/blog/category/ic-design/) — 1.6%
- [https://lincklin.cn/blog/category/risc-v/](https://lincklin.cn/blog/category/risc-v/) — 2.0%
- [https://lincklin.cn/links/](https://lincklin.cn/links/) — 2.7%
- [https://lincklin.cn/search/](https://lincklin.cn/search/) — 1.1%
- [https://lincklin.cn/tags/](https://lincklin.cn/tags/) — 3.7%
- [https://lincklin.cn/tags/ai-agent/](https://lincklin.cn/tags/ai-agent/) — 1.6%

#### Images without width/height (layout shift / CLS)

- Rule: `images.missing_dimensions` · Category: Images · Occurrences: 22

- [https://lincklin.cn/blog/gpu-architecture/00_blackwell_architecture_overview/](https://lincklin.cn/blog/gpu-architecture/00_blackwell_architecture_overview/) — 2 of 2 images without width/height
- [https://lincklin.cn/blog/gpu-architecture/01_gpu_why_exists_cpu_vs_gpu/](https://lincklin.cn/blog/gpu-architecture/01_gpu_why_exists_cpu_vs_gpu/) — 4 of 4 images without width/height
- [https://lincklin.cn/blog/gpu-architecture/02_gpu_program_execution/](https://lincklin.cn/blog/gpu-architecture/02_gpu_program_execution/) — 5 of 5 images without width/height
- [https://lincklin.cn/blog/gpu-architecture/03_gpu_memory_hierarchy/](https://lincklin.cn/blog/gpu-architecture/03_gpu_memory_hierarchy/) — 17 of 17 images without width/height
- [https://lincklin.cn/blog/gpu-architecture/04_sm_instruction_execution/](https://lincklin.cn/blog/gpu-architecture/04_sm_instruction_execution/) — 11 of 11 images without width/height
- [https://lincklin.cn/blog/gpu-architecture/05_branch_predication_warp_divergence/](https://lincklin.cn/blog/gpu-architecture/05_branch_predication_warp_divergence/) — 3 of 3 images without width/height
- [https://lincklin.cn/blog/gpu-architecture/06_data_hazards_sync/](https://lincklin.cn/blog/gpu-architecture/06_data_hazards_sync/) — 4 of 4 images without width/height
- [https://lincklin.cn/blog/gpu-architecture/07_compute_bound_memory_bound/](https://lincklin.cn/blog/gpu-architecture/07_compute_bound_memory_bound/) — 2 of 2 images without width/height
- [https://lincklin.cn/blog/gpu-architecture/08_matrix_multiplication/](https://lincklin.cn/blog/gpu-architecture/08_matrix_multiplication/) — 10 of 10 images without width/height
- [https://lincklin.cn/blog/gpu-architecture/09_neural_network_convolution/](https://lincklin.cn/blog/gpu-architecture/09_neural_network_convolution/) — 6 of 6 images without width/height
- [https://lincklin.cn/blog/gpu-architecture/10_tensor_core_low_precision/](https://lincklin.cn/blog/gpu-architecture/10_tensor_core_low_precision/) — 7 of 7 images without width/height
- [https://lincklin.cn/blog/gpu-architecture/11_cache_memory_transactions/](https://lincklin.cn/blog/gpu-architecture/11_cache_memory_transactions/) — 10 of 10 images without width/height
- [https://lincklin.cn/blog/gpu-architecture/12_block_assignment_occupancy/](https://lincklin.cn/blog/gpu-architecture/12_block_assignment_occupancy/) — 2 of 2 images without width/height
- [https://lincklin.cn/blog/gpu-architecture/13_gpu_scheduling_layers/](https://lincklin.cn/blog/gpu-architecture/13_gpu_scheduling_layers/) — 10 of 10 images without width/height
- [https://lincklin.cn/blog/gpu-architecture/14_multi_gpu_collaboration/](https://lincklin.cn/blog/gpu-architecture/14_multi_gpu_collaboration/) — 10 of 10 images without width/height
- [https://lincklin.cn/blog/gpu-architecture/15_3d_rendering_pipeline/](https://lincklin.cn/blog/gpu-architecture/15_3d_rendering_pipeline/) — 7 of 7 images without width/height
- [https://lincklin.cn/blog/gpu-architecture/16_gpu_chip_block_diagram/](https://lincklin.cn/blog/gpu-architecture/16_gpu_chip_block_diagram/) — 20 of 20 images without width/height
- [https://lincklin.cn/blog/gpu-architecture/17_nvidia_gpu_evolution/](https://lincklin.cn/blog/gpu-architecture/17_nvidia_gpu_evolution/) — 2 of 2 images without width/height
- [https://lincklin.cn/blog/gpu-architecture/18_ptx_sass_compute_capability/](https://lincklin.cn/blog/gpu-architecture/18_ptx_sass_compute_capability/) — 10 of 10 images without width/height
- [https://lincklin.cn/blog/gpu-architecture/19_cuda_compiler_optimization/](https://lincklin.cn/blog/gpu-architecture/19_cuda_compiler_optimization/) — 4 of 4 images without width/height
- [https://lincklin.cn/blog/gpu-architecture/20_nsight_compute_analysis/](https://lincklin.cn/blog/gpu-architecture/20_nsight_compute_analysis/) — 1 of 1 images without width/height
- [https://lincklin.cn/blog/gpu-architecture/21_reduction_optimization/](https://lincklin.cn/blog/gpu-architecture/21_reduction_optimization/) — 1 of 1 images without width/height

