←Search results

Search “Embodied AI”

By relevance · Total 27 results Clear search

Wednesday

  1. Leiphone

    LimX Dynamics' Open-Source FluxVLA Engine Update Connects the Entire Embodied Intelligence Pipeline from Data and Training to Evaluation and Deployment

    LimX Dynamics' open-source embodied intelligence engineering foundation FluxVLA Engine completed a major update on October 22. The technical report shows that it has incorporated heterogeneous policies such as OpenVLA, π0.5, DiT4DiT, and GR00T into a single pipeline for data, training, evaluation, and deployment. On the NVIDIA AGX Orin 64GB, GR00T's inference frequency increased from 3.2Hz to 7.4Hz, and π0.5 from 1.4Hz to 4.4Hz; together with RTC and a three-step trajectory post-processing mechanism, it connects to real-robot execution. In terms of online RL, SmolVLA's success rate on LIBERO-10 rose from 66.2% to 76.2%, and on 10 tasks in RoboTwin Clean from 45.9% to 61.3%.

    LimX Dynamics' open-source embodied intelligence engineering foundation FluxVLA Engine completed a major update on October 22. The technical report shows that it has incorporated heterogeneous policies such as OpenVLA, π0.5, DiT4DiT, and GR00T into a single pipeline for data, training, evaluation, and deployment. On the NVIDIA AGX Orin 64GB, GR00T's inference frequency increased from 3.2Hz to 7.4Hz, and π0.5 from 1.4Hz to 4.4Hz; together with RTC and a three-step trajectory post-processing mecha

    Matched fieldsTitle、Summary、Body

    #hot-signal#English text is being prepared.#Embodied AI#English text is being prepared.
  2. ZhiDongXi

    Iluvatar CoreX First Live Demonstration of TianGai 300 and On-Device Terminal TY1200

    Iluvatar CoreX held the first live demonstration of its general-purpose GPU TianGai 300 and the on-device compute terminal TY1200, and previewed the next-generation on-device product TY2000 with a nominal compute of 600T. The TianGai 300 has 144GB of memory, delivering Attention performance 1.05 times that of the H200 under a 64K long context and MoE throughput 1.15 times that of the H200 on the GLM-5.3 model; the TY1200 offers 300 TOPS INT8 dense compute, with measured TPS surpassing NVIDIA's Thor T5000. The TianGai 300 IXL144 supernode integrates 144 GPUs in a single rack and aggregates about 10.1TB of memory, reducing the cost per million effective tokens by about 40% compared with mainstream international solutions.

    Iluvatar CoreX held the first live demonstration of its general-purpose GPU TianGai 300 and the on-device compute terminal TY1200, and previewed the next-generation on-device product TY2000 with a nominal compute of 600T. The TianGai 300 has 144GB of memory, delivering Attention performance 1.05 times that of the H200 under a 64K long context and MoE throughput 1.15 times that of the H200 on the GLM-5.3 model; the TY1200 offers 300 TOPS INT8 dense compute, with measured TPS surpassing NVIDIA's T

    Matched fieldsBody

    #hot-signal#Product updates#English text is being prepared.#Reasoning

Monday

Thursday

Friday

Thursday

Tuesday

Sunday

Tuesday

Tuesday

Friday

  1. Story1 reports · 1 facts

    Zhiyuan releases GO-3 Preview, AHA3 and G2 Max

    On October 29, 2026, Zhiyuan released its new-generation embodied intelligence foundation model GO-3 Preview, along with the AHA3 hardware platform and the industrial robot G2 Max. According to a report by Shangguan News, GO-3 Preview connects the entire process from training and simulation practice to continuous learning after taking up a post, ranking first in three tests: RoboDojo, RoboTwin, and RoboColiseum, and is planned to be fully open-sourced by the end of 2026; in a real-world convenience store scenario test, the Elf G2 robot's task success rate was 28.6% on the first day and rose to 85.7% after five days. QbitAI reported that Zhiyuan also released the omni-modal interaction large model WITA-OMNI and the embodied native operating system Lingqu OS at the same event, and proposed the unit Worken for measuring a robot's effective intelligent working hours. On October 30, Zhiyuan co-founder Yao Maoqing said in an interview with Securities Times that general large models have obvious shortcomings in physical robot operation, and that future systems will move toward a layered architecture; he mentioned that GO-3 Preview adopts the WRAM native architecture, and its zero-shot average success rate on 100 tasks across 20 skill categories increased from 13.4% to 56.4%.

    On October 29, 2026, Zhiyuan released its new-generation embodied intelligence foundation model GO-3 Preview, along with the AHA3 hardware platform and the industrial robot G2 Max. According to a report by Shangguan News, GO-3 Preview connects the entire process from training and simulation practice to continuous learning after taking up a post, ranking first in three tests: RoboDojo, RoboTwin, and RoboColiseum, and is planned to be fully open-sourced by the end of 2026; in a real-world convenie

    Matched fieldsSummary

  2. Sina Tech

    Zhiyuan Unveils GO-3 Preview Foundation Model, Genie Master Framework, and Genie G2 Max Robot

    At the Embodied Work Productivity sub-forum of the 2026 Zhiyuan Innovation Day, Zhiyuan unveiled the embodied foundation model GO-3 Preview, the embodied Agent framework Genie Master, and the heavy-load embodied industrial robot Genie G2 Max. GO-3 Preview adopts the unified WRAM architecture for understanding, prediction, and action generation; on 100 tasks across 20 skill categories, its Zero-Shot average success rate rose from 13.4% to 56.4%, and in model-scale expansion experiments it rose from 56.2% to 73.2%; it has also topped the three major evaluation leaderboards RoboDojo, RoboTwin, and RoboColiseum. Co-President Yao Maoqing proposed three capability standards—measurable, replicable, and sustainable—and the event also presented data including a 98.3% success rate on a phone packaging task and the Genie G2's convenience-store task success rate rising from 28.6% on the first day to 85.7% by the fifth day.

