{"id":83104,"topic":"ai","source":"Sohu","title":"能跑30B大模型！天玑9600系列发布：AI性能大幅升级 - Sohu","url":"https://m.sohu.com/a/1076445217_122014422?scm=10001.325_13-325_13.0.0-0-0-0-0.5_1334","url_hash":"0682249830e30220cc44199ed9c6cead754bfbf6","author":"","summary":"<a href=\"https://news.google.com/rss/articles/CBMijAFBVV95cUxOSVNlaThaU2pxVExDeGNFaDJLOG92Yi1OVTNtV3VhVHlsaEs4X0JsOTRlSUhoTFIxTGdibEJ5UUdFVWRnUWVDWWZxdkxzRVlBdmZEU3VickI5NkVRSlZmSXA2cU5uLW12bVlVdEtfU3JZcDg4VWNJbFJXYzBXVEJEbVNycUlnZWh0VDJHcA?oc=5\" target=\"_blank\">能跑30B大模型！天玑9600系列发布：AI性能大幅升级</a>&nbsp;&nbsp;<font color=\"#6f6f6f\">Sohu</font>","content":"能跑30B大模型！天玑9600系列发布：AI性能大幅升级\n（来源：雷科技）\n9月15日，联发科发布天玑9600系列旗舰芯片，包含天玑9600 Pro和天玑9600M。这是联发科第一款台积电2nm手机芯片。同一天，vivo官宣X500 Pro系列全球首发这颗芯片，OPPO Find X10 Pro Max也在首批搭载名单中。\n参数层面，天玑9600 Pro用了2+3+3的全大核CPU架构，2颗4.55GHz的C2-Ultra超大核，外加6颗C2-Pro大核（3颗4.35GHz、3颗3.1GHz），缓存总容量34.5MB，率先支持LPDDR6内存和UFS 5.0闪存。官方给出的数字是单核性能提升17%、多核提升15%，多核在峰值性能下的功耗比上一代省了61%。\nAI性能是它的一大亮点，天玑9600 Pro把CPU和NPU绑在一起调度，官方称为「智算融合架构」，第二代超能效NPU功耗省40%，超性能NPU 1090的大模型预填充性能提升51%，每瓦生成词元数最高提升55%，可在端侧运行最高30B的MoE模型。","image_url":"//q5.itc.cn/q_70/images03/20260915/a2bf1ab3b1f843ce87d833acf66a59c2.jpeg","lang":"zh","published_at":"2026-09-15T11:35:26+00:00","fetched_at":"2026-09-15T18:15:08+00:00","status":"read","starred":0,"extract_state":"ok","summary_auto":"能跑30B大模型！天玑9600系列发布：AI性能大幅升级\n（来源：雷科技）\n9月15日，联发科发布天玑9600系列旗舰芯片，包含天玑9600 Pro和天玑9600M。这是联发科第一款台积电2nm手机芯片。同一天，vivo官宣X500 Pro系列全球首发这颗芯片，OPPO Find X10 Pro Max也在首批搭载名单中。\n参数层面，天玑9600 Pro用了2+3+3的全大核CPU架构，2颗4.55GHz的C2-Ultra超大核，外加6颗C2-Pro大核（3颗4.35GHz、3颗3.1GHz），缓存总容量34.5MB，率先支持LPDDR6内存和UFS 5.0闪存。官方给出的数字是单核性能提升17%、多核提升15%，多核在峰值性能下的功耗比上一代省了61%。\nAI性能是它的一大亮点，天玑9600 Pro把CPU和NPU绑在一起调度，官方称为「智算融合架构」，第二代超能效NPU功耗省40%，超性能NPU 1090的大模型预填充性能提升51%，每瓦生成词元数最高提升55%，可在端侧运行最高30B的MoE模型。","cluster_id":null,"extract_retries":0,"extract_error":null,"contract_version":"news_item.v1","format_contract_version":"news_item_formats.v1","dedup_url":"https://m.sohu.com/a/1076445217_122014422?scm=10001.325_13-325_13.0.0-0-0-0-0.5_1334","quality_profile":{"profile_version":"extraction_quality.v2","bucket":"high","confidence":0.9,"failure_kind":"none","retryable":false,"retry_after_attempts":0,"reason":"High confidence: full text extraction produced 458 characters.","operator_guidance":{"severity":"ok","recommended_action":"trust_full_text","next_step":"Use the extracted full text as the primary article source.","operator_label":"Ready","can_retry":false,"can_use_summary":false,"diagnostics_required":false},"content_depth":{"contract_version":"content_depth.v1","category":"full_text","label":"Full