Meta Andromeda：通过模型与系统协同设计驱动下一代广告检索引擎

• 本文介绍了 Meta 个性化广告检索引擎 Meta Andromeda 的下一代演进，以及如何通过协同设计模型、系统和 GPU 硬件，开启广告检索效率的新前沿。
  • 在生产环境中，该系统以检索阶段的规模对所有符合条件的广告进行评分，达到接近精排级别（ranking-level）的模型性能，并以每秒超过 10 亿个广告的吞吐量筛选出头部候选集，创下了广告检索阶段模型吞吐量的新基准（相比 2024 年的 Meta Andromeda 系统提升了 100 倍）。
  • 对于广告主而言，这意味着检索系统在相同延迟预算内能在更大的候选池中匹配广告，从而自动带来更好的效果。自 2025 年以来，Meta Andromeda 已帮助 Facebook 上的广告转化率提升了 2.4%。

Meta Andromeda 是 Meta 用于个性化广告检索的机器学习系统，也是广告推荐流水线的第一阶段。检索阶段会在毫秒级别内将庞大的符合条件的候选广告池缩减为一小部分优质候选集；随后，计算成本更高的高阶精排（ranking）阶段再对这些留存下来的候选广告进行细致打分。

从历史上看，精排阶段深度建模的高质量对于检索阶段的规模来说计算成本过于高昂。基于 2024 年 Andromeda 在 NVIDIA GH200 Grace Hopper 超级芯片上发布所建立的检索基础，本文将分享如何通过模型、系统与硬件的协同设计，将接近精排级别的性能带入检索规模。

检索扩展中的挑战与机遇

2024 年的 Andromeda 部署通过端到端优化的检索实现对系统进行了重构：涵盖从带有分层索引和 GPU 原生算子的定制 AI 硬件，到专用构建的机器学习模型，将模型容量提升了四个数量级。自那时起，生成式 AI 创意工具帮助广告主快速生成了数以亿计的新广告，使 Meta 的广告候选池规模达到了前所未有的程度。

要以更准确的模型和高得多的系统吞吐量对如此庞大的广告候选池进行打分，需要在模型质量、系统架构和硬件利用率方面共同取得基础性技术突破。

更高的模型性能需要新架构

广告检索在增量质量上具有极高的投资回报率（ROI）：检索模型性能的微小提升都能为广告主带来巨大的价值增长。广告候选池的迅速膨胀使得优化硬件算力 ROI 成为一个日益关键的因素。2024 年的 Meta Andromeda 检索模型凭借其针对 GPU 优化的深度神经网络架构奠定了高效的基础。然而，要在当前规模下以可持续的计算成本持续提升检索性能，我们必须超越这种范式。这需要从根本上推动架构演进，以在不按比例增加成本的前提下大幅推高模型性能边界。

十亿级规模下的新系统瓶颈

2024 年的 Meta Andromeda 系统单张 GPU 卡每秒能评估数千万个广告。而满足新的业务需求则需要系统中单卡达到每秒评估数十亿个广告的吞吐量。在每秒数十亿个广告的吞吐量下，在内存层级中传递广告张量的内存带宽成为了新的瓶颈。GPU 内核（kernel）之间的中间张量产生了每秒数 TB 的内存流量，并且在本地 CPU 主机与负责模型评估的远程硬件加速器之间传输广告数据使得网络链路严重饱和。这需要一种全新的系统协同设计，将这些碎片化、分布式的负载统一到单个 GPU 流水线中。

逼近硬件极限

随着检索运行在 GPU 上，关键的优化问题变成了：“我们离硬件天花板还有多近？未来的模型增长还剩下多少空间？”回答这个问题需要将检索流水线的各个组件推向当前硬件世代的极限，这一挑战促成了与 NVIDIA 的联合合作。

