将人工智能概念变成现实＃2121

加利福尼亚州圣何塞，2025年3月18日 /美通社 / - Adlink Technology Inc.，Edge Computing的全球领导者和NVIDIA Elite

SAN JOSE, Calif., March 18, 2025 /PRNewswire/ -- ADLINK Technology Inc. (TAIEX: 6166), a global leader in edge computing and an NVIDIA Elite Partner, will be showcasing its Edge AI, generative AI, and robotics solutions at the upcoming 2025 NVIDIA GTC conference, taking place in San Jose, California, from March 18-21, 2025. Visit ADLINK at booth #2121.

2025年3月18日，加利福尼亚州圣何塞 / - adlink Technology Inc.（Taiex：6166），Edge Computing的全球领导者和NVIDIA Elite合作伙伴将展示其边缘AI，生成AI，生成AI，并在2025年NVIDIA GTC COMPANTINC中的2025年CARCAMENIA，CARACNIA，CARACTARIA，CARACATINIA，CARACATION，CARACTARIA，CARACATION，CARACTARIA，CARACATION，CARACTARIA，CARACARCTIA，加利福尼亚州访问。在2121号展位。

Generative AI adoption at edge devices is often limited by DRAM capacity constraints when running large language models (LLM). However, ADLINK's new DLAP Supreme series, featuring NVIDIA® Jetson™ AGX Orin™, overcomes this challenge with Phison aiDAPTIV+ technology. This unique combination elevates AI power to new heights.

运行大型语言模型（LLM）时，Edge设备上的生成​​AI采用通常受到DRAM容量限制的限制。但是，ADLink的新DLAP Supreme系列以NVIDIA®JETSON™AGX ORIN™为特色，克服了Phison Aidaptiv+ Technology的这一挑战。这种独特的组合将AI功率提升到新的高度。

For instance, a standard configuration of the DLAP-411 Orin Supreme, equipped with an NVIDIA Jetson AGX Orin 64GB configuration and the Llama3.1-70B model, achieves an inference throughput of 70000 tokens per second with a maximum token length of 8192. In comparison, a typical edge server, like the AEA-S280, running the same model, achieves an average throughput of 134 tokens per second with a maximum token length of 2048. This translates to a 52 times improvement in inference speed and a 4 times increase in token length when using the DLAP Supreme series.

例如，配备了NVIDIA JETSON AGX ORIN 64GB配置和Llama3.1-70B模型的DLAP-411 ORIN最高配置的标准配置，可实现每秒70000个令牌的推入吞吐量，最大值的代币长度为8192。在使用DLAP Supreme Seriess时，最大令牌长度为2048的134个令牌，最大令牌长度为2048。

Moreover, the new series is optimized for specific GenAI models, like Stable Diffusion and BAE models, further increasing the AI TOPS by up to 1.7 times and improving the GenAI model performance by up to 2 times compared to the previous generation of products.

此外，新系列针对特定的Genai模型进行了优化，例如稳定的扩散和BAE模型，将AI顶部提高了1.7倍，而与上一代产品相比，Genai模型的性能最多可提高2倍。

On the other hand, for enterprises seeking scalable on-premise workstations, ADLINK's AXE AI Acceleration Server, certified by NVIDIA, is equipped with NVIDIA RTX 6000 and meets the varied AI performance requirements ranging from 100 to 6000 TOPS. As an NVIDIA-qualified system, it ensures hardware stability, scalability, and fast deployment for easy AI adoption.

另一方面，对于寻求可扩展本地工作站的企业，由NVIDIA认证的Adlink的AX AX AI Acceleration Server配备了NVIDIA RTX 6000，并满足了不同的AI性能要求，范围从100到6000台。作为NVIDIA合格的系统，它可确保硬件稳定性，可扩展性和快速部署，以轻松采用AI。

"We are excited to showcase our latest innovations in edge AI and robotics at NVIDIA GTC," said Rony Lee, Senior Division VP at ADLINK. "Our solutions are designed to help enterprises of all sizes deploy AI applications at the edge quickly and easily."

Adlink高级副总裁Rony Lee说：“我们很高兴在NVIDIA GTC展示我们在Edge AI和机器人技术方面的最新创新。” “我们的解决方案旨在帮助各种规模的企业快速轻松地在边缘部署AI应用程序。”

ADLINK's comprehensive Edge AI portfolio is ideal for diverse industries. The DLAP-211 Dev Kit, built on NVIDIA Jetson Orin NX or Orin Nano™ modules, is a robust AI computing platform for proof-of-concept developments in robotics, manufacturing, and retail.

Adlink的综合边缘AI投资组合非常适合多元化行业。 DLAP-211开发套件建于NVIDIA JETSON ORIN NX或ORIN NANO™模块，是一个可靠的AI计算平台，用于机器人技术，制造业和零售业的概念证明开发。

The MLB-IGX Medical Box PC, powered by the NVIDIA IGX platform with NVIDIA RTX 6000, enables low-latency AI inferencing for real-time surgical healthcare scenarios. For AI vision at the edge, the NEON AI Smart Camera, featuring NVIDIA Jetson Orin Nano, simplifies Edge AI vision in automation and manufacturing industries.

由NVIDIA IGX平台提供NVIDIA RTX 6000的MLB-IGX Medical Box PC，可为实时手术医疗保健方案推导低延迟AI。对于Edge的AI Vision，NEON AI智能摄像机（NVIDIA JETSON ORIN NANO）简化了自动化和制造行业的Edge AI视觉。

To unleash professional graphics on the go, the portable AI GPU, Pocket AI, equipped with NVIDIA RTX A500, improves work efficiency for AI developers and embedded industries.

为了释放旅途中的专业图形，配备了NVIDIA RTX A500的便携式AI GPU，Pocket AI提高了AI开发人员和嵌入式行业的工作效率。

To learn more about ADLINK's NVIDIA-based solutions, visit the ADLINK website or join us at the upcoming NVIDIA GTC conference in San Jose, California.

要了解有关ADLink基于NVIDIA的解决方案的更多信息，请访问Adlink网站，或加入我们即将在加利福尼亚州圣何塞举行的NVIDIA GTC会议上。

About ADLINK Technology

关于AdLink技术

ADLINK Technology Inc. (TAIEX: 6166) is a global leader in edge computing. Since 1995, ADLINK has been a trusted edge hardware and software provider for over 1,600 customers in 80+ countries. The company's products and services are used in a wide range of industries, including manufacturing, telecommunications, healthcare, energy, defense, transportation, and infotainment.

Adlink Technology Inc.（TAIEX：6166）是边缘计算的全球领导者。自1995年以来，Adlink一直是80多个国家 /地区的1,600多个客户的值得信赖的边缘硬件和软件提供商。该公司的产品和服务用于广泛的行业，包括制造，电信，医疗保健，能源，国防，运输和信息娱乐。

ADLINK is also a member of several industry consortia and standards bodies, where it contributes to the development of new technologies and standards. The company is committed to delivering innovative solutions that meet the needs of its customers and help to advance the industry as a whole.

Adlink还是几个行业财团和标准机构的成员，在那里它有助于开发新技术和标准。该公司致力于提供创新的解决方案，以满足客户的需求，并帮助整个行业推动。