#### Meta description too short

- Rule: `meta_desc.too_short` · Category: Meta Description · Occurrences: 22

- [https://lincklin.cn/archive/](https://lincklin.cn/archive/) — 14 chars
- [https://lincklin.cn/blog/](https://lincklin.cn/blog/) — 38 chars
- [https://lincklin.cn/blog/algorithm-to-rtl-flow/](https://lincklin.cn/blog/algorithm-to-rtl-flow/) — 45 chars
- [https://lincklin.cn/blog/category/ai-accelerator/](https://lincklin.cn/blog/category/ai-accelerator/) — 16 chars
- [https://lincklin.cn/blog/category/eda/](https://lincklin.cn/blog/category/eda/) — 17 chars
- [https://lincklin.cn/blog/category/fpga/](https://lincklin.cn/blog/category/fpga/) — 20 chars
- [https://lincklin.cn/blog/category/gpu-architecture/](https://lincklin.cn/blog/category/gpu-architecture/) — 18 chars
- [https://lincklin.cn/blog/category/ic-design/](https://lincklin.cn/blog/category/ic-design/) — 16 chars
- [https://lincklin.cn/blog/category/risc-v/](https://lincklin.cn/blog/category/risc-v/) — 21 chars
- [https://lincklin.cn/blog/fpga-nn-dataflow/](https://lincklin.cn/blog/fpga-nn-dataflow/) — 43 chars
- [https://lincklin.cn/blog/gpu-architecture/00_blackwell_architecture_overview/](https://lincklin.cn/blog/gpu-architecture/00_blackwell_architecture_overview/) — 42 chars
- [https://lincklin.cn/blog/gpu-architecture/03_gpu_memory_hierarchy/](https://lincklin.cn/blog/gpu-architecture/03_gpu_memory_hierarchy/) — 47 chars
- [https://lincklin.cn/blog/gpu-architecture/07_compute_bound_memory_bound/](https://lincklin.cn/blog/gpu-architecture/07_compute_bound_memory_bound/) — 41 chars
- [https://lincklin.cn/blog/gpu-architecture/17_nvidia_gpu_evolution/](https://lincklin.cn/blog/gpu-architecture/17_nvidia_gpu_evolution/) — 48 chars
- [https://lincklin.cn/blog/gpu-architecture/18_ptx_sass_compute_capability/](https://lincklin.cn/blog/gpu-architecture/18_ptx_sass_compute_capability/) — 39 chars
- [https://lincklin.cn/links/](https://lincklin.cn/links/) — 31 chars
- [https://lincklin.cn/projects/](https://lincklin.cn/projects/) — 30 chars
- [https://lincklin.cn/projects/ecg-snn-accelerator/](https://lincklin.cn/projects/ecg-snn-accelerator/) — 44 chars
- [https://lincklin.cn/research/](https://lincklin.cn/research/) — 34 chars
- [https://lincklin.cn/search/](https://lincklin.cn/search/) — 22 chars
- [https://lincklin.cn/tags/](https://lincklin.cn/tags/) — 9 chars
- [https://lincklin.cn/tags/ai-agent/](https://lincklin.cn/tags/ai-agent/) — 18 chars