上述每一个挑战都推动了相应的技术突破。

通过模型-系统协同设计释放检索性能

下一代检索模型架构——检索混合专家模型（Retrieval MoE）

Meta Andromeda 的最新迭代引入了三项协同的架构创新，将受大语言模型（LLM）启发的混合专家模型（Mixture-of-Experts, MoE）引入到检索中（见图 1）。依托这种表达能力极高且推理高效的架构，系统建立了新的检索性能-效率帕累托前沿：在典型的“检索-精排”系统中，模型性能接近精排级别，同时推理阶段的 FLOPs 相比 2024 年 Meta Andromeda 检索模型减少了约 300 倍。

在用户建模方面，我们从单一用户塔过渡到由专门的用户塔组成的专家集成（ensemble）。这通过整合海量的高保真信号捕捉多样化、随时间演变且具有上下文感知的用户兴趣：时序广告互动事件、用户群组表征、LLM 衍生的语义特征以及跨业务场景的共同互动信号。

在候选广告塔方面，我们采用联合编码策略，引入实时投放上下文，确保下游交互模型获得上下文感知的表征。

在交互顶层结构（overarch layer）方面，交互模型从稠密神经网络演进为基于 MoE 注意力的门控网络，可动态将候选广告路由至最相关的用户专家，以极低比例的 FLOPs 实现了接近精排级别的性能。

新架构的一个决定性特征在于其性能与底层硬件执行模型进行了协同设计。在单张 GH200 GPU 上，模型推理吞吐量已超过每秒 100 亿个广告，这种高效率使得模型容量可以通过更丰富的特征交互、额外的专家子网络以及更深层的架构进一步提升，而不会导致 GPU 成本同比例上升。模型每一次微小的进步都能直接转化为更大的广告主价值。

图 1：Retrieval MoE 架构。多样化的高保真用户信号输入到异构用户专家集成中。广告内容与实时服务上下文融合形成上下文相关的广告输入，驱动广告感知的 MoE 路由器，仅激活 Top-K 最相关的专家以生成用户-广告相关性得分。结果是在检索规模上实现了精排级别的匹配质量——相比 2024 年 Meta Andromeda 减少约 300 倍的推理 FLOPs，并在单颗 Grace Hopper 超级芯片上达到超过 100 亿广告/秒的吞吐量。

将模型与系统协同设计于单一 GPU 流水线上

2026 年版 Meta Andromeda 通过两项创新将模型-系统协同设计推向了新高度——针对每次用户-广告交互融合了模型与系统负载的单 GPU 内核，以及将单个用户请求的逻辑与每次用户-广告的 GPU 内核协同部署在单台 GPU 服务器（single GPU box）内的策略。为此，构建了一套高度优化且全融合的 GPU 流水线，承载端到端的 Meta Andromeda 工作流。

融合内核（Fused Kernel）

从高层来看，Meta Andromeda 在推理期处理每个用户-广告的负载是跨模型执行和系统逻辑的一组紧凑操作序列——汇集 Embedding、执行用户-广告交互模型以及评分。与 2024 年仅融合模型执行内部操作相比，Retrieval MoE 门控架构中约 300 倍的 FLOPs 减少创造了必要的空间，并且为将模型执行与非模型系统负载完全融合到单个 GPU 内核中减少了阻力。因此，融合内核如今在单张 GH200 上可以达到每秒超过 100 亿个广告的运行速度。

这至关重要，因为检索负载从根本上是受内存带宽限制的（memory-bandwidth bound）。在每秒数十亿广告的规模下，即使每个广告产生一个微小的中间张量，累积起来也会导致每秒数 TB 的高带宽内存（HBM）流量。在多内核检索流水线中，这些中间张量的内存读写负载——而非算术运算——成为了主要瓶颈。

融合内核彻底消除了这一瓶颈。它无需为模型和系统操作分派多个小 GPU 内核，而是在单次遍历（one pass）中执行针对每个用户-广告的完整计算。中间结果保留在 GPU 寄存器和线程块共享内存中，绝不会物化到 HBM 中，从而实现了对 GPU 峰值内存带宽的高效利用，而不是浪费在冗余的中间流量上。

