{"id":90509,"topic":"ai","source":"搜狐网","title":"大模型应用很热，华为为何选择“向下扎根”｜解码广东经济生态 - 搜狐网","url":"https://m.sohu.com/a/1080990788_100116740?scm=10001.325_13-325_13.0.0-0-0-0-0.5_1334","url_hash":"83627d6db05bf2b205e1e5068189e179b513f999","author":"","summary":"<a href=\"https://news.google.com/rss/articles/CBMijAFBVV95cUxNcWtsZkRxSnR2UGlMcHZsTXc4dER1cVQxZXVjSmJUT0xRaTBqVThWQWJHdm5RNzk3aGd0SkxhQmkzMWMzSFRtWXhlYTZqVVdUbFBPTmlGeTQzUGdCOHpLTGF4dTJVZUhEcnFwd3VsOGU2bFJZbEFmX2ZmSjMzQ1VsWW8ybE9wVHRyYWtzTw?oc=5\" target=\"_blank\">大模型应用很热，华为为何选择“向下扎根”｜解码广东经济生态</a>&nbsp;&nbsp;<font color=\"#6f6f6f\">搜狐网</font>","content":"大模型应用很热，华为为何选择“向下扎根”｜解码广东经济生态\n“今天中国日均Token（词元）消耗量约500万亿，而支撑100万亿Token，大约需要7万台通算服务器。”在近日举行的华为全联接大会2026期间，华为计算首席战略官朱照生算了一笔账。\n今年3月，来自国家数据局的数据显示，中国市场日均Token（词元）消耗量还只是140万亿。当Token需求以数量级增长，单靠提升芯片性能已经不够，算力的供给方式需要从底层重新设计。\n受访者 供图\n在先进制程不可得的前提下，中国大模型的算力底座到底能不能撑住？华为给出了答案：华为副董事长、轮值董事长汪涛在全联接大会上表示，华为围绕“超节点+集群”，通过系统架构创新构建竞争力，打造中国坚实算力底座，为世界构建新的选择。\n这份布局未来的组合拳，不仅为中国的AI产业积蓄力量，也传递出中国“硬核科技”突围的信号：当单一技术节点遭遇天花板，系统创新成为新的增长引擎。","image_url":"https://q2.itc.cn/q_70/images03/20260926/9cc92dfad5914d7daff1f8a278bcf595.jpeg","lang":"zh","published_at":"2026-09-25T23:01:16+00:00","fetched_at":"2026-09-26T00:15:08+00:00","status":"read","starred":0,"extract_state":"ok","summary_auto":"大模型应用很热，华为为何选择“向下扎根”｜解码广东经济生态\n“今天中国日均Token（词元）消耗量约500万亿，而支撑100万亿Token，大约需要7万台通算服务器。”在近日举行的华为全联接大会2026期间，华为计算首席战略官朱照生算了一笔账。\n今年3月，来自国家数据局的数据显示，中国市场日均Token（词元）消耗量还只是140万亿。当Token需求以数量级增长，单靠提升芯片性能已经不够，算力的供给方式需要从底层重新设计。\n受访者…","cluster_id":3515017,"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/1080990788_100116740?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 403 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":403,"summary_length":218,"usable_text_length":403,"source_field":"content"},"legacy_collapsed":false,"signals":{"extract_state":"ok","extract_error":null,"extract_retries":0,"content_length":403,"summary_length":218}},"news_item":{"id":90509,"canonical_url":"https://m.sohu.com/a/1080990788_100116740?scm=10001.325_13-325_13.0.0-0-0-0-0.5_1334","source_url":"https://m.sohu.com/a/1080990788_100116740?scm=10001.325_13-325_13.0.0-0-0-0-0.5_1334","title":"大模型应用很热，华为为何选择“向下扎根”｜解码广东经济生态 - 搜狐网","source_name":"搜狐网","author":null,"published_at":"2026-09-25T23:01:16+00:00","locale":"zh","topic":"ai","tags":[],"rss_summary":"<a href=\"https://news.google.com/rss/articles/CBMijAFBVV95cUxNcWtsZkRxSnR2UGlMcHZsTXc4dER1cVQxZXVjSmJUT0xRaTBqVThWQWJHdm5RNzk3aGd0SkxhQmkzMWMzSFRtWXhlYTZqVVdUbFBPTmlGeTQzUGdCOHpLTGF4dTJVZUhEcnFwd3VsOGU2bFJZbEFmX2ZmSjMzQ1VsWW8ybE9wVHRyYWtzTw?oc=5\" target=\"_blank\">大模型应用很热，华为为何选择“向下扎根”｜解码广东经济生态</a>&nbsp;&nbsp;<font color=\"#6f6f6f\">搜狐网</font>","full_text":"大模型应用很热，华为为何选择“向下扎根”｜解码广东经济生态\n“今天中国日均Token（词元）消耗量约500万亿，而支撑100万亿Token，大约需要7万台通算服务器。”