text","has_full_text":true,"has_summary":true,"content_length":458,"summary_length":458,"usable_text_length":458,"source_field":"content"},"legacy_collapsed":false,"signals":{"extract_state":"ok","extract_error":null,"extract_retries":0,"content_length":458,"summary_length":458}},"news_item":{"id":83104,"canonical_url":"https://m.sohu.com/a/1076445217_122014422?scm=10001.325_13-325_13.0.0-0-0-0-0.5_1334","source_url":"https://m.sohu.com/a/1076445217_122014422?scm=10001.325_13-325_13.0.0-0-0-0-0.5_1334","title":"能跑30B大模型！天玑9600系列发布：AI性能大幅升级 - Sohu","source_name":"Sohu","author":null,"published_at":"2026-09-15T11:35:26+00:00","locale":"zh","topic":"ai","tags":[],"rss_summary":"<a href=\"https://news.google.com/rss/articles/CBMijAFBVV95cUxOSVNlaThaU2pxVExDeGNFaDJLOG92Yi1OVTNtV3VhVHlsaEs4X0JsOTRlSUhoTFIxTGdibEJ5UUdFVWRnUWVDWWZxdkxzRVlBdmZEU3VickI5NkVRSlZmSXA2cU5uLW12bVlVdEtfU3JZcDg4VWNJbFJXYzBXVEJEbVNycUlnZWh0VDJHcA?oc=5\" target=\"_blank\">能跑30B大模型！天玑9600系列发布：AI性能大幅升级</a>&nbsp;&nbsp;<font color=\"#6f6f6f\">Sohu</font>","full_text":"能跑30B大模型！天玑9600系列发布：AI性能大幅升级\n（来源：雷科技）\n9月15日，联发科发布天玑9600系列旗舰芯片，包含天玑9600 Pro和天玑9600M。这是联发科第一款台积电2nm手机芯片。同一天，vivo官宣X500 Pro系列全球首发这颗芯片，OPPO Find X10 Pro Max也在首批搭载名单中。\n参数层面，天玑9600 Pro用了2+3+3的全大核CPU架构，2颗4.55GHz的C2-Ultra超大核，外加6颗C2-Pro大核（3颗4.35GHz、3颗3.1GHz），缓存总容量34.5MB，率先支持LPDDR6内存和UFS 5.0闪存。官方给出的数字是单核性能提升17%、多核提升15%，多核在峰值性能下的功耗比上一代省了61%。\nAI性能是它的一大亮点，天玑9600 Pro把CPU和NPU绑在一起调度，官方称为「智算融合架构」，第二代超能效NPU功耗省40%，超性能NPU 1090的大模型预填充性能提升51%，每瓦生成词元数最高提升55%，可在端侧运行最高30B的MoE模型。","excerpt":"能跑30B大模型！天玑9600系列发布：AI性能大幅升级\n（来源：雷科技）\n9月15日，联发科发布天玑9600系列旗舰芯片，包含天玑9600 Pro和天玑9600M。这是联发科第一款台积电2nm手机芯片。同一天，vivo官宣X500 Pro系列全球首发这颗芯片，OPPO Find X10 Pro Max也在首批搭载名单中。\n参数层面，天玑9600 Pro用了2+3+3的全大核CPU架构，2颗4.55GHz的C2-Ultra超大核，外加6颗C2-Pro大核（3颗4.35GHz、3颗3.1GHz），缓存总容量34.5MB，率先支持LPDDR6内存和UFS 5.0闪存。官方给出的数字是单核性能提升17%、多核提升15%，多核在峰值性能下的功耗比上一代省了61%。\nAI性能是它的一大亮点，天玑9600 Pro把CPU和NPU绑在一起调度，官方称为「智算融合架构」，第二代超能效NPU功耗省40%，超性能NPU 1090的大模型预填充性能提升51%，每瓦生成词元数最高提升55%，可在端侧运行最高30B的MoE模型。","extraction":{"state":"ok","confidence":0.9,"error":null,"explanation":"High confidence: full text extraction produced 458 characters.","diagnostics_url":"/api/diagnose?url=https%3A//m.sohu.com/a/1076445217_122014422%3Fscm%3D10001.325_13-325_13.0.0-0-0-0-0.5_1334","quality_profile":{"profile_version":"extraction_quality.v2","bucket":"high","confidence":0.9,"failure_kind":"none","retryable":false,"retry_after_attempts":0,"reason":"High confidence: full text extraction produced 458 characters.","operator_guidance":{"severity":"ok","recommended_action":"trust_full_text","next_step":"Use