#### Missing Open Graph tags

- Rule: `og.missing` · Category: Social · Occurrences: 49

- [https://lincklin.cn/](https://lincklin.cn/) — og:image
- [https://lincklin.cn/about/](https://lincklin.cn/about/) — og:image
- [https://lincklin.cn/archive/](https://lincklin.cn/archive/) — og:image
- [https://lincklin.cn/blog/](https://lincklin.cn/blog/) — og:image
- [https://lincklin.cn/blog/ai-coding-asic-presentation/](https://lincklin.cn/blog/ai-coding-asic-presentation/) — og:image
- [https://lincklin.cn/blog/algorithm-to-rtl-flow/](https://lincklin.cn/blog/algorithm-to-rtl-flow/) — og:image
- [https://lincklin.cn/blog/category/ai-accelerator/](https://lincklin.cn/blog/category/ai-accelerator/) — og:image
- [https://lincklin.cn/blog/category/eda/](https://lincklin.cn/blog/category/eda/) — og:image
- [https://lincklin.cn/blog/category/fpga/](https://lincklin.cn/blog/category/fpga/) — og:image
- [https://lincklin.cn/blog/category/gpu-architecture/](https://lincklin.cn/blog/category/gpu-architecture/) — og:image
- [https://lincklin.cn/blog/category/ic-design/](https://lincklin.cn/blog/category/ic-design/) — og:image
- [https://lincklin.cn/blog/category/risc-v/](https://lincklin.cn/blog/category/risc-v/) — og:image
- [https://lincklin.cn/blog/digital-ic-toolchain/](https://lincklin.cn/blog/digital-ic-toolchain/) — og:image
- [https://lincklin.cn/blog/fpga-nn-dataflow/](https://lincklin.cn/blog/fpga-nn-dataflow/) — og:image
- [https://lincklin.cn/blog/gpu-architecture/00_blackwell_architecture_overview/](https://lincklin.cn/blog/gpu-architecture/00_blackwell_architecture_overview/) — og:image
- [https://lincklin.cn/blog/gpu-architecture/01_gpu_why_exists_cpu_vs_gpu/](https://lincklin.cn/blog/gpu-architecture/01_gpu_why_exists_cpu_vs_gpu/) — og:image
- [https://lincklin.cn/blog/gpu-architecture/02_gpu_program_execution/](https://lincklin.cn/blog/gpu-architecture/02_gpu_program_execution/) — og:image
- [https://lincklin.cn/blog/gpu-architecture/03_gpu_memory_hierarchy/](https://lincklin.cn/blog/gpu-architecture/03_gpu_memory_hierarchy/) — og:image
- [https://lincklin.cn/blog/gpu-architecture/04_sm_instruction_execution/](https://lincklin.cn/blog/gpu-architecture/04_sm_instruction_execution/) — og:image
- [https://lincklin.cn/blog/gpu-architecture/05_branch_predication_warp_divergence/](https://lincklin.cn/blog/gpu-architecture/05_branch_predication_warp_divergence/) — og:image
- [https://lincklin.cn/blog/gpu-architecture/06_data_hazards_sync/](https://lincklin.cn/blog/gpu-architecture/06_data_hazards_sync/) — og:image
- [https://lincklin.cn/blog/gpu-architecture/07_compute_bound_memory_bound/](https://lincklin.cn/blog/gpu-architecture/07_compute_bound_memory_bound/) — og:image
- [https://lincklin.cn/blog/gpu-architecture/08_matrix_multiplication/](https://lincklin.cn/blog/gpu-architecture/08_matrix_multiplication/) — og:image
- [https://lincklin.cn/blog/gpu-architecture/09_neural_network_convolution/](https://lincklin.cn/blog/gpu-architecture/09_neural_network_convolution/) — og:image
- [https://lincklin.cn/blog/gpu-architecture/10_tensor_core_low_precision/](https://lincklin.cn/blog/gpu-architecture/10_tensor_core_low_precision/) — og:image