在近日举行的华为全联接大会2026期间，华为计算首席战略官朱照生算了一笔账。\n今年3月，来自国家数据局的数据显示，中国市场日均Token（词元）消耗量还只是140万亿。当Token需求以数量级增长，单靠提升芯片性能已经不够，算力的供给方式需要从底层重新设计。\n受访者 供图\n在先进制程不可得的前提下，中国大模型的算力底座到底能不能撑住？华为给出了答案：华为副董事长、轮值董事长汪涛在全联接大会上表示，华为围绕“超节点+集群”，通过系统架构创新构建竞争力，打造中国坚实算力底座，为世界构建新的选择。\n这份布局未来的组合拳，不仅为中国的AI产业积蓄力量，也传递出中国“硬核科技”突围的信号：当单一技术节点遭遇天花板，系统创新成为新的增长引擎。","excerpt":"大模型应用很热，华为为何选择“向下扎根”｜解码广东经济生态\n“今天中国日均Token（词元）消耗量约500万亿，而支撑100万亿Token，大约需要7万台通算服务器。”在近日举行的华为全联接大会2026期间，华为计算首席战略官朱照生算了一笔账。\n今年3月，来自国家数据局的数据显示，中国市场日均Token（词元）消耗量还只是140万亿。当Token需求以数量级增长，单靠提升芯片性能已经不够，算力的供给方式需要从底层重新设计。\n受访者…","extraction":{"state":"ok","confidence":0.9,"error":null,"explanation":"High confidence: full text extraction produced 403 characters.","diagnostics_url":"/api/diagnose?url=https%3A//m.sohu.com/a/1080990788_100116740%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 403 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":403,"summary_length":218,"usable_text_length":403,"source_field":"content"},"legacy_collapsed":false,"signals":{"extract_state":"ok","extract_error":null,"extract_retries":0,"content_length":403,"summary_length":218}}},"display_formats":["compact","card","full","digest_section","json"]},"daily_stack_record":{"title":"大模型应用很热，华为为何选择“向下扎根”｜解码广东经济生态 - 搜狐网","url":"https://m.sohu.com/a/1080990788_100116740?scm=10001.325_13-325_13.0.0-0-0-0-0.5_1334","summary":"大模型应用很热，华为为何选择“向下扎根”｜解码广东经济生态\n“今天中国日均Token（词元）消耗量约500万亿，而支撑100万亿Token，大约需要7万台通算服务器。”在近日举行的华为全联接大会2026期间，华为计算首席战略官朱照生算了一笔账。\n今年3月，来自国家数据局的数据显示，中国市场日均Token（词元）消耗量还只是140万亿。当Token需求以数量级增长，单靠提升芯片性能已经不够，算力的供给方式需要从底层重新设计。\n受访者…","source":"搜狐网","date":"2026-09-25T23:01:16+00:00","content":"大模型应用很热，华为为何选择“向下扎根”｜解码广东经济生态\n“今天中国日均Token（词元）消耗量约500万亿，而支撑100万亿Token，大约需要7万台通算服务器。”在近日举行的华为全联接大会2026期间，华为计算首席战略官朱照生算了一笔账。\n今年3月，来自国家数据局的数据显示，中国市场日均Token（词元）消耗量还只是140万亿。当Token需求以数量级增长，单靠提升芯片性能已经不够，算力的供给方式需要从底层重新设计。\n受访者 供图\n在先进制程不可得的前提下，中国大模型的算力底座到底能不能撑住？华为给出了答案：华为副董事长、轮值董事长汪涛在全联接大会上表示，华为围绕“超节点+集群”，通过系统架构创新构建竞争力，打造中国坚实算力底座，为世界构建新的选择。\n这份布局未来的组合拳，不仅为中国的AI产业积蓄力量，也传递出中国“硬核科技”突围的信号：当单一技术节点遭遇天花板，系统创新成为新的增长引擎。","confidence":0.9,"diagnostics_url":"/api/diagnose?url=https%3A//m.sohu.com/a/1080990788_100116740%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 403 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 403 