{"_canonicalization":{"envelope_id":"axm_ + sha256(envelope minus {signature, axiom_id, anchors})","envelope_signature":"ed25519(envelope minus {signature, axiom_id})","json":"sort_keys=True, separators=(',',':'), ensure_ascii=False, allow_nan=False, utf-8","leaf_hash":"sha256(0x00 || canonical_json(envelope_full))","seal_signature":"ed25519(seal minus {signature, sig_algorithm})"},"axiom_id":"axm_35e1146475cbee51c4c01169da6d7114b0451e5df070d768aafd4caf0bcabc1b","bitcoin_anchor":{"bitcoin_attestations":[],"calendar_attestations":[],"ots_url":"","stamped_at":"","status":"pending_next_stamp"},"envelope":{"anchors":[{"chain":"crovia.axiom_graph","height":0,"merkle_proof":"spider_vendor_press_v1","root_at_anchor":"spider_vendor_press_v1"}],"axiom_id":"axm_35e1146475cbee51c4c01169da6d7114b0451e5df070d768aafd4caf0bcabc1b","axiom_type":"AX.OBS","body":{"axiom_subtype":"news.vendor_press.v1","category":"news","fingerprint":"d49764038b46e00bdb77c53415d97076ff6b36aa16fe7fbe6894a18f5b2ea4a3","published":"Tue, 30 Jun 2026 00:00:00 -0400","receipt_hash":"d49764038b46e00bdb77c53415d97076ff6b36aa16fe7fbe6894a18f5b2ea4a3","schema":"spider.news.vendor_press.v1","spider":"vendor_press","spider_record":{"axiom_subtype":"news.vendor_press.v1","category":"news","decision_hint":"POSITIVE","envelope_target":"AX.OBS","fingerprint":"d49764038b46e00bdb77c53415d97076ff6b36aa16fe7fbe6894a18f5b2ea4a3","observed_at":"2026-06-30T04:43:04.087680Z","parent_run_hash":"74f7ab392cc702044101fe24a76a2fdad11164cd79ce725aad6c446a477e89c5","published":"Tue, 30 Jun 2026 00:00:00 -0400","runtime_version":"0.1.0","schema":"spider.news.vendor_press.v1","source_status":200,"source_url":"https://export.arxiv.org/rss/cs.AI","spider":"vendor_press","summary_excerpt":"arXiv:2606.08761v2 Announce Type: replace-cross \nAbstract: W4A4 quantization promises full utilization of INT4 Tensor Cores, yet group dequantization overhead on CUDA Cores has driven existing systems to mixed-precision fallbacks. We present the first systematic study of how intra-SM compute balance governs this bottleneck. Through controlled benchmarks across four GPUs from Ampere and Ada architectures, we identify the Tensor Cores to CUDA Cores throughput ratio ($\\rho$) as the primary hardware indicator: the W4A4-g128 kernel yields $2.0$--$2.5\\times$ speedup on RTX~3090 ($\\rho=16$) yet degrades to $0.43$--$0.47\\times$ on A100 ($\\rho=64$) in compute-bond scenarios, establishing W4A4 viability as platform-dependent rather than universally infeasible. Guided by this finding, we build \\textbf{APEX4}, which co-designs pure INT4 GEMM kernels with $\\rho$-aware granularity adaptation to mitigate the CUDA Cores dequantization bottleneck. APEX4 achieves perplexity within 0.63 of FP16 on LLaMA-","title":"APEX4: Efficient Pure W4A4 LLM Inference via Intra-SM Compute Rebalancing","url":"https://arxiv.org/abs/2606.08761","vendor":"arxiv_cs_ai"},"summary":"arXiv:2606.08761v2 Announce Type: replace-cross \nAbstract: W4A4 quantization promises full utilization of INT4 Tensor Cores, yet group dequantization overhead on CUDA Cores has driven existing systems to mixed-precision fallbacks. We present the first systematic study of how intra-SM compute balance governs this bottleneck. Through controlled benchmarks across four GPUs from Ampere and Ada architectures, we identify the Tensor Cores to CUDA Cores throughput ratio ($\\rho$) as the primary hardware indicator: the W4A4-g128 kernel yields $2.0$--$2.5\\times$ speedup on RTX~3090 ($\\rho=16$) yet degrades to $0.43$--$0.47\\times$ on A100 ($\\rho=64$) in compute-bond scenarios, establishing W4A4 viability as platform-dependent rather than universally infeasible. Guided by this finding, we build \\textbf{APEX4}, which co-designs pure INT4 GEMM kernels with $\\rho$-aware granularity adaptation to mitigate the CUDA Cores dequantization bottleneck. APEX4 achieves perplexity within 0.63 of FP16 on LLaMA-","title":"APEX4: Efficient Pure W4A4 LLM Inference via Intra-SM Compute Rebalancing","vendor":"arxiv_cs_ai"},"confidence":{"method":"deterministic"},"decision":"POSITIVE","issued_at":"2026-06-30T04:43:04Z","notes":"Spider