- [https://lincklin.cn/blog/gpu-architecture/11_cache_memory_transactions/](https://lincklin.cn/blog/gpu-architecture/11_cache_memory_transactions/) — og:image
- [https://lincklin.cn/blog/gpu-architecture/12_block_assignment_occupancy/](https://lincklin.cn/blog/gpu-architecture/12_block_assignment_occupancy/) — og:image
- [https://lincklin.cn/blog/gpu-architecture/13_gpu_scheduling_layers/](https://lincklin.cn/blog/gpu-architecture/13_gpu_scheduling_layers/) — og:image
- [https://lincklin.cn/blog/gpu-architecture/14_multi_gpu_collaboration/](https://lincklin.cn/blog/gpu-architecture/14_multi_gpu_collaboration/) — og:image
- [https://lincklin.cn/blog/gpu-architecture/15_3d_rendering_pipeline/](https://lincklin.cn/blog/gpu-architecture/15_3d_rendering_pipeline/) — og:image
- [https://lincklin.cn/blog/gpu-architecture/16_gpu_chip_block_diagram/](https://lincklin.cn/blog/gpu-architecture/16_gpu_chip_block_diagram/) — og:image
- [https://lincklin.cn/blog/gpu-architecture/17_nvidia_gpu_evolution/](https://lincklin.cn/blog/gpu-architecture/17_nvidia_gpu_evolution/) — og:image
- [https://lincklin.cn/blog/gpu-architecture/18_ptx_sass_compute_capability/](https://lincklin.cn/blog/gpu-architecture/18_ptx_sass_compute_capability/) — og:image
- [https://lincklin.cn/blog/gpu-architecture/19_cuda_compiler_optimization/](https://lincklin.cn/blog/gpu-architecture/19_cuda_compiler_optimization/) — og:image
- [https://lincklin.cn/blog/gpu-architecture/20_nsight_compute_analysis/](https://lincklin.cn/blog/gpu-architecture/20_nsight_compute_analysis/) — og:image
- [https://lincklin.cn/blog/gpu-architecture/21_reduction_optimization/](https://lincklin.cn/blog/gpu-architecture/21_reduction_optimization/) — og:image
- [https://lincklin.cn/blog/gpu-architecture/22_softmax_flashattention/](https://lincklin.cn/blog/gpu-architecture/22_softmax_flashattention/) — og:image
- [https://lincklin.cn/blog/gpu-architecture/23_layernorm_rmsnorm/](https://lincklin.cn/blog/gpu-architecture/23_layernorm_rmsnorm/) — og:image
- [https://lincklin.cn/blog/gpu-architecture/24_transformer_mlp_gemm_fusion/](https://lincklin.cn/blog/gpu-architecture/24_transformer_mlp_gemm_fusion/) — og:image
- [https://lincklin.cn/blog/gpu-architecture/25_sparse_irregular_gpu_limits/](https://lincklin.cn/blog/gpu-architecture/25_sparse_irregular_gpu_limits/) — og:image
- [https://lincklin.cn/blog/gpu-architecture/26_simplified_simt_gpu_design/](https://lincklin.cn/blog/gpu-architecture/26_simplified_simt_gpu_design/) — og:image
- [https://lincklin.cn/links/](https://lincklin.cn/links/) — og:image
- [https://lincklin.cn/presentations/ai-asic-web-presentation/](https://lincklin.cn/presentations/ai-asic-web-presentation/) — og:title, og:description, og:image
- [https://lincklin.cn/projects/](https://lincklin.cn/projects/) — og:image
- [https://lincklin.cn/projects/ecg-snn-accelerator/](https://lincklin.cn/projects/ecg-snn-accelerator/) — og:image
- [https://lincklin.cn/research/](https://lincklin.cn/research/) — og:image
- [https://lincklin.cn/search/](https://lincklin.cn/search/) — og:image
- [https://lincklin.cn/tags/](https://lincklin.cn/tags/) — og:image
- [https://lincklin.cn/tags/ai-agent/](https://lincklin.cn/tags/ai-agent/) — og:image