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":403,"summary_length":218,"usable_text_length":403,"source_field":"content"},"legacy_collapsed":false,"signals":{"extract_state":"ok","extract_error":null,"extract_retries":0,"content_length":403,"summary_length":218}},"tags":[]},"fallback_formats":["markdown","json","html"],"actions":{"read":"/item/90509","export_markdown":"/api/items/90509/export?format=markdown","export_json":"/api/items/90509/export?format=json","diagnose":"/api/diagnose?url=https%3A//m.sohu.com/a/1080990788_100116740%3Fscm%3D10001.325_13-325_13.0.0-0-0-0-0.5_1334"},"formats":{"full":{"id":90509,"title":"大模型应用很热，华为为何选择“向下扎根”｜解码广东经济生态 - 搜狐网","url":"https://m.sohu.com/a/1080990788_100116740?scm=10001.325_13-325_13.0.0-0-0-0-0.5_1334","source":"搜狐网","author":null,"published_at":"2026-09-25T23:01:16+00:00","locale":"zh","topic":"ai","tags":[],"excerpt":"大模型应用很热，华为为何选择“向下扎根”｜解码广东经济生态\n“今天中国日均Token（词元）消耗量约500万亿，而支撑100万亿Token，大约需要7万台通算服务器。”在近日举行的华为全联接大会2026期间，华为计算首席战略官朱照生算了一笔账。\n今年3月，来自国家数据局的数据显示，中国市场日均Token（词元）消耗量还只是140万亿。当Token需求以数量级增长，单靠提升芯片性能已经不够，算力的供给方式需要从底层重新设计。\n受访者…","full_text":"大模型应用很热，华为为何选择“向下扎根”｜解码广东经济生态\n“今天中国日均Token（词元）消耗量约500万亿，而支撑100万亿Token，大约需要7万台通算服务器。”在近日举行的华为全联接大会2026期间，华为计算首席战略官朱照生算了一笔账。\n今年3月，来自国家数据局的数据显示，中国市场日均Token（词元）消耗量还只是140万亿。当Token需求以数量级增长，单靠提升芯片性能已经不够，算力的供给方式需要从底层重新设计。\n受访者 供图\n在先进制程不可得的前提下，中国大模型的算力底座到底能不能撑住？华为给出了答案：华为副董事长、轮值董事长汪涛在全联接大会上表示，华为围绕“超节点+集群”，通过系统架构创新构建竞争力，打造中国坚实算力底座，为世界构建新的选择。\n这份布局未来的组合拳，不仅为中国的AI产业积蓄力量，也传递出中国“硬核科技”突围的信号：当单一技术节点遭遇天花板，系统创新成为新的增长引擎。","reading_time_min":1,"extraction":{"state":"ok","confidence":0.9,"error":null,"explanation":"High confidence: full text extraction produced 403 characters.","diagnostics_url":"/api/diagnose?url=https%3A//m.sohu.com/a/1080990788_100116740%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 403 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":403,"summary_length":218,"usable_text_length":403,"source_field":"content"},"legacy_collapsed":false,"signals":{"extract_state":"ok","extract_error":null,"extract_retries":0,"content_length":403,"summary_length":218}}},"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 403 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":403,"summary_length":218,"usable_text_length":403,"source_field":"content"},"legacy_collapsed":false,"signals":{"extract_state":"ok","extract_error":null,"extract_retries":0,"content_length":403,"summary_length":218}},"actions":{"read":"/item/90509","export_markdown":"/api/items/90509/export?format=markdown","export_json":"/api/items/90509/export?format=json","diagnose":"/api/diagnose?url=https%3A//m.sohu.com/a/1080990788_100116740%3Fscm%3D10001.325_13-325_13.0.0-0-0-0-0.5_1334"}},"digest":{"id":90509,"title":"大模型应用很热，华为为何选择“向下扎根”｜解码广东经济生态 - 搜狐网","url":"https://m.sohu.com/a/1080990788_100116740?scm=10001.325_13-325_13.0.0-0-0-0-0.5_1334","source":"搜狐网","topic":"ai","published_at":"2026-09-25T23:01:16+00:00","excerpt":"大模型应用很热，华为为何选择“向下扎根”｜解码广东经济生态 “今天中国日均Token（词元）消耗量约500万亿，而支撑100万亿Token，大约需要7万台通算服务器。”