the extracted full text as the primary article source.","operator_label":"Ready","can_retry":false,"can_use_summary":false,"diagnostics_required":false},"content_depth":{"contract_version":"content_depth.v1","category":"full_text","label":"Full text","has_full_text":true,"has_summary":true,"content_length":458,"summary_length":458,"usable_text_length":458,"source_field":"content"},"legacy_collapsed":false,"signals":{"extract_state":"ok","extract_error":null,"extract_retries":0,"content_length":458,"summary_length":458}}},"display_formats":["compact","card","full","digest_section","json"]},"daily_stack_record":{"title":"能跑30B大模型！天玑9600系列发布：AI性能大幅升级 - Sohu","url":"https://m.sohu.com/a/1076445217_122014422?scm=10001.325_13-325_13.0.0-0-0-0-0.5_1334","summary":"能跑30B大模型！天玑9600系列发布：AI性能大幅升级\n（来源：雷科技）\n9月15日，联发科发布天玑9600系列旗舰芯片，包含天玑9600 Pro和天玑9600M。这是联发科第一款台积电2nm手机芯片。同一天，vivo官宣X500 Pro系列全球首发这颗芯片，OPPO Find X10 Pro Max也在首批搭载名单中。\n参数层面，天玑9600 Pro用了2+3+3的全大核CPU架构，2颗4.55GHz的C2-Ultra超大核，外加6颗C2-Pro大核（3颗4.35GHz、3颗3.1GHz），缓存总容量34.5MB，率先支持LPDDR6内存和UFS 5.0闪存。官方给出的数字是单核性能提升17%、多核提升15%，多核在峰值性能下的功耗比上一代省了61%。\nAI性能是它的一大亮点，天玑9600 Pro把CPU和NPU绑在一起调度，官方称为「智算融合架构」，第二代超能效NPU功耗省40%，超性能NPU 1090的大模型预填充性能提升51%，每瓦生成词元数最高提升55%，可在端侧运行最高30B的MoE模型。","source":"Sohu","date":"2026-09-15T11:35:26+00:00","content":"能跑30B大模型！天玑9600系列发布：AI性能大幅升级\n（来源：雷科技）\n9月15日，联发科发布天玑9600系列旗舰芯片，包含天玑9600 Pro和天玑9600M。这是联发科第一款台积电2nm手机芯片。同一天，vivo官宣X500 Pro系列全球首发这颗芯片，OPPO Find X10 Pro Max也在首批搭载名单中。\n参数层面，天玑9600 Pro用了2+3+3的全大核CPU架构，2颗4.55GHz的C2-Ultra超大核，外加6颗C2-Pro大核（3颗4.35GHz、3颗3.1GHz），缓存总容量34.5MB，率先支持LPDDR6内存和UFS 5.0闪存。官方给出的数字是单核性能提升17%、多核提升15%，多核在峰值性能下的功耗比上一代省了61%。\nAI性能是它的一大亮点，天玑9600 Pro把CPU和NPU绑在一起调度，官方称为「智算融合架构」，第二代超能效NPU功耗省40%，超性能NPU 1090的大模型预填充性能提升51%，每瓦生成词元数最高提升55%，可在端侧运行最高30B的MoE模型。","confidence":0.9,"diagnostics_url":"/api/diagnose?url=https%3A//m.sohu.com/a/1076445217_122014422%3Fscm%3D10001.325_13-325_13.0.0-0-0-0-0.5_1334","quality_bucket":"high","failure_kind":"none","retryable":false,"quality_reason":"High confidence: full text extraction produced 458 characters.","quality_profile":{"profile_version":"extraction_quality.v2","bucket":"high","confidence":0.9,"failure_kind":"none","retryable":false,"retry_after_attempts":0,"reason":"High confidence: full text extraction produced 458 characters.","operator_guidance":{"severity":"ok","recommended_action":"trust_full_text","next_step":"Use the extracted full text as the primary article source.","operator_label":"Ready","can_retry":false,"can_use_summary":false,"diagnostics_required":false},"content_depth":{"contract_version":"content_depth.v1","category":"full_text","label":"Full