Background
As cyberattacks become more sophisticated, AI’s role in defense is expected to be more crucial. Due to fast-evolving cyber threats and data surge, traditional methods are hard to keep pace with them. The integration of AI and machine learning technologies enable advanced cybersecurity solutions with excellent efficiency and enhanced defensing power to deal with the dynamic environment.

Features that AI Brings into Cybersecurity
AI enhances cybersecurity by providing advanced threat detection, automating responses, predictive capability, and evolvable adaptability against cyber-attacks. Features are listed below to show why AI is indispensable to cybersecurity.

• Real-time Detection
  Compared to traditional rule-based security systems, AI offers greater accuracy and efficiency. AI can scan vast numbers of devices for vulnerabilities and analyze the behaviors of user and system to establish a baseline, making it easier to identify unusual or malicious activities that traditional approaches might miss.
• Intelligent Analysis & Prediction
  AI-driven analytics can process and analyze large datasets to identify trends, vulnerabilities, and potential risks. By analyzing historical data and user behavior, it is capable of predicting potential attacks and provide insights which allow organization to proactively take preventive measures before an attack occurs.
• Automated Response
  AI-powered systems are capable of automating the detection and quickly response to security threats such as isolating infected devices and initiating remediation procedures which help to free up security professionals to focus on more complex tasks. Real-time reaction and better resource utilization can be achieved.
• Evolvable Adaptability
  Training machine learning algorithms continuously can minimize false positives by refining their understanding of normal system behavior. Furthermore, learning from historical data and new data can also improve the accuracy of detecting evolving cyber threats.

Performant Platforms for AI-powered Cybersecurity Solutions
The Edge AI server is one of the key elements in AI-powered cybersecurity solutions. To accomplish real-time processing massive amount of data, rapid detection of cyber threats, and automatic reaction to minimize the risk, the features that are demanded to the edge AI servers are listed below.

• High Processing Power
  AI algorithms require significant processing power, especially for tasks like real-time threat analysis and complex model training. High-core-count CPUs and potentially GPUs or accelerators are demanded for computing-intensive tasks.
• High Bandwidth and Memory/Storage Density
  High bandwidth and memory capacity are necessary for high-speed data processing. Sufficient storage to house training data, historical security logs, and potentially threat intelligence information are also highly required.
• Scalability
  The requirements of the system of cybersecurity increases with the fast-evolving cyber threats. Server with options that allow users to easily scale up according to the demands provides convenience and flexibility for adapting dynamic environments.
• Security
  Security features such as encryption and decryption that help to protect the sensitive data it stores and processes are essential to cybersecurity. The reliability of the hardware itself is also important to maintain the system with reduced OPEX.

In addition to high performance edge computing servers, ultra-compact devices are suitable for real-time inference at the edge. From edge to cloud, computing systems with high performance per watt and optimized resource efficiency apply AI to optimize the efficiency and accuracy of cybersecurity.

Summary
With AI, cybersecurity has been revolutionized. AI-powered cybersecurity solutions provide effective and adaptive defenses against increasingly sophisticated cyber threats to minimize security risk with low OPEX.

這篇文章 Deploying AI to Optimize Cybersecurity 最早出現於 AEWIN。

Why is Intelligent Storage required?
Explosive growth of data leads to complex challenges requiring high-performance storage solution to access, process, and manage the data fast, securely, and cost effectively. Intelligent storage is a system or solution that leverages artificial intelligence (AI) to optimize performance for real-time management and proactive response.

To maximize the value of the data and to support innovative applications, systems featuring secured data access, efficient management, excellent reliability, and predictive management are highly demanded. Intelligent storage features optimized performance, excellent scalability, enhanced security, and the ability to analyze, predict, and react in real-time.

Features of Intelligent Storage

• Performance Optimization
  Intelligent Storage enables smart caching to accelerate data access, automated data tiering to classify data with its pattern/usage with low latency, and data deduplication & dynamic provisioning to optimize storage utilization for better efficiency and TCO. It is beneficial to performance-sensitive applications.