在近日举行的华为全联接大会2026期间，华为计算首席战略官朱照生算了一笔账。 今年3月，来自国家数据局的数据显示，中国市场日均Token（词元）消耗量还只是140万亿。当Token需求以数量级增长，单靠提升芯片性能已经不够，算力的供给方式需要从底层重新设计。 受访者…","quality_bucket":"high","quality_reason":"High confidence: full text extraction produced 403 characters.","reading_time_min":1,"cluster_id":3515017},"card":{"display_title":"大模型应用很热，华为为何选择“向下扎根”｜解码广东经济生态 - 搜狐网","subtitle":"搜狐网 · 2026-09-25","summary":"大模型应用很热，华为为何选择“向下扎根”｜解码广东经济生态 “今天中国日均Token（词元）消耗量约500万亿，而支撑100万亿Token，大约需要7万台通算服务器。”在近日举行的华为全联接大会2026期间，华为计算首席战略官朱照生算了一笔账。 今年3月，来自国家数据局的数据显示，中国市场日均Token（词元）消耗量还只是140万亿。当Token需求以数量级增长，单靠提升芯片性能已经不够，算力的供给方式需要从底层重新设计。 受访者…","badges":["quality:high"],"links":{"read":"/item/90509","original":"https://m.sohu.com/a/1080990788_100116740?scm=10001.325_13-325_13.0.0-0-0-0-0.5_1334","diagnose":"/api/diagnose?url=https%3A//m.sohu.com/a/1080990788_100116740%3Fscm%3D10001.325_13-325_13.0.0-0-0-0-0.5_1334"},"quality_warning":null},"export":{"title":"大模型应用很热，华为为何选择“向下扎根”｜解码广东经济生态 - 搜狐网","url":"https://m.sohu.com/a/1080990788_100116740?scm=10001.325_13-325_13.0.0-0-0-0-0.5_1334","summary":"大模型应用很热，华为为何选择“向下扎根”｜解码广东经济生态\n“今天中国日均Token（词元）消耗量约500万亿，而支撑100万亿Token，大约需要7万台通算服务器。”在近日举行的华为全联接大会2026期间，华为计算首席战略官朱照生算了一笔账。\n今年3月，来自国家数据局的数据显示，中国市场日均Token（词元）消耗量还只是140万亿。当Token需求以数量级增长，单靠提升芯片性能已经不够，算力的供给方式需要从底层重新设计。\n受访者…","source":"搜狐网","date":"2026-09-25T23:01:16+00:00","content":"大模型应用很热，华为为何选择“向下扎根”｜解码广东经济生态\n“今天中国日均Token（词元）消耗量约500万亿，而支撑100万亿Token，大约需要7万台通算服务器。”在近日举行的华为全联接大会2026期间，华为计算首席战略官朱照生算了一笔账。\n今年3月，来自国家数据局的数据显示，中国市场日均Token（词元）消耗量还只是140万亿。当Token需求以数量级增长，单靠提升芯片性能已经不够，算力的供给方式需要从底层重新设计。\n受访者 供图\n在先进制程不可得的前提下，中国大模型的算力底座到底能不能撑住？华为给出了答案：华为副董事长、轮值董事长汪涛在全联接大会上表示，华为围绕“超节点+集群”，通过系统架构创新构建竞争力，打造中国坚实算力底座，为世界构建新的选择。\n这份布局未来的组合拳，不仅为中国的AI产业积蓄力量，也传递出中国“硬核科技”突围的信号：当单一技术节点遭遇天花板，系统创新成为新的增长引擎。","confidence":0.9,"diagnostics_url":"/api/diagnose?url=https%3A//m.sohu.com/a/1080990788_100116740%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 403 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 403 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":403,"summary_length":218,"usable_text_length":403,"source_field":"content"},"legacy_collapsed":false,"signals":{"extract_state":"ok","extract_error":null,"extract_retries":0,"content_length":403,"summary_length":218}},"tags":[],"format_contract_version":"news_item_formats.v1"}}}