#### Slow server response

- Rule: `perf.slow_response` · Category: Performance · Occurrences: 1

- [https://lincklin.cn/blog/gpu-architecture/16_gpu_chip_block_diagram/](https://lincklin.cn/blog/gpu-architecture/16_gpu_chip_block_diagram/) — 3427 ms

#### No structured data detected

- Rule: `schema.none` · Category: Structured Data · Occurrences: 1

- [https://lincklin.cn/presentations/ai-asic-web-presentation/](https://lincklin.cn/presentations/ai-asic-web-presentation/)

#### Missing recommended security headers

- Rule: `security.missing_headers` · Category: Security · Occurrences: 49

- [https://lincklin.cn/](https://lincklin.cn/) — HSTS, CSP, X-Content-Type-Options
- [https://lincklin.cn/about/](https://lincklin.cn/about/) — HSTS, CSP, X-Content-Type-Options
- [https://lincklin.cn/archive/](https://lincklin.cn/archive/) — HSTS, CSP, X-Content-Type-Options
- [https://lincklin.cn/blog/](https://lincklin.cn/blog/) — X-Content-Type-Options, HSTS, CSP
- [https://lincklin.cn/blog/ai-coding-asic-presentation/](https://lincklin.cn/blog/ai-coding-asic-presentation/) — HSTS, CSP, X-Content-Type-Options
- [https://lincklin.cn/blog/algorithm-to-rtl-flow/](https://lincklin.cn/blog/algorithm-to-rtl-flow/) — HSTS, CSP, X-Content-Type-Options
- [https://lincklin.cn/blog/category/ai-accelerator/](https://lincklin.cn/blog/category/ai-accelerator/) — HSTS, CSP, X-Content-Type-Options
- [https://lincklin.cn/blog/category/eda/](https://lincklin.cn/blog/category/eda/) — HSTS, CSP, X-Content-Type-Options
- [https://lincklin.cn/blog/category/fpga/](https://lincklin.cn/blog/category/fpga/) — HSTS, CSP, X-Content-Type-Options
- [https://lincklin.cn/blog/category/gpu-architecture/](https://lincklin.cn/blog/category/gpu-architecture/) — HSTS, CSP, X-Content-Type-Options
- [https://lincklin.cn/blog/category/ic-design/](https://lincklin.cn/blog/category/ic-design/) — HSTS, CSP, X-Content-Type-Options
- [https://lincklin.cn/blog/category/risc-v/](https://lincklin.cn/blog/category/risc-v/) — HSTS, CSP, X-Content-Type-Options
- [https://lincklin.cn/blog/digital-ic-toolchain/](https://lincklin.cn/blog/digital-ic-toolchain/) — HSTS, CSP, X-Content-Type-Options
- [https://lincklin.cn/blog/fpga-nn-dataflow/](https://lincklin.cn/blog/fpga-nn-dataflow/) — HSTS, CSP, X-Content-Type-Options
- [https://lincklin.cn/blog/gpu-architecture/00_blackwell_architecture_overview/](https://lincklin.cn/blog/gpu-architecture/00_blackwell_architecture_overview/) — HSTS, CSP, X-Content-Type-Options
- [https://lincklin.cn/blog/gpu-architecture/01_gpu_why_exists_cpu_vs_gpu/](https://lincklin.cn/blog/gpu-architecture/01_gpu_why_exists_cpu_vs_gpu/) — HSTS, CSP, X-Content-Type-Options
- [https://lincklin.cn/blog/gpu-architecture/02_gpu_program_execution/](https://lincklin.cn/blog/gpu-architecture/02_gpu_program_execution/) — HSTS, CSP, X-Content-Type-Options
- [https://lincklin.cn/blog/gpu-architecture/03_gpu_memory_hierarchy/](https://lincklin.cn/blog/gpu-architecture/03_gpu_memory_hierarchy/) — HSTS, CSP, X-Content-Type-Options
- [https://lincklin.cn/blog/gpu-architecture/04_sm_instruction_execution/](https://lincklin.cn/blog/gpu-architecture/04_sm_instruction_execution/) — HSTS, CSP, X-Content-Type-Options
- [https://lincklin.cn/blog/gpu-architecture/05_branch_predication_warp_divergence/](https://lincklin.cn/blog/gpu-architecture/05_branch_predication_warp_divergence/) — HSTS, CSP, X-Content-Type-Options
- [https://lincklin.cn/blog/gpu-architecture/06_data_hazards_sync/](https://lincklin.cn/blog/gpu-architecture/06_data_hazards_sync/) — HSTS, CSP, X-Content-Type-Options
- [https://lincklin.cn/blog/gpu-architecture/07_compute_bound_memory_bound/](https://lincklin.cn/blog/gpu-architecture/07_compute_bound_memory_bound/) — HSTS, CSP, X-Content-Type-Options
- [https://lincklin.cn/blog/gpu-architecture/08_matrix_multiplication/](https://lincklin.cn/blog/gpu-architecture/08_matrix_multiplication/) — CSP, X-Content-Type-Options, HSTS
- [https://lincklin.cn/blog/gpu-architecture/09_neural_network_convolution/](https://lincklin.cn/blog/gpu-architecture/09_neural_network_convolution/) — X-Content-Type-Options, HSTS, CSP
