Realize more flexible generation AI utilization
*Zero Field Co., Ltd.*
Press release: September 4, 2024
**
Realize more flexible generation AI utilization
*Zero Field starts offering “GPU Server for AI”*
Wednesday, September 4, 2024, Zero Field Co., Ltd. (Headquarters: Minato-ku, Tokyo, Representative Director and CEO: Hirashima) Yosuke (hereinafter referred to as Zero Field) is a GPU server specialized for IT vendors and AI developers who develop generative AI.
We are pleased to announce that we have started providing AI. * Securing computing resources has become a major topic *
The US semiconductor giant NVIDIA announced its financial results for the May-July period of 2024 on August 28th, with sales of $30.04 billion (approximately 4.35 trillion yen), approximately 2.2 times the same period last year. Net income increased 2.7 times to $16,599 million, both exceeding market expectations. This shows the high demand for semiconductors for AI, and it is expected that securing computing resources will become a major topic for companies in the future.
*What AI solutions companies need*
As companies increasingly introduce AI, the use of generative AI is expected to become a mainstay for improving operational efficiency and creating new business. Recently, there has been a growing need for on-premises GPU servers that do not rely on major cloud services, but it is also true that there are some hurdles to implementing GPU servers in-house. There are several issues such as high operational costs, inability to ensure security, lack of server operation know-how, lack of human resources, and uncertainty about the amount of resources required. To take advantage of generative AI, you need a flexible GPU server that fits your business and is truly
developer-friendly.
* “GPU Server for AI” to accelerate AI development *
“GPU Server for
“AI” is a service that provides customers with the optimal server based on both software and hardware. Based on the extensive experience Zero Field has cultivated over the years in GPU sales/management and data center operations, we provide GPU servers that meet all needs. We design and design the optimal GPU server according to the use case, such as AI applications such as generative AI, large-scale language models (LLM), and image analysis, advanced simulation applications such as 3D CAD and fluid analysis, and visualization applications such as 3D and metaverse. We will build it.
* Three features of GPU Server for AI *
* ・Proposals from an AI developer’s perspective *
In cooperation with Triple Eyes, a group company with strengths in AI development, we will propose the optimal server for running AI. We can also provide PoC development support by AI engineers for AI
utilization. By utilizing Triple Eyes’ AI implementation support service “AI Lab”, it is possible to construct an AI server that is even closer to the customer.
* ・GPU utilization and engineering*
Zero Field has a cumulative track record of operating over 15,000 GPUs. Because we have know-how on various uses and operations of GPUs, we are able to make proposals from a position close to the users. Customization is possible taking into account the compatibility of the housing and various parts and ease of maintenance.
* ・Data center and server operation results*
Because Zero Field operates data centers and manages GPU servers in-house, we have knowledge about the operating environment of GPU servers, such as cooling, electricity, networks, and security. We can also discuss the operating environment and installation work.
*Example of assumed usage of GPU Server for AI*
We design and build GPU servers suitable for a wide range of uses such as HPC, AI/deep learning, design simulation, visualization, and media/entertainment.
・Generative AI (image generation, video generation)
・Image analysis such as OCR and annotation
・Large-scale language model construction (LLM) such as GPT
・Increasing the resolution of videos and images (Super resolution and noise removal)
・3D, 4D CAD simulation
・Encoding processing of high-resolution video data, etc.
* Assumed hardware for GPU Server for AI *
* ・GPU server for generated AI for businesses (operated at data center) * [GIGABYTE GPU server equipped with NVIDIA100]
The GPU server for generating AI for businesses is intended for use in data centers. We support the installation of Zero Field into container-type data centers and into data centers that customers are already using.
* ・Workstation for AI developers*
[AMD Threadripper Pro. 5955WX 64GB memory installed]
The AI Developer Workstation is a workstation starter kit that can accommodate up to 4 GPUs in a 4U chassis.
* About Zero Field ─ No.1 sales and operation record of GPU machines * Since 2017, we have developed and operated over 3,500 GPU machines in a variety of environments, and as of 2024, we boast the No. 1* operating track record in Japan.
(* ** Tokyo Shoko Research Survey implementation period: November 24, 2023 to February 28, 2024)
Additionally, we have opened and operate five data centers in Japan and two in Washington State, USA. In September 2023, we will become a group company of AI system developer Triple Eyes, and aim to develop our data center and computing businesses by leveraging our past experience in design, construction, and operation.
* [GPU Server for AI inquiries] *
Zerofield Co., Ltd. GPU Server Division
Phone: 03-6424-7616 (10:00-19:00, excluding Saturdays, Sundays, holidays, and specific days)
Mail：info@zerofield.biz
*Zero Field Company Profile*
[Company name] ZEROFIELD, Inc.
[Location] 32F, Grand Park Tower, 3-4-1 Shibaura, Minato-ku, Tokyo [Established] April 2017
[Representative] Representative Director and CEO Yosuke Hirajima [Business content] AI/big data related system development and operation business/Multi-purpose high performance PC sales/
Operation business
[Corporate site] https://zerofield.biz/
Zero Field Co., Ltd. is Triple Eyes Co., Ltd. (Headquarters: Chiyoda-ku, Tokyo, Representative Director: Yamada)
Yuichiro is a group company of the Tokyo Stock Exchange Growth Market (stock code: 5026).