{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:FLKY2LKR7DV5SXK2EORYRAIBJA","short_pith_number":"pith:FLKY2LKR","schema_version":"1.0","canonical_sha256":"2ad58d2d51f8ebd95d5a23a38881014834a7054b19e68851e8ef12417ae383ea","source":{"kind":"arxiv","id":"2409.11424","version":1},"attestation_state":"computed","paper":{"title":"LlamaF: An Efficient Llama2 Architecture Accelerator on Embedded FPGAs","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.AR","authors_text":"Han Xu, Shihao Ji, Yutong Li","submitted_at":"2024-09-12T17:53:37Z","abstract_excerpt":"Large language models (LLMs) have demonstrated remarkable abilities in natural language processing. However, their deployment on resource-constrained embedded devices remains difficult due to memory and computational demands. In this paper, we present an FPGA-based accelerator designed to improve LLM inference performance on embedded FPGAs. We employ post-training quantization to reduce model size and optimize for off-chip memory bandwidth. Our design features asynchronous computation and a fully pipelined accelerator for matrix-vector multiplication. Experiments of the TinyLlama 1.1B model on"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2409.11424","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AR","submitted_at":"2024-09-12T17:53:37Z","cross_cats_sorted":[],"title_canon_sha256":"7b94b99b70933aca34275671fdf19d41d7d024bf06b5eedfe382f3154422bf0a","abstract_canon_sha256":"bc27b6142f9504646d8a39ef23c998a8854a0438ef1a6a7a871d5000ebbd4098"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:08:30.648598Z","signature_b64":"Esu9/FL3iX5f/570YQ+PSVb/MmxinrOMqC9Ej5XVOFBlU2Fe0sFlu+vnfokZNuz6dNY3XjJSsgtkqjdZLW1qAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"2ad58d2d51f8ebd95d5a23a38881014834a7054b19e68851e8ef12417ae383ea","last_reissued_at":"2026-07-05T09:08:30.648075Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:08:30.648075Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"LlamaF: An Efficient Llama2 Architecture Accelerator on Embedded FPGAs","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.AR","authors_text":"Han Xu, Shihao Ji, Yutong Li","submitted_at":"2024-09-12T17:53:37Z","abstract_excerpt":"Large language models (LLMs) have demonstrated remarkable abilities in natural language processing. However, their deployment on resource-constrained embedded devices remains difficult due to memory and computational demands. In this paper, we present an FPGA-based accelerator designed to improve LLM inference performance on embedded FPGAs. We employ post-training quantization to reduce model size and optimize for off-chip memory bandwidth. Our design features asynchronous computation and a fully pipelined accelerator for matrix-vector multiplication. Experiments of the TinyLlama 1.1B model on"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2409.11424","kind":"arxiv","version":1},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2409.11424/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"aliases":[{"alias_kind":"arxiv","alias_value":"2409.11424","created_at":"2026-07-05T09:08:30.648141+00:00"},{"alias_kind":"arxiv_version","alias_value":"2409.11424v1","created_at":"2026-07-05T09:08:30.648141+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2409.11424","created_at":"2026-07-05T09:08:30.648141+00:00"},{"alias_kind":"pith_short_12","alias_value":"FLKY2LKR7DV5","created_at":"2026-07-05T09:08:30.648141+00:00"},{"alias_kind":"pith_short_16","alias_value":"FLKY2LKR7DV5SXK2","created_at":"2026-07-05T09:08:30.648141+00:00"},{"alias_kind":"pith_short_8","alias_value":"FLKY2LKR","created_at":"2026-07-05T09:08:30.648141+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2412.19925","citing_title":"HADES: Hardware Accelerated Decoding for Efficient Speculation in Large Language Models","ref_index":22,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/FLKY2LKR7DV5SXK2EORYRAIBJA","json":"https://pith.science/pith/FLKY2LKR7DV5SXK2EORYRAIBJA.json","graph_json":"https://pith.science/api/pith-number/FLKY2LKR7DV5SXK2EORYRAIBJA/graph.json","events_json":"https://pith.science/api/pith-number/FLKY2LKR7DV5SXK2EORYRAIBJA/events.json","paper":"https://pith.science/paper/FLKY2LKR"},"agent_actions":{"view_html":"https://pith.science/pith/FLKY2LKR7DV5SXK2EORYRAIBJA","download_json":"https://pith.science/pith/FLKY2LKR7DV5SXK2EORYRAIBJA.json","view_paper":"https://pith.science/paper/FLKY2LKR","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2409.11424&json=true","fetch_graph":"https://pith.science/api/pith-number/FLKY2LKR7DV5SXK2EORYRAIBJA/graph.json","fetch_events":"https://pith.science/api/pith-number/FLKY2LKR7DV5SXK2EORYRAIBJA/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/FLKY2LKR7DV5SXK2EORYRAIBJA/action/timestamp_anchor","attest_storage":"https://pith.science/pith/FLKY2LKR7DV5SXK2EORYRAIBJA/action/storage_attestation","attest_author":"https://pith.science/pith/FLKY2LKR7DV5SXK2EORYRAIBJA/action/author_attestation","sign_citation":"https://pith.science/pith/FLKY2LKR7DV5SXK2EORYRAIBJA/action/citation_signature","submit_replication":"https://pith.science/pith/FLKY2LKR7DV5SXK2EORYRAIBJA/action/replication_record"}},"created_at":"2026-07-05T09:08:30.648141+00:00","updated_at":"2026-07-05T09:08:30.648141+00:00"}