{"id":1181587,"date":"2026-08-12T22:25:53","date_gmt":"2026-08-13T05:25:53","guid":{"rendered":"https:\/\/research.codeghost.online\/en-us\/research\/?post_type=msr-research-item&#038;p=1181587"},"modified":"2026-08-12T23:46:11","modified_gmt":"2026-08-13T06:46:11","slug":"improving-training-time-and-gpu-utilization-in-geo-distributed-language-model-training","status":"publish","type":"msr-research-item","link":"https:\/\/research.codeghost.online\/en-us\/research\/publication\/improving-training-time-and-gpu-utilization-in-geo-distributed-language-model-training\/","title":{"rendered":"Improving training time and GPU utilization in geo-distributed language model training"},"content":{"rendered":"\n\n\n<p class=\"wp-block-paragraph\">The widespread adoption of language models (LMs) has caused a huge surge in demand for GPUs. Training large LMs requires tens of thousands of GPUs and housing them in the same datacenter (DC) is a challenge due to many constraints including availability of peak power. We focus on training such models across multiple DCs connected via the Wide-Area-Network (WAN). We built Atlas that speeds up the training time using novel workload-aware temporal bandwidth sharing and other design choices. While Atlas improves the training time, it does not completely eliminate the bubbles (idle GPU cycles). We built BubbleTea that runs prefill-as-a-service (part of LM inference) during the bubbles thus improving the GPU utilization without any impact on training. Compared to state-of-the-art designs, Atlas and BubbleTea together achieve up to 17x faster training, and up to 94% GPU utilization.\u00a0<\/p>\n","protected":false},"excerpt":{"rendered":"<p>The widespread adoption of language models (LMs) has caused a huge surge in demand for GPUs. Training large LMs requires tens of thousands of GPUs and housing them in the same datacenter (DC) is a challenge due to many constraints including availability of peak power. We focus on training such models across multiple DCs connected [&hellip;]<\/p>\n","protected":false},"featured_media":0,"template":"","meta":{"msr-url-field":"","msr-podcast-episode":"","msrModifiedDate":"","msrModifiedDateEnabled":false,"ep_exclude_from_search":false,"_classifai_error":"","msr-author-ordering":[{"type":"text","value":"Palak LNU","user_id":0},{"type":"text","value":"Tella Rajashekhar Reddy","user_id":0},{"type":"text","value":"Bhaskar Kataria","user_id":0},{"type":"user_nicename","value":"Rohan Gandhi","user_id":"42372"},{"type":"text","value":"Karan Tandon","user_id":0},{"type":"text","value":"Debopam Bhattacherjee","user_id":0},{"type":"user_nicename","value":"Venkat Padmanabhan","user_id":"33180"}],"msr_publishername":"","msr_publisher_other":"","msr_booktitle":"","msr_chapter":"","msr_edition":"","msr_editors":"","msr_how_published":"","msr_isbn":"","msr_issue":"","msr_journal":"","msr_number":"","msr_organization":"","msr_pages_string":"","msr_page_range_start":"","msr_page_range_end":"","msr_series":"","msr_volume":"","msr_copyright":"","msr_conference_name":"","msr_doi":"","msr_arxiv_id":"","msr_mag_id":"","msr_other_authors":"","msr_other_contributors":"","msr_speaker":"","msr_award":"","msr_affiliation":"","msr_institution":"","msr_host":"","msr_version":"","msr_duration":"","msr_release_tracker_id":"","msr_highlight_type":"","msr_date_display_format":"