协同部署策略（Co-Location Strategy）

单独衡量时，单 GPU 上的用户-广告融合内核能够以每秒数十亿广告的速度运行。然而，每次用户请求的操作（例如索引匹配——即执行广告主的定向规则）以及 Top-K 选择通常在独立的本地 CPU 主机上运行，并通过网络连接与远程加速器上的融合 GPU 内核打分服务通信。在每秒数十亿广告的吞吐量下，仅这些流量就会使网络链路饱和，导致网络带宽取代 GPU 计算成为新的瓶颈。

协同设计的应对方案是收缩系统架构以匹配模型架构：将索引匹配、打分和 Top-K 选择协同部署在同一块 GPU 上。系统不再跨主机传输庞大的候选列表，而是向单台 GPU 服务器发送一份紧凑的请求载荷（用户特征与定向查询），由该服务器在本地运行完整的检索流水线，并且仅返回最终的 Top-K 结果。这使网络流量减少了约 100 倍，并将系统重新拉回到受 GPU 计算约束的区间。

将索引匹配迁移到 GPU 并非易事。广告索引匹配需要跨数十个特征类别评估广告主定义的复杂定向逻辑，并且必须支持全天随着广告创建、修改或过期而进行的实时索引更新。在 GPU 上支持这一功能需要重新设计面向 GPU 并行执行的索引结构，同时保持生产环境定向语义的完整保真度。

Meta Andromeda 的单 GPU 流水线

结合上述两项创新，完全协同部署的单 GPU 流水线在单台 GPU 服务器上可处理每秒超过 10 亿个广告，同时对候选池中的每一个符合条件的广告进行评分。这是广告检索阶段模型的一个全新吞吐量基准，正如下一节所述，系统仍保留着巨大的提升空间。

图 2：Meta Andromeda 单一融合 GPU 流水线的系统架构。该流水线将索引匹配、融合打分和 Top-K 选择协同部署在单台 GPU 服务器上。融合打分内核在芯片内单次遍历中执行完整的每用户-广告计算——从 Embedding 提取到 MoE 交互和打分——将中间结果保留在寄存器和共享内存中以消除 HBM 瓶颈。这种协同设计通过仅发送紧凑的请求载荷将网络流量减少了约 100 倍——在对 Facebook 和 Instagram 上每个符合条件的广告进行评分的同时，突破了每秒 10 亿个广告的吞吐量。

GPU 加速检索：Meta 与 NVIDIA 的联合合作

随着生产流水线以每秒超过 10 亿个广告的速度运行，一个自然的问题随之而来：该系统离当前一代硬件的天花板还有多近？未来的模型增长还剩下多少空间？2026 年初，Meta 和 NVIDIA 展开联合合作来回答这个问题——系统地优化检索流水线的每个组件，将吞吐量推向硬件极限。NVIDIA 提供了优化后的 GPU 内核，Meta 负责集成并在完整流水线上进行基准测试。每个组件均实现了相比基线 2 倍至 5 倍的加速。

交互网络（Retrieval MoE 门控网络）

经过优化的门控内核性能提升了 5.3 倍，从每秒 24 亿提升至超过 100 亿个广告，达到了 GPU 峰值 HBM 带宽的约 85%。在这一吞吐量下，门控机制的执行时间 / 延迟与在相同硬件上使用简单点积打分内核一样快——而模型的表达能力却大幅提升。这就是协同设计的有效验证：在 Tensor Core 硬件上，额外的模型复杂度基本上是“免费”的。INT8 量化相比优化后的 FP16 内核进一步实现了高达 2 倍的吞吐量提升，且精度与 16 位精度相当。

索引匹配

通过针对性的 GPU 内核优化，索引匹配吞吐量提升了 2 倍以上，达到每秒超过 1000 亿个广告——比打分网络高出一个数量级。这种充裕的裕度意味着索引匹配可以在多条检索路径中重复运行而不会成为瓶颈。

端到端 GPU 流水线

除了单个内核之外，双方合作还对完整的检索流程进行了性能分析（profiling），并发现流水线组件之间的数据移动——而非计算本身——占据了端到端延迟的相当大一部分。通过跨索引匹配与打分协同设计数据布局以消除不必要的间接寻址，团队消除了这一开销。在应用所有优化后，完整流水线在单台 GPU 服务器上的基准测试表现达到了当前生产吞吐量的数倍，随着更多优化集成到生产环境中，预计还会获得进一步提升。