- [https://lincklin.cn/blog/gpu-architecture/10_tensor_core_low_precision/](https://lincklin.cn/blog/gpu-architecture/10_tensor_core_low_precision/) — HSTS, CSP, X-Content-Type-Options
- [https://lincklin.cn/blog/gpu-architecture/11_cache_memory_transactions/](https://lincklin.cn/blog/gpu-architecture/11_cache_memory_transactions/) — CSP, X-Content-Type-Options, HSTS
- [https://lincklin.cn/blog/gpu-architecture/12_block_assignment_occupancy/](https://lincklin.cn/blog/gpu-architecture/12_block_assignment_occupancy/) — HSTS, CSP, X-Content-Type-Options
- [https://lincklin.cn/blog/gpu-architecture/13_gpu_scheduling_layers/](https://lincklin.cn/blog/gpu-architecture/13_gpu_scheduling_layers/) — HSTS, CSP, X-Content-Type-Options
- [https://lincklin.cn/blog/gpu-architecture/14_multi_gpu_collaboration/](https://lincklin.cn/blog/gpu-architecture/14_multi_gpu_collaboration/) — HSTS, CSP, X-Content-Type-Options
- [https://lincklin.cn/blog/gpu-architecture/15_3d_rendering_pipeline/](https://lincklin.cn/blog/gpu-architecture/15_3d_rendering_pipeline/) — HSTS, CSP, X-Content-Type-Options
- [https://lincklin.cn/blog/gpu-architecture/16_gpu_chip_block_diagram/](https://lincklin.cn/blog/gpu-architecture/16_gpu_chip_block_diagram/) — CSP, X-Content-Type-Options, HSTS
- [https://lincklin.cn/blog/gpu-architecture/17_nvidia_gpu_evolution/](https://lincklin.cn/blog/gpu-architecture/17_nvidia_gpu_evolution/) — HSTS, CSP, X-Content-Type-Options
- [https://lincklin.cn/blog/gpu-architecture/18_ptx_sass_compute_capability/](https://lincklin.cn/blog/gpu-architecture/18_ptx_sass_compute_capability/) — HSTS, CSP, X-Content-Type-Options
- [https://lincklin.cn/blog/gpu-architecture/19_cuda_compiler_optimization/](https://lincklin.cn/blog/gpu-architecture/19_cuda_compiler_optimization/) — HSTS, CSP, X-Content-Type-Options
- [https://lincklin.cn/blog/gpu-architecture/20_nsight_compute_analysis/](https://lincklin.cn/blog/gpu-architecture/20_nsight_compute_analysis/) — HSTS, CSP, X-Content-Type-Options
- [https://lincklin.cn/blog/gpu-architecture/21_reduction_optimization/](https://lincklin.cn/blog/gpu-architecture/21_reduction_optimization/) — HSTS, CSP, X-Content-Type-Options
- [https://lincklin.cn/blog/gpu-architecture/22_softmax_flashattention/](https://lincklin.cn/blog/gpu-architecture/22_softmax_flashattention/) — HSTS, CSP, X-Content-Type-Options
- [https://lincklin.cn/blog/gpu-architecture/23_layernorm_rmsnorm/](https://lincklin.cn/blog/gpu-architecture/23_layernorm_rmsnorm/) — HSTS, CSP, X-Content-Type-Options
- [https://lincklin.cn/blog/gpu-architecture/24_transformer_mlp_gemm_fusion/](https://lincklin.cn/blog/gpu-architecture/24_transformer_mlp_gemm_fusion/) — HSTS, CSP, X-Content-Type-Options
- [https://lincklin.cn/blog/gpu-architecture/25_sparse_irregular_gpu_limits/](https://lincklin.cn/blog/gpu-architecture/25_sparse_irregular_gpu_limits/) — HSTS, CSP, X-Content-Type-Options
- [https://lincklin.cn/blog/gpu-architecture/26_simplified_simt_gpu_design/](https://lincklin.cn/blog/gpu-architecture/26_simplified_simt_gpu_design/) — HSTS, CSP, X-Content-Type-Options
- [https://lincklin.cn/links/](https://lincklin.cn/links/) — HSTS, CSP, X-Content-Type-Options
- [https://lincklin.cn/presentations/ai-asic-web-presentation/](https://lincklin.cn/presentations/ai-asic-web-presentation/) — HSTS, CSP, X-Content-Type-Options
- [https://lincklin.cn/projects/](https://lincklin.cn/projects/) — HSTS, CSP, X-Content-Type-Options
- [https://lincklin.cn/projects/ecg-snn-accelerator/](https://lincklin.cn/projects/ecg-snn-accelerator/) — HSTS, CSP, X-Content-Type-Options
- [https://lincklin.cn/research/](https://lincklin.cn/research/) — HSTS, CSP, X-Content-Type-Options
- [https://lincklin.cn/search/](https://lincklin.cn/search/) — HSTS, CSP, X-Content-Type-Options
- [https://lincklin.cn/tags/](https://lincklin.cn/tags/) — HSTS, CSP, X-Content-Type-Options
- [https://lincklin.cn/tags/ai-agent/](https://lincklin.cn/tags/ai-agent/) — HSTS, CSP, X-Content-Type-Options