• Scalability and Flexibility
  With storage virtualization techniques and dynamic provisioning, it enables better management and allocation of multiple storage sources to adapt to the fast-evolving environment. Furthermore, cloud integration is featured which allows it to extend the storage capabilities beyond on-premises infrastructure.

• Enhanced Security
  Implemented encryption and access control protect stored data from unauthorized activity which are essential to handle sensitive data. Automated data back-up ensures data integrity, and data recovering further strengthen the data protection to minimize data loss.

• Advanced Analytical and Predictive Capability
  Incorporate analytics and machine learning capabilities, intelligent search can be achieved to increase overall efficiency. Moreover, it can analyze historical data access patterns to predict further storage needs and generate future plan with proactive reaction.

Applications that Intelligent Storage is crucial to

• Generative AI
  Generative AI and other large language model applications address high demands of large-scale data processing and real-time analytics which intelligent storage can efficiently handle. Powerful computing, high throughput, and great scalability of intelligent storage are also indispensable to efficient access and model training.

• Cybersecurity
  As security attacks becomes more complex and diverse, evolving training model for intelligent detection can help to strengthen the overall security. Intelligent storage ensures automatic threat detection, identification, and response in real-time with advanced algorithms which make them critical to AI-powered cybersecurity.

• Smart City
  Diverse data sets are generated by sensors, cameras, and other IoT devices every day. The great scalability of intelligent storage is perfect for accommodating the tremendous amount of data. Intelligent analysis and prediction make more value of the video data which are beneficial for the city planner to come up with appropriate plan for better city development.

• Scientific Research
  Climate modeling, medicine development, and further scientific simulations all involve massive datasets. Intelligent storage could help with quick data access, processing, and retrieval with secure. Moreover, it enables real-time analytics which allow researchers to gain more insights and make smart decisions accordingly.

AEWIN platform as intelligent storage with the best TCO
AEWIN provides several general-purpose servers featuring extraordinary computing power, high throughput, great scalability, and enhanced security. Multiple models are offered to help customer tailor their configuration to meet specific workload required with great flexibility and the best TCO. Some of AEWIN general-purpose models are listed below for quick comparison. Further details can be found in the bottom of this article.

Conclusion
Featured intelligent storage plays a crucial role in enabling various AI applications that require rapid data access, real-time data analytics, and advanced data security. AEWIN offers several general-purpose servers with extraordinary computing power, high throughput, great scalability, high storage density, extraordinary reliability, and the best TCO for customers to build up tailored solutions effortlessly.

Check out further AEWIN intelligent storage by clicking below.
BIS-5121-1U: Single 3rd Gen Intel® Xeon® Scalable Processor (Ice Lake-SP/ Whitley platform) Server, supporting 8x DDR4, 4x 3.5” Hot-Swap HDD, 2x PCIe Gen4
BIS-5121-2U: Single 3rd Gen Intel® Xeon® Scalable Processor (Ice Lake-SP/ Whitley platform) Server, supporting 8x DDR4, 12x 3.5” Hot-Swap HDD, 3x PCIe Gen4
BIS-5222-2U: Dual 3rd Gen Intel® Xeon® Scalable Processor (Ice Lake-SP/ Whitley platform) Server, supporting 16x DDR4, 12x 3.5” Hot-Swap HDD, 5x PCIe Gen4
BIS-5222-4U: Dual 3rd Gen Intel® Xeon® Scalable Processor (Ice Lake-SP/ Whitley platform) Storage Server, supporting 16x DDR4, 24x 3.5” Hot-Swap HDD, 5x PCIe Gen4
BAS-6101B: AMD Zen4/Zen4c EPYC 9004 Processor (Genoa/Bergamo) Server, supporting 24x DDR5, 12x 3.5” SATA slots, 8x PCIe slots (6x Gen5, 2x Gen4), 1x OCP3.0 SFF

這篇文章 Introduction of Intelligent Storage 最早出現於 AEWIN。

Background
The amount of data surges unprecedented by increased connectivity and the development of digitization of various aspects of daily life. Higher throughput platforms are required to handle and process the exponential growth of digital data not only in data transfer, but also in storage and analysis. Real-time communication and high-performance computing are major to effectively execute various kinds of applications. As there are increasing sophistication of cyber threats, Network appliances/storage server/edge AI servers with higher throughput are crucial to efficiently inspect and filter network traffic for potential threat and attack to maintain the integrity and security of the network.