图 3：交互网络、索引匹配和端到端流水线优化后的每秒广告数（ads/sec）。

在下一代 NVIDIA 平台上的持续工作

基于 Hopper 架构取得的成果……

• This post describes the next generation of Meta Andromeda, Meta’s personalized ads retrieval engine, and how co-designing the model, system, and GPU hardware unlocks a new efficiency frontier for ads retrieval.
  • In production, the system scores every eligible ad with near ranking-level model performance at retrieval scale and surfaces the top candidates at over one billion ads per second, a new throughput benchmark for ads retrieval-stage models (i.e., a 100x lift compared to the 2024 Meta Andromeda system).
  • For advertisers, this means better performance automatically from the retrieval system matching ads across a larger candidate pool, within the same latency budget. Since 2025, Meta Andromeda has delivered a 2.4% increase in conversions on Facebook.

Meta Andromeda is Meta’s machine learning system for personalized ads retrieval, the first stage of the ads recommendation pipeline. Retrieval narrows a vast pool of eligible ads down to a small set of strong candidates in milliseconds; a later, more computationally expensive ranking stage then scores those survivors in fine detail.

Historically, the deep modeling quality of the ranking stage has been too costly to apply at retrieval’s scale. Building on the retrieval foundation established with Andromeda’s 2024 launch on the NVIDIA GH200 Grace Hopper Superchip, this post shares how co-designing the model, system, and hardware together brings near ranking-level performance to the retrieval scale.

Challenges and Opportunities in Scaling Retrieval

The 2024 Andromeda deployment reimplemented retrieval with an end-to-end optimized retrieval implementation from custom AI hardware with hierarchical indexing and GPU-native operators to a purpose-built ML model, growing model capacity by four orders of magnitude. Since then, generative AI creative tools have helped advertisers rapidly create hundreds of millions of new ads, growing Meta’s ads candidate pool to an unprecedented scale.

Scoring such a large ad candidate pool with a more accurate model at a much higher system throughput requires fundamental technology breakthroughs in model quality, system architecture, and hardware utilization jointly.

Higher Model Performance Requires a New Architecture

Ads retrieval has an outsized ROI on incremental quality: Small improvements in retrieval model performance can drive large gains in advertiser value. The rapid expansion of our ads candidate pool makes optimizing for hardware capacity ROI an increasingly critical factor. The 2024 Meta Andromeda retrieval models have established a highly effective foundation with their GPU-optimized deep neural network architecture. To keep improving retrieval performance at a sustainable compute cost at our current scale, however, we must move beyond that paradigm. This requires a fundamental architectural evolution to push model performance boundaries significantly without a proportional increase in costs.

New System Bottlenecks at Billion Scale

The 2024 Meta Andromeda system evaluates at tens of millions of ads per second per GPU card. Meeting the new business demand requires a throughput at multi-billion ads evaluated per second per card in the system. At billions of ads per second throughput, the memory bandwidth of passing ad tensors through memory hierarchy becomes the new bottleneck. Intermediate tensors between GPU kernels account for terabytes per second of memory traffic, and shipping ads between local CPU hosts and remote hardware accelerators evaluating models saturates the network link. This requires a new system co-design that unifies these fragmented, distributed workloads into a single GPU pipeline.

Pushing Towards the Hardware Limits

With retrieval running on GPU, the important optimization question is “How close are we to the hardware ceiling, and how much room remains for future model growth?” Answering this requires pushing each component of the retrieval pipeline to the limits of the current hardware generation, a challenge that motivated a joint collaboration with NVIDIA.

Each of these challenges drove a corresponding advance.