vendor_press (news) news.vendor_press.v1","object":{"captured_by":"crovia.spider.vendor_press","primary_source_url":"https://arxiv.org/abs/2606.08761"},"predecessors":[],"schema":"crovia.axiom.v1","signature":"ed25519:6e1b929197acd65b75d5413fce700c77e1ecf4b8a0d40928e26418ed1f081b744afc37d54edf5193c2be5144678689506c64b7347d321252f85e215422b56b0c","signer":"crovia.substrate","subject":{"observed_at":"2026-06-30T04:43:04Z","source_collector":"spider:vendor_press","target_id":"https://arxiv.org/abs/2606.08761"},"tsa":{"authority":"crovia.substrate.bootstrap","rfc3161_token":"{\"kind\":\"crovia.bootstrap.tsa\",\"source_jsonl\":\"/opt/crovia/spider/data/news/vendor_press_v1.jsonl\",\"source_seal_merkle_root\":\"spider_vendor_press_v1\",\"upgrade_path\":\"Sessione H \\u2014 OpenTimestamps weekly anchor\"}"},"zk_mode":"clear","zk_proof":null},"ledger":{"leaf_hash":"2bc83b6b71e17a695b6d7daf4e447de411a27085425edfe38f0d31a1a6df3293","leaf_index":265227,"ledger_path":"/opt/crovia/substrate/axiom_ledger.jsonl"},"merkle_proof":{"hash_alg":"sha256","leaf_prefix":"0x00","node_prefix":"0x01","odd_leaf_rule":"duplicate_last","path":[{"sibling":"85955cb40684b41b100e6dad56eb071e2f51f20e6c1c5dc0b27fec351428e9f5","side":"left"},{"sibling":"143df79eb0b0757fff375a9b83f335cc0e546dd989216623fe8c366f6358c23e","side":"left"},{"sibling":"c3347878d017decf1680868596d00af0944d234e65f51985deef11014654b2ab","side":"right"},{"sibling":"ee84fab4995bbf16ee814433a62c3155e4c92a69f6891e9e53aeea7aebba0cda","side":"left"},{"sibling":"750af34a5e4fd1f98e08b05ae7a486ede017c6a168c7636c84ba8054d5ded484","side":"right"},{"sibling":"a495e190d223e69b24e0d5ad23a15e735fff86f2c73b282efad1181b007e3623","side":"right"},{"sibling":"ff273c8201c286328de95b4c3f7ad04c9fb005c67fda3de45c54b72f9f5db5b3","side":"right"},{"sibling":"3cde187de7516743e6cd60012ce336384e260530220514cdd49599ef710e14e5","side":"right"},{"sibling":"2da116eff026ee371a88b8eb1a0aeb87679b487f1fbf1623083121809f754234","side":"right"},{"sibling":"eef13ac8a36eb836ad77846f6ef08d9afa4dcc380a64b88c542b57231522db0a","side":"right"},{"sibling":"a2cf20d97c095bc4fa7c4f5bb4f2266dd953e3d733d7e185aa164763491b312c","side":"left"},{"sibling":"f9b4bed84fa6990c71ad2887c91bda183001f05f6b648f21d1273045a26b11fd","side":"left"},{"sibling":"173d2dc4b29ee04ea41d6d0ebc334c4bc2d46e7ee4230c94765413f24fb4bc42","side":"right"},{"sibling":"112461f7c0ec411116fb5c6c90fe95cea9d8f188b9fe08afe25a837ac02d0071","side":"right"},{"sibling":"ea9488204352c49db8f7daf05eefcd7628ecf9413830346674801a99d0654a94","side":"right"},{"sibling":"6261c13b9922cb657f10d1e5d36ec15d8771cf8766e36c61dcbffb7bed57e396","side":"right"},{"sibling":"fa19aa3faf287618b820bcfceebb366152ad521dd20ef9f51e977816663e448b","side":"right"},{"sibling":"c32f943406b62d1fc59b7f7e243492174c8e1caba8c8a2705f86c773315736e0","side":"right"},{"sibling":"1cecb7f447febd025aac272837c80de218aecc6485d2395a509b2a1f1b9c746e","side":"left"}]},"schema":"crovia.axiom_proof.v1","seal":{"first_collector_run_id":"","first_receipt_hash":"","jsonl_path":"/opt/crovia/substrate/axiom_ledger.jsonl","key_id":"430895f101d38164","last_collector_run_id":"","last_receipt_hash":"","leaf_count":265374,"merkle_root":"9636001ecab173cb6af10dc7c71eb14585daa62f9c0a6f027046f05633156891","public_key_hex":"cf742e26f75669dc673cb5c0786a1ae23ae8ca19c347317192ce40c28a7ff25c","run_id":"hourly_json_retrofit_20260630T053701Z","schema":"crovia.seal.v1","seal_family_version":"crovia-seal-family/1","seal_kind":"substrate_batch","sealed_at":"2026-06-30T05:38:03Z","sig_algorithm":"ed25519","signature":"6d4b8fd9b9da856cbb5fba7540877c6a63fa18a5ec3eaf28dc4d6d1c64921c9c0565f8c95f4b7aec0e7744fd7754844f051baf863db708cad768765b416a7b0c","signer_version":"1.1.0"},"trust_root":{"key_id":"430895f101d38164","public_key_hex":"cf742e26f75669dc673cb5c0786a1ae23ae8ca19c347317192ce40c28a7ff25c","signature_algorithm":"ed25519","url":"/registry/canon/TRUST_ROOT.md"},"verifier":{"spec":"/registry/canon/AXIOM_RECEIPT_v1.md","url":"/v/axm_35e1146475cbee51c4c01169da6d7114b0451e5df070d768aafd4caf0bcabc1b"}}