High Throughput Network Appliances with PCIe Gen5 NIC
AEWIN offers high performance network appliances/storage server/edge AI server with flexible PCIe Gen5 expansion slots for excellent scalability. It is ready for NVIDIA Mellanox ConnectX-7 PCIe Gen5 NIC to enable high throughput with low TCO. Here we introduce a pre-tested solution of BAS-6101 with NVIDIA ConnectX-7 200GbE NIC.

• BAS-6101

BAS-6101A is a 2U edge server. It supports the most advanced AMD Bergamo/Genoa processor, 24x DDR5 slots, 6x PCIe Gen5 x16 slots (4x FHFL, 2x FHHL), 2x HHHL PCIe Gen4 x8 slots, 6x SATA/NVMe SSD, and 1x OCP 3.0 SFF Gen5 slot. It also offers IPMI function to simplify local/remote management. AEWIN develops BAS-6101A as general-purpose server or edge AI server for diverse applications. Different add-on cards (e.g. GPU/AI card, QAT card, encryption card, smart NIC such as ConnectX-7 NIC or accelerator card) can be installed with ease to satisfy various kinds of requirements.

• NVIDIA Mellanox ConnectX-7 NIC

ConnectX-7 NIC series are the PCIe Gen5 ethernet cards provided by NVIDIA Mellanox. There are multiple options including 25GbE, 50GbE, 100GbE, 200GbE, and 400GbE. By installing selected NVIDIA ConnectX-7 NICs in AEWIN platforms, tailored solution with the best TCO can be deployed. ConnectX-7 family provides hardware-accelerated networking, storage, security, and manageability services with features in including RoCE & SR-IOV for networking, NVMe-of over TCP/RoCE for storage accelerations, and IPsec/TLS/MACsec encryption/decryption & root-of-trust for cybersecurity.

• High Throughput Edge Server Solution: BAS-6101+ NVIDIA ConnectX-7 NIC

AEWIN installed and tested NVIDIA Mellanox ConnectX-7 200GbE NIC (Dual QSFP112) in BAS-6101 and the test results will be shared in white paper for your reference. With multiple NVIDIA ConnectX-7 NIC installed, the system is capable of providing smooth transmission and reliable performance for bandwidth-intensive applications of Networking and Storage. With additional GPU cards installed, AI powered cybersecurity solution can efficiently identify, detect, and react to cyber-attacks with the ability to learn automatically and further analyze the data for better response.

Summary
Tremendous amount of data has been generated in an extremely high speed and requires real-time process to enable more innovative applications. Performant systems with high-speed transmission and affordable cost are highly essential. AEWIN network appliances/storage server/edge AI servers with PCIe Gen5 NIC such as NVIDIA Mellanox ConnectX-7 NIC features high throughput and excellent performance with the best TCO which make them perfect for AI powered Cybersecurity, Intelligent Storage, Networking, and more.

這篇文章 High Throughput Solution with PCIe Gen5 NIC 最早出現於 AEWIN。

Background
Exponential growth of the data has been a consistent trend. Diverse applications including AIoT, intellectual storage, AI powered cybersecurity, generative AI have been developed rapidly along with it. It is crucial to transmit, store, analysis, and make use of them in a smarter way. Innovative technologies, Intel Accelerator Engines that will be introduced in this Tech Blog, help to accomplish complex tasks of data center, cloud, and edge with better TCO.