Unlocking Retrieval Performance Through Model-System Co-Design

Next-Gen Retrieval Model Architecture — Retrieval Mixture-of-Experts

The latest iteration of Meta Andromeda introduces three synergistic architectural innovations that bring large language model (LLM)-inspired mixture-of-experts (MoE) into the retrieval (see Figure 1). With this highly expressive yet inference-efficient architecture, a new retrieval performance-efficiency Pareto frontier has been established: model performance approaches ranking-level in a typical retrieval-then-ranking system, at approximately 300x fewer inference-time FLOPs than the 2024 Meta Andromeda retrieval model.

On the user modeling side , we transitioned from a single user tower to an ensemble of specialized user towers as experts. This captures diverse, time-evolving, and contextualized user interests by integrating a vast array of high-fidelity signals: sequential ad engagement events, user-cohort representations, LLM-derived semantic features, and cross-surface co-engagement signals.

In the candidate ads tower , we employ a joint encoding strategy that incorporates real-time serving context, ensuring that downstream interaction models receive context-aware representations.

In the interaction overarch layer , the interaction model evolved from dense neural networks into an MoE attention-based gating network that dynamically routes candidates across the most relevant user experts, delivering near-ranking-level performance at a fraction of the FLOPs.

A defining property of the new architecture is that its performance is co-designed with the underlying hardware execution model. Model inference throughput has reached over 10 billion ads per second on a single GH200 GPU, and this efficiency allows model capacity to grow further, through richer feature interactions, additional expert sub-networks, and deeper architectures, without a proportional increase in GPU cost. Each incremental model advancement translates into greater advertiser value.

Figure 1: Retrieval MoE architecture. Diverse high-fidelity user signals feed an ensemble of heterogeneous user experts. Ad content fused with real-time serving context forms a contextualized ad input that drives an ad-aware MoE router, activating only the top-K most relevant experts to produce the user–ad relevance score. The result is ranking-level matching quality at retrieval scale — ~300x fewer inference-time FLOPs than the 2024 Meta Andromeda and over 10 billion ads/sec on a single Grace Hopper Superchip.

Co-Designing the Model and System Onto a Single GPU Pipeline

The 2026 Meta Andromeda pushes the model-system co-design to the next level with two innovations — a single GPU kernel fusing both model and system workloads per user-ad interaction and a strategy to co-locate per user request logics with per user-ad GPU kernels in a single GPU box. To this end, a highly optimized and fully fused GPU pipeline is built to host the entire end-to-end Meta Andromeda workflow.

Fused Kernel

At a high level, Meta Andromeda’s inference time per user-ad workload is a compact sequence of operations across model execution and system logics — gathering embeddings, executing user-ad interaction models, and scoring. The ~300x FLOPs reduction in the retrieval MoE gating architecture creates the necessary headroom and is less resistant to fully fuse both model execution and non-model system workloads into a single GPU kernel, compared to the 2024 launch where only operations within the model execution were fused. Consequently, the fused kernel now runs at over 10 billion ads per second on a single GH200.

This matters because retrieval workloads are fundamentally memory-bandwidth bound. At billions of ads per second, even a small intermediate tensor per ad multiplies into terabytes per second of high bandwidth memory (HBM) traffic. The memory workload of these intermediate tensors, not arithmetic, becomes the dominant bottleneck in multi-kernel retrieval pipelines.

The fused kernel eliminates this bottleneck entirely. Rather than dispatching multiple small GPU kernels for model and system operations, it executes the entire per user-ad computation in one pass. Intermediate results stay in GPU registers and thread block shared memory and are never materialized to HBM, achieving high utilization of the GPU’s peak memory bandwidth rather than wasting it on redundant intermediate traffic.

Co-Location Strategy

Measured in isolation, the per user-ad fused kernel runs at multi-billion ads per second on a single GPU. However, per-user request operations such as the index matching (i.e., enforcing advertisers’ targeting rules) and top-K selection run on separate local CPU hosts and communicate with the fused GPU kernel scoring service on remote accelerators through the network connection. At multi-billion ads per second throughput, this traffic alone saturates the network link, making the network bandwidth become the bottleneck rather than GPU compute.