    At the Embodied Work Productivity sub-forum of the 2026 Zhiyuan Innovation Day, Zhiyuan unveiled the embodied foundation model GO-3 Preview, the embodied Agent framework Genie Master, and the heavy-load embodied industrial robot Genie G2 Max. GO-3 Preview adopts the unified WRAM architecture for understanding, prediction, and action generation; on 100 tasks across 20 skill categories, its Zero-Shot average success rate rose from 13.4% to 56.4%, and in model-scale expansion experiments it rose fr

    Matched fieldsBody

    #chinese#Industry news#Embodied AI#Agents
  3. Phoenix Technology

    XHEART Releases Large-Model Chip X7 Ultra with 700 TOPS of Compute

    XHEART has released the X7 Ultra, a new-generation chip designed for billion-parameter-level smart-driving large models, with 700 TOPS of compute. The company says that in real-world tests using a partner's production Transformer smart-driving model, compute utilization reached 98.7%, and that while matching the model performance of industry flagship chips, memory bandwidth consumption dropped by 80% and inference energy efficiency improved by 100%. The X7 series has shipped a cumulative 200,000 units in 8 months of mass production, equipping 10 production vehicle models, with the first launch model being the SAIC Volkswagen ID.ERA 9X; at the same event, Momenta released the R7 Ultra world model, whose parameter count is 5 to 10 times higher than the previous generation.

    XHEART has released the X7 Ultra, a new-generation chip designed for billion-parameter-level smart-driving large models, with 700 TOPS of compute. The company says that in real-world tests using a partner's production Transformer smart-driving model, compute utilization reached 98.7%, and that while matching the model performance of industry flagship chips, memory bandwidth consumption dropped by 80% and inference energy efficiency improved by 100%. The X7 series has shipped a cumulative 200,000

    Matched fieldsBody

    #chinese#Product updates#English text is being prepared.#Reasoning

Thursday

  1. FactRelated story

    Zhiyuan releases GO-3 Preview, AHA3 and G2 Max

    On October 29, 2026, Zhiyuan released its new-generation embodied intelligence foundation model GO-3 Preview, along with the AHA3 hardware platform and the industrial robot G2 Max. According to a report by Shangguan News, GO-3 Preview connects the entire process from training and simulation practice to continuous learning after taking up a post, ranking first in three tests: RoboDojo, RoboTwin, and RoboColiseum, and is planned to be fully open-sourced by the end of 2026; in a real-world convenience store scenario test, the Elf G2 robot's task success rate was 28.6% on the first day and rose to 85.7% after five days. QbitAI reported that Zhiyuan also released the omni-modal interaction large model WITA-OMNI and the embodied native operating system Lingqu OS at the same event, and proposed the unit Worken for measuring a robot's effective intelligent working hours. On October 30, Zhiyuan co-founder Yao Maoqing said in an interview with Securities Times that general large models have obvious shortcomings in physical robot operation, and that future systems will move toward a layered architecture; he mentioned that GO-3 Preview adopts the WRAM native architecture, and its zero-shot average success rate on 100 tasks across 20 skill categories increased from 13.4% to 56.4%.

    Matched fieldsSummary

    View story:Zhiyuan releases GO-3 Preview, AHA3 and G2 Max

Wednesday

  1. Story1 reports · 1 facts

    SenseTime Open-Sources SenseNova-RoboRSI Framework and 27B Embodied Model

    SenseTime has open-sourced the SenseNova-RoboRSI framework and a 27B embodied model, while also releasing an empirical study on embodied intelligence for GPT-6-Astra. The study covers 12 simulation benchmarks and 8 real robot platforms. Under a unified protocol (ICL plus up to 10x interaction budget), GPT-6-Astra achieved SOTA on all 12 benchmarks, with RoboDojo's success rate rising from 27.6% at 1x budget to 81.9% at 10x, and DexJoCo reaching 87.3%, compared with the previous SOTA of 40.3%. The study also notes that simple tasks such as putting a pen into a pen holder cost about $55 in inference, and that real robots overheat and wait too long; the conclusion is that GPT-6-Astra has essentially crossed the threshold of embodied capability, but practical-grade embodied intelligence has not yet arrived. The open-source framework packages each benchmark as an independent Docker image and provides environment interaction and scoring through a unified HTTP API; the 27B model raises the success rate on RoboCasa365 from 26.0% to 46.5%, reduces task token overhead by 73%, and costs about 1/50 of GPT-6-Astra per completed task.

    SenseTime has open-sourced the SenseNova-RoboRSI framework and a 27B embodied model, while also releasing an empirical study on embodied intelligence for GPT-6-Astra. The study covers 12 simulation benchmarks and 8 real robot platforms. Under a unified protocol (ICL plus up to 10x interaction budget), GPT-6-Astra achieved SOTA on all 12 benchmarks, with RoboDojo's success rate rising from 27.6% at 1x budget to 81.9% at 10x, and DexJoCo reaching 87.3%, compared with the previous SOTA of 40.3%. Th

    Matched fieldsSummary