Introduction of Intel Accelerator Engines
Intel Xeon Scalable processors are designed for data-intensive use cases and there are Intel Accelerator Engines embedded for higher performance, higher efficiency, strengthened data security, lower cost, and lower energy consumption. Here we would introduce the Intel Accelerator Engine that is beneficial to Network, Security, Storage, AI, High Performance Computing (HPC), and Analytics.

• Intel QuickAssist Technology, Intel QAT
  The latest 4th and 5th Gen Intel Xeon processors are with Intel QAT which can offload compute-intensive workloads of cryptography and data compression/decompression operations from the CPU. It helps to reduce CPU utilization and significantly enahnce the performance of networking, storage, security, HPC, and analytics.
• Intel Data Streaming Accelerator, Intel DSA
  Intel DSA delivers high performance for networking, storage, analytics and data-intensive workloads by optimizing transmission and speeding up streaming data movement. It offloads compute-intensive operations from the CPU and lower the overall latency with reduced cost.
• Intel Dynamic Load Balancer, Intel DLB
  Apart from Intel QAT and Intel DSA, Intel DLB is also one of the Intel Network Engines to help move data faster. Intel DLB balances and distributes network traffic across CPU cores and optimize the performance of data streaming, packet processing and compute-intensive networking tasks.
• Intel Volume Management Device (Intel VMD)
  Intel VMD is another Intel storage engines that can unleash the performance of the NVMe storage solution. It supports direct control and management of NVMe SSDs from the PCIe bus. Intel VMD Direct Assign can reduce storage latency and increase bandwidth as it allows an Intel VMD Domain to be directly assigned to a VM.
• Intel In-Memory Analytics Accelerator, Intel IAA
  Intel IAA speeds up big data analytics by scanning and filtering large datasets within queries to offload work from cores. It accelerates compression and decompression for faster data transfers and analytics workload processing with better energy efficiency and lower expense.
• Intel AMX & Intel AVX-512
  Regarding AI/HPC acceleration, there are Intel Advanced Matrix Extensions (Intel AMX) which improves deep learning training/inference on the CPU and a set of new instructions called Intel Advanced Vector Extensions (Intel AVX-512) that can boost performance for machine learning workloads.
• Further Intel Security Engines: Intel SGX, Intel TDX, Intel PFR
  In addition to QAT, there are further security engines for enhanced protection and strengthened defenses. Intel Software Guard Extensions (Intel SGX) isolates sensitive data and protect them from attack while Intel Trust Domain Extensions (Intel TDX) enable a trusted domain in a virtual machine (VM). Intel Platform Firmware Resilience (Intel PFR) ensure firmware integrity against disruption.

Summary
Intel Accelerator Engines are beneficial to address the growing demand for high-performance and energy-efficient solutions from data centers to edge devices. They enhance overall system performance, efficiency, and the ability to meet the critical requirements of modern computing workloads.

這篇文章 Introduction to Intel Accelerator Engines 最早出現於 AEWIN。

In a server/embedded platform environment, storage devices play a critical role in managing and providing access to vast amounts of data that servers need to store and retrieve. The selection of storage devices in a server setup is crucial for achieving optimal performance, reliability, and scalability. Here’s an introduction to common storage options in current standard AEWIN platforms.

Table1. Comparison of storage interfaces

There are three main interfaces/protocol of storages in AEWIN Server.

SATA, Serial ATA
Serial AT Attachment, SATA, is a computer bus interface that connects host bus adapters to mass storage devices. It is a widely used interface for connecting hard disk drives (HDDs) and solid-state drives (SSDs) to servers. It is a cost-effective solution suitable for most applications with moderate performance requirements. SATA interfaces are commonly found in servers and storage systems.

SAS, Serial Attached SCSI
Serial Attached SCSI, SAS, is a point-to-point serial protocol that transfer data to and from computer-storage devices. SAS error-recovery and error-reporting uses SCSI commands, which have more functionality than the ATA commands used by SATA drives. SAS is commonly used in mission-critical applications where speed and reliability are paramount.