The co-design response was to collapse the system architecture to match the model architecture: co-locate index matching, scoring, and top-K selection on the same GPU. Instead of shipping large candidate lists across hosts, the system sends a compact request payload — user features and a targeting query — to a single GPU server, which runs the entire retrieval pipeline locally and returns only the final top-K results. This reduces network traffic by ~100x and shifts the system back to a GPU-compute-bound regime.

Moving index matching onto the GPU is not straightforward. Ads index matching evaluates complex advertiser-defined targeting logic across tens of feature categories, and must support real-time index updates as ads are created, modified, or expired throughout the day. Supporting this on GPU required redesigning the index structure for GPU-parallel execution while preserving the full fidelity of production targeting semantics.

A Single GPU Pipeline for Meta Andromeda

Combining the aforementioned two innovations, the fully co-located single GPU pipeline processes over one billion ads per second on a single GPU server while scoring every eligible ad in the candidate pool. This is a new throughput benchmark for ads retrieval-stage models, and as the next section describes, significant headroom remains.

Figure 2: System architecture of the single fused GPU pipeline for Meta Andromeda. This pipeline co-locates index matching, fused scoring, and top-K selection on a single GPU server. The fused scoring kernel executes the entire per-user-ad computation — from embedding gathering to MoE interaction and scoring — in a single on-chip pass, keeping intermediate results in registers and shared memory to eliminate HBM bottlenecks. This co-design reduces network traffic by approximately 100x by sending only compact request payloads — surpassing 1 billion ads per second while scoring every eligible ad across Facebook and Instagram.

GPU-Accelerated Retrieval: The Meta-NVIDIA Collaboration

With the production pipeline running at over one billion ads per second, a natural question follows: How close is the system to the ceiling of the current generation hardware, and how much headroom remains for future model growth? In early 2026, Meta and NVIDIA launched a joint collaboration to answer this question — systematically optimizing each component of the retrieval pipeline to push throughput toward hardware limits. NVIDIA contributed optimized GPU kernels, and Meta integrated and benchmarked the full pipeline. Each component achieved 2x to 5x speedup over the baseline.

Interaction Network (Retrieval MoE Gating Network)

The optimized gating kernel improved by 5.3x, from 2.4 billion to over 10 billion ads per second, reaching ~85% of the GPU’s peak HBM bandwidth. At this throughput, the execution time / latency of the gating mechanism is just as fast as what a simple dot-product scoring kernel achieves on the same hardware — while the model is substantially more expressive. This is the co-design validation: the additional model complexity is essentially free on Tensor Core hardware. INT8 quantization achieved a further up to 2x throughput increase over the optimized FP16 kernels, with accuracy on par with 16-bit precision.

Index Matching

Through targeted GPU kernel optimizations, index matching throughput improved over 2x, reaching over 100 billion ads per second — an order of magnitude higher than the scoring network. This headroom means index matching can be re-run across multiple retrieval pathways without becoming a bottleneck.

End-to-End GPU Pipeline

Beyond individual kernels, the collaboration profiled the full retrieval flow and identified that data movement between pipeline components — not computation — accounted for a significant fraction of end-to-end latency. By co-designing data layouts across index matching and scoring to eliminate unnecessary indirection, the team removed this overhead. With all optimizations applied, the full pipeline benchmarks at multiple times the current production throughput on a single GPU server, with further gains expected as additional optimizations integrate into production.

Figure 3: Optimized ads/sec for interaction network, index matching, and E2E pipeline.

Continuing Work on Next-Generation NVIDIA Platforms

Building on the Hopper results above, we are extending our optimization and co-design work to Grace Blackwell. Here, we are focusing on platform design optimizations that enable different CPU-to-GPU compute ratios, which our work so far suggests is an important factor for Andromeda-like workloads. We are evaluating how sensitive end-to-end application performance is to these ratios, using Grace-class CPU performance and Blackwell-class GPU performance for their respective parts of the workload.

More broadly, we are exploring how these findings extend to the next platform generation, where greater flexibility in CPU+GPU execution, resource ratios, and coupling could bring more of these architectural improvements to bear on Andromeda-like workloads.