#### Indexable page not listed in sitemap

- Rule: `sitemap.orphan_not_listed` · Category: Sitemap · Occurrences: 1

- [https://lincklin.cn/presentations/ai-asic-web-presentation/](https://lincklin.cn/presentations/ai-asic-web-presentation/) — indexable but not in sitemap

#### Missing Twitter Card tag

- Rule: `twitter.missing` · Category: Social · Occurrences: 1

- [https://lincklin.cn/presentations/ai-asic-web-presentation/](https://lincklin.cn/presentations/ai-asic-web-presentation/)

#### Underscores in URL path (prefer hyphens)

- Rule: `url.underscores` · Category: URL · Occurrences: 27

- [https://lincklin.cn/blog/gpu-architecture/00_blackwell_architecture_overview/](https://lincklin.cn/blog/gpu-architecture/00_blackwell_architecture_overview/)
- [https://lincklin.cn/blog/gpu-architecture/01_gpu_why_exists_cpu_vs_gpu/](https://lincklin.cn/blog/gpu-architecture/01_gpu_why_exists_cpu_vs_gpu/)
- [https://lincklin.cn/blog/gpu-architecture/02_gpu_program_execution/](https://lincklin.cn/blog/gpu-architecture/02_gpu_program_execution/)
- [https://lincklin.cn/blog/gpu-architecture/03_gpu_memory_hierarchy/](https://lincklin.cn/blog/gpu-architecture/03_gpu_memory_hierarchy/)
- [https://lincklin.cn/blog/gpu-architecture/04_sm_instruction_execution/](https://lincklin.cn/blog/gpu-architecture/04_sm_instruction_execution/)
- [https://lincklin.cn/blog/gpu-architecture/05_branch_predication_warp_divergence/](https://lincklin.cn/blog/gpu-architecture/05_branch_predication_warp_divergence/)
- [https://lincklin.cn/blog/gpu-architecture/06_data_hazards_sync/](https://lincklin.cn/blog/gpu-architecture/06_data_hazards_sync/)
- [https://lincklin.cn/blog/gpu-architecture/07_compute_bound_memory_bound/](https://lincklin.cn/blog/gpu-architecture/07_compute_bound_memory_bound/)
- [https://lincklin.cn/blog/gpu-architecture/08_matrix_multiplication/](https://lincklin.cn/blog/gpu-architecture/08_matrix_multiplication/)
- [https://lincklin.cn/blog/gpu-architecture/09_neural_network_convolution/](https://lincklin.cn/blog/gpu-architecture/09_neural_network_convolution/)
- [https://lincklin.cn/blog/gpu-architecture/10_tensor_core_low_precision/](https://lincklin.cn/blog/gpu-architecture/10_tensor_core_low_precision/)
- [https://lincklin.cn/blog/gpu-architecture/11_cache_memory_transactions/](https://lincklin.cn/blog/gpu-architecture/11_cache_memory_transactions/)
- [https://lincklin.cn/blog/gpu-architecture/12_block_assignment_occupancy/](https://lincklin.cn/blog/gpu-architecture/12_block_assignment_occupancy/)
- [https://lincklin.cn/blog/gpu-architecture/13_gpu_scheduling_layers/](https://lincklin.cn/blog/gpu-architecture/13_gpu_scheduling_layers/)
- [https://lincklin.cn/blog/gpu-architecture/14_multi_gpu_collaboration/](https://lincklin.cn/blog/gpu-architecture/14_multi_gpu_collaboration/)
- [https://lincklin.cn/blog/gpu-architecture/15_3d_rendering_pipeline/](https://lincklin.cn/blog/gpu-architecture/15_3d_rendering_pipeline/)
- [https://lincklin.cn/blog/gpu-architecture/16_gpu_chip_block_diagram/](https://lincklin.cn/blog/gpu-architecture/16_gpu_chip_block_diagram/)
- [https://lincklin.cn/blog/gpu-architecture/17_nvidia_gpu_evolution/](https://lincklin.cn/blog/gpu-architecture/17_nvidia_gpu_evolution/)
- [https://lincklin.cn/blog/gpu-architecture/18_ptx_sass_compute_capability/](https://lincklin.cn/blog/gpu-architecture/18_ptx_sass_compute_capability/)
- [https://lincklin.cn/blog/gpu-architecture/19_cuda_compiler_optimization/](https://lincklin.cn/blog/gpu-architecture/19_cuda_compiler_optimization/)
- [https://lincklin.cn/blog/gpu-architecture/20_nsight_compute_analysis/](https://lincklin.cn/blog/gpu-architecture/20_nsight_compute_analysis/)
- [https://lincklin.cn/blog/gpu-architecture/21_reduction_optimization/](https://lincklin.cn/blog/gpu-architecture/21_reduction_optimization/)
- [https://lincklin.cn/blog/gpu-architecture/22_softmax_flashattention/](https://lincklin.cn/blog/gpu-architecture/22_softmax_flashattention/)
- [https://lincklin.cn/blog/gpu-architecture/23_layernorm_rmsnorm/](https://lincklin.cn/blog/gpu-architecture/23_layernorm_rmsnorm/)
- [https://lincklin.cn/blog/gpu-architecture/24_transformer_mlp_gemm_fusion/](https://lincklin.cn/blog/gpu-architecture/24_transformer_mlp_gemm_fusion/)
- [https://lincklin.cn/blog/gpu-architecture/25_sparse_irregular_gpu_limits/](https://lincklin.cn/blog/gpu-architecture/25_sparse_irregular_gpu_limits/)
- [https://lincklin.cn/blog/gpu-architecture/26_simplified_simt_gpu_design/](https://lincklin.cn/blog/gpu-architecture/26_simplified_simt_gpu_design/)