text","has_full_text":true,"has_summary":true,"content_length":458,"summary_length":458,"usable_text_length":458,"source_field":"content"},"legacy_collapsed":false,"signals":{"extract_state":"ok","extract_error":null,"extract_retries":0,"content_length":458,"summary_length":458}},"tags":[]},"fallback_formats":["markdown","json","html"],"actions":{"read":"/item/83104","export_markdown":"/api/items/83104/export?format=markdown","export_json":"/api/items/83104/export?format=json","diagnose":"/api/diagnose?url=https%3A//m.sohu.com/a/1076445217_122014422%3Fscm%3D10001.325_13-325_13.0.0-0-0-0-0.5_1334"},"formats":{"full":{"id":83104,"title":"能跑30B大模型！天玑9600系列发布：AI性能大幅升级 - Sohu","url":"https://m.sohu.com/a/1076445217_122014422?scm=10001.325_13-325_13.0.0-0-0-0-0.5_1334","source":"Sohu","author":null,"published_at":"2026-09-15T11:35:26+00:00","locale":"zh","topic":"ai","tags":[],"excerpt":"能跑30B大模型！天玑9600系列发布：AI性能大幅升级\n（来源：雷科技）\n9月15日，联发科发布天玑9600系列旗舰芯片，包含天玑9600 Pro和天玑9600M。这是联发科第一款台积电2nm手机芯片。同一天，vivo官宣X500 Pro系列全球首发这颗芯片，OPPO Find X10 Pro Max也在首批搭载名单中。\n参数层面，天玑9600 Pro用了2+3+3的全大核CPU架构，2颗4.55GHz的C2-Ultra超大核，外加6颗C2-Pro大核（3颗4.35GHz、3颗3.1GHz），缓存总容量34.5MB，率先支持LPDDR6内存和UFS 5.0闪存。官方给出的数字是单核性能提升17%、多核提升15%，多核在峰值性能下的功耗比上一代省了61%。\nAI性能是它的一大亮点，天玑9600 Pro把CPU和NPU绑在一起调度，官方称为「智算融合架构」，第二代超能效NPU功耗省40%，超性能NPU 1090的大模型预填充性能提升51%，每瓦生成词元数最高提升55%，可在端侧运行最高30B的MoE模型。","full_text":"能跑30B大模型！天玑9600系列发布：AI性能大幅升级\n（来源：雷科技）\n9月15日，联发科发布天玑9600系列旗舰芯片，包含天玑9600 Pro和天玑9600M。这是联发科第一款台积电2nm手机芯片。同一天，vivo官宣X500 Pro系列全球首发这颗芯片，OPPO Find X10 Pro Max也在首批搭载名单中。\n参数层面，天玑9600 Pro用了2+3+3的全大核CPU架构，2颗4.55GHz的C2-Ultra超大核，外加6颗C2-Pro大核（3颗4.35GHz、3颗3.1GHz），缓存总容量34.5MB，率先支持LPDDR6内存和UFS 5.0闪存。官方给出的数字是单核性能提升17%、多核提升15%，多核在峰值性能下的功耗比上一代省了61%。\nAI性能是它的一大亮点，天玑9600 Pro把CPU和NPU绑在一起调度，官方称为「智算融合架构」，第二代超能效NPU功耗省40%，超性能NPU 1090的大模型预填充性能提升51%，每瓦生成词元数最高提升55%，可在端侧运行最高30B的MoE模型。","reading_time_min":1,"extraction":{"state":"ok","confidence":0.9,"error":null,"explanation":"High confidence: full text extraction produced 458 characters.","diagnostics_url":"/api/diagnose?url=https%3A//m.sohu.com/a/1076445217_122014422%3Fscm%3D10001.325_13-325_13.0.0-0-0-0-0.5_1334","quality_profile":{"profile_version":"extraction_quality.v2","bucket":"high","confidence":0.9,"failure_kind":"none","retryable":false,"retry_after_attempts":0,"reason":"High confidence: full text extraction produced 458 characters.","operator_guidance":{"severity":"ok","recommended_action":"trust_full_text","next_step":"Use the extracted full text as the primary article