These results confirm that the co-designed retrieval pipeline is not yet at its ceiling. With quantization unlocking further scoring throughput and index matching operating at an order of magnitude above the scoring network, the architecture is positioned to absorb significantly more model complexity — deeper interaction networks, richer feature spaces, and additional retrieval pathways — without proportional infrastructure growth.

The Path Forward for Ads Retrieval

The advances described here follow a single principle: The largest gains come from designing model, system, and hardware as one integrated stack, rather than optimizing each in isolation. The retrieval MoE gating architecture delivers near-ranking-level expressiveness at no additional serving cost. Fusing the entire retrieval pipeline onto a single GPU eliminates network and memory bottlenecks. And pushing the pipeline towards its hardware ceiling confirms substantial headroom remains.

Looking ahead, the co-design principle that drove these results applies equally to next-generation model architectures and hardware platforms, with the demonstrated headroom enabling continued scaling without proportional infrastructure growth.

For people who use Meta’s products, this translates directly into better ads that are more relevant to them, delivered within the same latency budget. For advertisers, it means better customer matches and more conversions. For Meta, it means increased business from advertisers and lower costs, as each GPU now evaluates more ads per second. As the volume of eligible ads continues to grow, driven by Advantage+ automation and generative AI creative tools, Meta Andromeda’s co-designed architecture scales to meet it.

Acknowledgements

We would like to thank Gosh Arzumanyan, Habiya Beg, Andrew Bocking, Zain Brohi, Rohit Chandel, Claire Chen, Jason Chen, Wenlin Chen, Zhenyu Chen, Yuze Dai, Zhou Fang, Junjie Feng, Stephen Fitz, Chunli Fu, Yicong Fu, Xiaorui Gan, Sohan Gantala, Yingsheng Gao, Golnaz Ghasemiesfeh, Ishu Goyal, Wentian Guo, George Han, Yihan He, Jun Hu, Xuan Hu, Kun Huang,Qin Huang , Tianchen Ji, Crystal Jin, Tao Ju, Rajat Karnwal, Santanu Kolay, Michael Kwan, Ilia Lewis, Baiyu Li, Chao Li, Jim Li, Li Li, Lin Li, Liyuan Li, Mingda Li, Ted Li, Xiaotian Li, Yanyu Liang, Yue Liang, Irene Liu, Gaoxiang Liu, Jason Liu, Siqi Liu, Wei Liu, Xiao Liu, Yuheng Long, Yining Lu, Siyi Luo, Anuj Madan, Phalguna Mankar, Jiachen Mao, Yuji Mo, GP Musumeci, Guruswamy Namasivayam, Kexin Nie, Sandeep Pandey, Uladzimir Pashkevich, Clementine Plati, Heng Qiao, Hang Qu, Lei Qu, Anand Rajan, Jatin Sethi, Puneet Sharma, Ruchit Sharma, Yan Shi, Chang Shu, Sean Song, Alec Stashevsky, Matt Steiner, Zheng Sun, Nikhil Vinayak Supekar, Jim Tao, Chintan Thakkar, Tian Tong, Aleksandr Ulanov, Daksh Vasistha, Simha Venkataramaiah, Abhishek Verma, Ruud Vinke, Vidhoon Viswanathan, Sasha Vorontsov, Eric Walden, Dong Wang, Feng Wang, Haole Wang, Jeff Wang, Senhong Wang, Ye Wang, Di Wu, Haocheng Wu, Shawn Wu, Jimmy Xiang, Nan Xiao, Ruichao Xiao, Chao Xie, Yudi Xie, Fangzhou Xu, Nathan Yan, Tak Yan, Hang Yang, Song Yang, Yuhan Yang, Yavuz Yetim, Bowen Yu, Li Yu, Qifan Yuan, Xiaodong Yuan, Lin Zeng, Siqi (Cindy) Zeng, Caijie Zhang, Joanna Zhang, Min Zhang, Qing Zhang, Zhengkai Zhang, Fangyu Zou, and the entire v-team behind the success of Meta Andromeda.