","msr_main_download_label":"","msr_external_link_label":"","msr_doi_label":"","msr_published_date":"2026-01","msr_startdate":"","msr_presentation_date":"","msr_highlight_text":"","msr_notes":"","msr_longbiography":"","msr_publicationurl":"","msr_external_url":"","msr_secondary_video_url":"","msr_conference_url":"","msr_journal_url":"","msr_year":2026,"msr_month":1,"msr_day":0,"msr_microsoftintellectualproperty":false,"msr_pub_id":"","msr_publication_uploader":[{"type":"file","title":"geo-dist-arxiv.pdf","label_id":243132,"id":1181588,"viewUrl":"https:\/\/research.codeghost.online\/en-us\/research\/wp-content\/uploads\/2026\/08\/geo-dist-arxiv.pdf"}],"msr_related_uploader":[],"msr_original_fields_of_study":[],"msr_s2_paper_id":"","msr_s2_pdf_url":"","msr_citation_count_updated":"","msr_citation_count":0,"msr_influential_citations":0,"msr_reference_count":0,"msr_s2_open_access":false,"msr_s2_author_ids":[],"msr_pub_ids":[],"msr_hide_image_in_river":0,"footnotes":""},"msr-research-highlight":[],"research-area":[13556,13547],"msr-publication-type":[193715],"msr-publisher":[],"msr-publication-cta":[],"msr-focus-area":[],"msr-locale":[268875],"msr-post-option":[],"msr-field-of-study":[],"msr-conference":[],"msr-journal":[268188],"msr-impact-theme":[],"msr-pillar":[],"class_list":["post-1181587","msr-research-item","type-msr-research-item","status-publish","hentry","msr-research-area-artificial-intelligence","msr-research-area-systems-and-networking","msr-locale-en_us"],"msr_publishername":"","msr_edition":"","msr_affiliation":"","msr_published_date":"2026-01","msr_host":"","msr_duration":"","msr_version":"","msr_speaker":"","msr_other_contributors":"","msr_booktitle":"","msr_pages_string":"","msr_chapter":"","msr_isbn":"","msr_journal":"","msr_volume":"","msr_number":"","msr_editors":"","msr_series":"","msr_issue":"","msr_organization":"","msr_how_published":"","msr_notes":"","msr_highlight_text":"","msr_release_tracker_id":"","msr_original_fields_of_study":"","msr_download_urls":"","msr_external_url":"","msr_secondary_video_url":"","msr_longbiography":"","msr_microsoftintellectualproperty":0,"msr_main_download":"","msr_publicationurl":"","msr_doi":"","msr_publication_uploader":[{"type":"file","title":"geo-dist-arxiv.pdf","label_id":243132,"id":1181588,"viewUrl":"https:\/\/research.codeghost.online\/en-us\/research\/wp-content\/uploads\/2026\/08\/geo-dist-arxiv.pdf"}],"msr_related_uploader":[],"msr_citation_count":0,"msr_citation_count_updated":"","msr_s2_paper_id":"","msr_influential_citations":0,"msr_reference_count":0,"msr_arxiv_id":"","msr_s2_author_ids":[],"msr_s2_open_access":false,"msr_s2_pdf_url":null,"msr_attachments":[{"id":1181588,"url":"https:\/\/research.codeghost.online\/en-us\/research\/wp-content\/uploads\/2026\/08\/geo-dist-arxiv.pdf"}],"msr-author-ordering":[{"type":"text","value":"Palak LNU","user_id":0,"rest_url":false},{"type":"text","value":"Tella Rajashekhar Reddy","user_id":0,"rest_url":false},{"type":"text","value":"Bhaskar Kataria","user_id":0,"rest_url":false},{"type":"user_nicename","value":"Rohan Gandhi","user_id":42372,"rest_url":"https:\/\/research.codeghost.online\/en-us\/research\/wp-json\/microsoft-research\/v1\/researchers?person=Rohan Gandhi"},{"type":"text","value":"Karan Tandon","user_id":0,"rest_url":false},{"type":"text","value":"Debopam