source.","operator_label":"Ready","can_retry":false,"can_use_summary":false,"diagnostics_required":false},"content_depth":{"contract_version":"content_depth.v1","category":"full_text","label":"Full text","has_full_text":true,"has_summary":true,"content_length":458,"summary_length":458,"usable_text_length":458,"source_field":"content"},"legacy_collapsed":false,"signals":{"extract_state":"ok","extract_error":null,"extract_retries":0,"content_length":458,"summary_length":458}}},"quality_profile":{"profile_version":"extraction_quality.v2","bucket":"high","confidence":0.9,"failure_kind":"none","retryable":false,"retry_after_attempts":0,"reason":"High confidence: full text extraction produced 458 characters.","operator_guidance":{"severity":"ok","recommended_action":"trust_full_text","next_step":"Use the extracted full text as the primary article source.","operator_label":"Ready","can_retry":false,"can_use_summary":false,"diagnostics_required":false},"content_depth":{"contract_version":"content_depth.v1","category":"full_text","label":"Full text","has_full_text":true,"has_summary":true,"content_length":458,"summary_length":458,"usable_text_length":458,"source_field":"content"},"legacy_collapsed":false,"signals":{"extract_state":"ok","extract_error":null,"extract_retries":0,"content_length":458,"summary_length":458}},"actions":{"read":"/item/83104","export_markdown":"/api/items/83104/export?format=markdown","export_json":"/api/items/83104/export?format=json","diagnose":"/api/diagnose?url=https%3A//m.sohu.com/a/1076445217_122014422%3Fscm%3D10001.325_13-325_13.0.0-0-0-0-0.5_1334"}},"digest":{"id":83104,"title":"能跑30B大模型！天玑9600系列发布：AI性能大幅升级 - Sohu","url":"https://m.sohu.com/a/1076445217_122014422?scm=10001.325_13-325_13.0.0-0-0-0-0.5_1334","source":"Sohu","topic":"ai","published_at":"2026-09-15T11:35:26+00:00","excerpt":"能跑30B大模型！天玑9600系列发布：AI性能大幅升级 （来源：雷科技） 9月15日，联发科发布天玑9600系列旗舰芯片，包含天玑9600 Pro和天玑9600M。这是联发科第一款台积电2nm手机芯片。同一天，vivo官宣X500 Pro系列全球首发这颗芯片，OPPO Find X10 Pro Max也在首批搭载名单中。 参数层面，天玑9600 Pro用了2+3+3的全大核CPU架构，2颗4.55GHz的C2-Ultra超大核，外加6颗C2-Pro大核（3颗4.35GHz、3颗3.1GHz），缓存总容量34.5MB，率先支持LPDDR6内存和UFS…","quality_bucket":"high","quality_reason":"High confidence: full text extraction produced 458 characters.","reading_time_min":1,"cluster_id":null},"card":{"display_title":"能跑30B大模型！天玑9600系列发布：AI性能大幅升级 - Sohu","subtitle":"Sohu · 2026-09-15","summary":"能跑30B大模型！天玑9600系列发布：AI性能大幅升级 （来源：雷科技） 9月15日，联发科发布天玑9600系列旗舰芯片，包含天玑9600 Pro和天玑9600M。这是联发科第一款台积电2nm手机芯片。同一天，vivo官宣X500 Pro系列全球首发这颗芯片，OPPO Find X10 Pro Max也在首批搭载名单中。 