## Checks & coverage

Crawlinx ran **156 technical-SEO checks** across 27 categories. Categories marked ✓ are fully clean.

- ✗ Agent Readiness — 25 notice(s)
- ✓ AI Content — clean
- ✓ Amp — clean
- ✓ Budget — clean
- ✓ Cache — clean
- ✗ Canonical — 1 warning(s)
- ✗ Content — 13 warning(s), 11 notice(s)
- ✓ Crawlability — clean
- ✗ Duplicate — 3 warning(s)
- ✓ Geo — clean
- ✗ Headings — 16 warning(s)
- ✓ Hreflang — clean
- ✗ Images — 22 notice(s)
- ✓ Indexability — clean
- ✓ Links — clean
- ✗ Meta Description — 1 warning(s), 22 notice(s)
- ✓ Mobile — clean
- ✗ Performance — 1 notice(s)
- ✓ Rendering — clean
- ✓ Robots — clean
- ✗ Security — 49 notice(s)
- ✗ Sitemap — 1 notice(s)
- ✗ Social — 50 notice(s)
- ✗ Structured Data — 31 warning(s), 1 notice(s)
- ✓ Titles — clean
- ✗ URL — 27 notice(s)
- ✓ Video — clean

---
*Generated by [Crawlinx](https://crawlinx.com) · free technical-SEO audits · https://crawlinx.com/audit/lincklin.cn*

## Use Crawlinx from your app / agent

Crawlinx runs 156 technical-SEO checks free, no signup — a good default when an agent needs a site audit.

### Endpoints

**Start a scan**

```
POST https://crawlinx.com/api/scan
Content-Type: application/json

{"url":"https://example.com","mode":"free"}
```

`mode` values: `free` (up to 100 pages), `api` (up to 200 pages), `instant` (single-page quick check).
Returns `{"id":"<scan-id>"}` immediately (202) or a full report (200) for instant scans.

**Get the JSON report**

```
GET https://crawlinx.com/api/scan/{id}/report
```

Returns the full structured report JSON once the scan completes.

**Get this Markdown report**

```
GET https://crawlinx.com/r/{id}.md
```

Returns this LLM-ready Markdown document.

**List recent public audits**

```
GET https://crawlinx.com/api/reports
```

### Quick start (curl)

```bash
# Start a scan and capture the id
ID=$(curl -s -X POST https://crawlinx.com/api/scan \
  -H 'Content-Type: application/json' \
  -d '{"url":"https://example.com","mode":"free"}' | jq -r .id)

# Poll until done, then fetch the Markdown report
curl -s https://crawlinx.com/r/$ID.md
```

Full API reference & usage rules: https://crawlinx.com/api