Bhattacherjee","user_id":0,"rest_url":false},{"type":"user_nicename","value":"Venkat Padmanabhan","user_id":33180,"rest_url":"https:\/\/research.codeghost.online\/en-us\/research\/wp-json\/microsoft-research\/v1\/researchers?person=Venkat Padmanabhan"}],"msr_impact_theme":[],"msr_research_lab":[199562],"msr_event":[],"msr_group":[144939],"msr_project":[],"publication":[],"video":[],"msr-tool":[],"msr_publication_type":"article","related_content":[],"_links":{"self":[{"href":"https:\/\/research.codeghost.online\/en-us\/research\/wp-json\/wp\/v2\/msr-research-item\/1181587","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/research.codeghost.online\/en-us\/research\/wp-json\/wp\/v2\/msr-research-item"}],"about":[{"href":"https:\/\/research.codeghost.online\/en-us\/research\/wp-json\/wp\/v2\/types\/msr-research-item"}],"version-history":[{"count":2,"href":"https:\/\/research.codeghost.online\/en-us\/research\/wp-json\/wp\/v2\/msr-research-item\/1181587\/revisions"}],"predecessor-version":[{"id":1181590,"href":"https:\/\/research.codeghost.online\/en-us\/research\/wp-json\/wp\/v2\/msr-research-item\/1181587\/revisions\/1181590"}],"wp:attachment":[{"href":"https:\/\/research.codeghost.online\/en-us\/research\/wp-json\/wp\/v2\/media?parent=1181587"}],"wp:term":[{"taxonomy":"msr-research-highlight","embeddable":true,"href":"https:\/\/research.codeghost.online\/en-us\/research\/wp-json\/wp\/v2\/msr-research-highlight?post=1181587"},{"taxonomy":"msr-research-area","embeddable":true,"href":"https:\/\/research.codeghost.online\/en-us\/research\/wp-json\/wp\/v2\/research-area?post=1181587"},{"taxonomy":"msr-publication-type","embeddable":true,"href":"https:\/\/research.codeghost.online\/en-us\/research\/wp-json\/wp\/v2\/msr-publication-type?post=1181587"},{"taxonomy":"msr-publisher","embeddable":true,"href":"https:\/\/research.codeghost.online\/en-us\/research\/wp-json\/wp\/v2\/msr-publisher?post=1181587"},{"taxonomy":"msr-publication-cta","embeddable":true,"href":"https:\/\/research.codeghost.online\/en-us\/research\/wp-json\/wp\/v2\/msr-publication-cta?post=1181587"},{"taxonomy":"msr-focus-area","embeddable":true,"href":"https:\/\/research.codeghost.online\/en-us\/research\/wp-json\/wp\/v2\/msr-focus-area?post=1181587"},{"taxonomy":"msr-locale","embeddable":true,"href":"https:\/\/research.codeghost.online\/en-us\/research\/wp-json\/wp\/v2\/msr-locale?post=1181587"},{"taxonomy":"msr-post-option","embeddable":true,"href":"https:\/\/research.codeghost.online\/en-us\/research\/wp-json\/wp\/v2\/msr-post-option?post=1181587"},{"taxonomy":"msr-field-of-study","embeddable":true,"href":"https:\/\/research.codeghost.online\/en-us\/research\/wp-json\/wp\/v2\/msr-field-of-study?post=1181587"},{"taxonomy":"msr-conference","embeddable":true,"href":"https:\/\/research.codeghost.online\/en-us\/research\/wp-json\/wp\/v2\/msr-conference?post=1181587"},{"taxonomy":"msr-journal","embeddable":true,"href":"https:\/\/research.codeghost.online\/en-us\/research\/wp-json\/wp\/v2\/msr-journal?post=1181587"},{"taxonomy":"msr-impact-theme","embeddable":true,"href":"https:\/\/research.codeghost.online\/en-us\/research\/wp-json\/wp\/v2\/msr-impact-theme?post=1181587"},{"taxonomy":"msr-pillar","embeddable":true,"href":"https:\/\/research.codeghost.online\/en-us\/research\/wp-json\/wp\/v2\/msr-pillar?post=1181587"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}