参数层面，天玑9600…","badges":["quality:high"],"links":{"read":"/item/83104","original":"https://m.sohu.com/a/1076445217_122014422?scm=10001.325_13-325_13.0.0-0-0-0-0.5_1334","diagnose":"/api/diagnose?url=https%3A//m.sohu.com/a/1076445217_122014422%3Fscm%3D10001.325_13-325_13.0.0-0-0-0-0.5_1334"},"quality_warning":null},"export":{"title":"能跑30B大模型！天玑9600系列发布：AI性能大幅升级 - Sohu","url":"https://m.sohu.com/a/1076445217_122014422?scm=10001.325_13-325_13.0.0-0-0-0-0.5_1334","summary":"能跑30B大模型！天玑9600系列发布：AI性能大幅升级\n（来源：雷科技）\n9月15日，联发科发布天玑9600系列旗舰芯片，包含天玑9600 Pro和天玑9600M。这是联发科第一款台积电2nm手机芯片。同一天，vivo官宣X500 Pro系列全球首发这颗芯片，OPPO Find X10 Pro Max也在首批搭载名单中。\n参数层面，天玑9600 Pro用了2+3+3的全大核CPU架构，2颗4.55GHz的C2-Ultra超大核，外加6颗C2-Pro大核（3颗4.35GHz、3颗3.1GHz），缓存总容量34.5MB，率先支持LPDDR6内存和UFS 5.0闪存。官方给出的数字是单核性能提升17%、多核提升15%，多核在峰值性能下的功耗比上一代省了61%。\nAI性能是它的一大亮点，天玑9600 Pro把CPU和NPU绑在一起调度，官方称为「智算融合架构」，第二代超能效NPU功耗省40%，超性能NPU 1090的大模型预填充性能提升51%，每瓦生成词元数最高提升55%，可在端侧运行最高30B的MoE模型。","source":"Sohu","date":"2026-09-15T11:35:26+00:00","content":"能跑30B大模型！天玑9600系列发布：AI性能大幅升级\n（来源：雷科技）\n9月15日，联发科发布天玑9600系列旗舰芯片，包含天玑9600 Pro和天玑9600M。这是联发科第一款台积电2nm手机芯片。同一天，vivo官宣X500 Pro系列全球首发这颗芯片，OPPO Find X10 Pro Max也在首批搭载名单中。\n参数层面，天玑9600 Pro用了2+3+3的全大核CPU架构，2颗4.55GHz的C2-Ultra超大核，外加6颗C2-Pro大核（3颗4.35GHz、3颗3.1GHz），缓存总容量34.5MB，率先支持LPDDR6内存和UFS 5.0闪存。官方给出的数字是单核性能提升17%、多核提升15%，多核在峰值性能下的功耗比上一代省了61%。\nAI性能是它的一大亮点，天玑9600 Pro把CPU和NPU绑在一起调度，官方称为「智算融合架构」，第二代超能效NPU功耗省40%，超性能NPU 1090的大模型预填充性能提升51%，每瓦生成词元数最高提升55%，可在端侧运行最高30B的MoE模型。","confidence":0.9,"diagnostics_url":"/api/diagnose?url=https%3A//m.sohu.com/a/1076445217_122014422%3Fscm%3D10001.325_13-325_13.0.0-0-0-0-0.5_1334","quality_bucket":"high","failure_kind":"none","retryable":false,"quality_reason":"High confidence: full text extraction produced 458 characters.","quality_profile":{"profile_version":"extraction_quality.v2","bucket":"high","confidence":0.9,"failure_kind":"none","retryable":false,"retry_after_attempts":0,"reason":"High confidence: full text extraction produced 458 characters.","operator_guidance":{"severity":"ok","recommended_action":"trust_full_text","next_step":"Use the extracted full text as the primary article source.","operator_label":"Ready","can_retry":false,"can_use_summary":false,"diagnostics_required":false},"content_depth":{"contract_version":"content_depth.v1","category":"full_text","label":"Full text","has_full_text":true,"has_summary":true,"content_length":458,"summary_length":458,"usable_text_length":458,"source_field":"content"},"legacy_collapsed":false,"signals":{"extract_state":"ok","extract_error":null,"extract_retries":0,"content_length":458,"summary_length":458}},"tags":[],"format_contract_version":"news_item_formats.v1"}}}