NVMe, Non-Volatile Memory Express
Non-Volatile Memory Express or NVM Express, NVMe, is a storage interface specifically designed for modern, high-performance solid-state drives (SSDs). It is a protocol that leverages PCIe to provide direct communication between the storage device and the server’s CPU, reducing latency and improving overall performance.

Various kinds of form factors for each storage interfaces have been listed in table1 listed above. Each form factor has its own advantages and is suitable for specific use cases. For example, HDD is applied to achieve high capacity with the best TOC. As for mSATA and M.2, they are suitable for compact systems (further details of the dimension can be found in table2 below. U.2, with its compatibility with 2.5-inch bays, is often used in enterprise environments where high-performance, hot-swappable storage is essential.

Table2. Reference dimensions of storage devices with different form factors

Summary
As a Network Appliance provider, AEWIN designs platforms with great flexibility for installing a variety of the storage devices per customer’s requirements. Selected based on the features including speed/scalability/cost/dimension mentioned in this article, the most suitable storage options can be installed in AEWIN system for the best TCO.

這篇文章 Introduction of storage options for AEWIN’s Platform 最早出現於 AEWIN。

Wi-Fi, the ubiquitous wireless technology that keeps us connected, has come a long way since its beginning. The evolution of Wi-Fi standards has taken us from these dial-up days to an era of blazing-fast connectivity. Today, let’s take a trip through the history of Wi-Fi, exploring its past and evolution, and what the future holds for our wireless world.

 Evolution of Wi-Fi Standards

• The Early Days: 802.11b Wi-Fi 1* (1999)
  Imagine dial-up internet, but slower. That’s essentially what 802.11b offered, with a maximum speed of 11 Mbps (megabits per second) on the 2.4 GHz frequency. It was enough for basic web browsing and email, but forget about streaming or video calls.
• Opened Door to the 5 GHz Band: 11a Wi-Fi 2* (1999)
  802.11a focus on the 5 GHz band offered better performance but faced challenges with range and penetration through walls and obstacles. Additionally, its higher power consumption and limited device compatibility meant it wasn’t a universal solution.
• Picking Up the Pace: 802.11g Wi-Fi 3* (2003)
  11g standard has extended link rate to up to 54 Mbps using the same 20 MHz bandwidth as 802.11b uses to achieve 11 Mbps. 802.11g hardware is fully backward compatible with 802.11b hardware. It was rapidly adopted by consumers starting in January 2003, due to the desire for higher speeds and reductions in manufacturing costs. Despite its major acceptance, 802.11g suffers from the same interference as 802.11b in the already crowded 2.4 GHz range.

Note: Wi‑Fi 1, 2, and 3 are named by retroactive inference. They do not exist in the official nomenclature.

• The Wireless Revol802ution: 802.11n Wi-Fi 4 (2008)
  11n Wi-Fi 4 was a game-changer. It introduced MIMO (Multiple-Input, Multiple-Output) technology, which used multiple antennas to send and receive data simultaneously, boosting speeds to a theoretical maximum of 600 Mbps.
• Push the Boundaries Further: 802.11ac Wi-Fi 5 (2014)
  11ac, also known as Wi-Fi 5, pushed the boundaries further with wider channels on the 5 GHz band and even more advanced MIMO techniques. This translated to speeds of up to 6.9 Gbps, making it ideal for demanding applications like 4K streaming and online gaming with minimal lag.
• Faster than Ever: 802.11ax Wi-Fi 6(2019) & Wi-Fi 6E(2020)
  11ax is the latest IEEE standard from the Wi-Fi Alliance for wireless networks. It operates in the 2.4 GHz and 5 GHz bands, with an extended version, Wi-Fi 6E, that adds the 6 GHz band. Wi-Fi 6 delivers theoretical speeds of up to 9.6 Gbps, ensuring a seamless and reliable connection for multiple devices simultaneously.

The Future is Now: 802.11be Wi-Fi 7 (2024)
With the higher networking performance demands and a more responsive connectivity environment, Wi-Fi 7 represents the next step into the future.

Wi-Fi 7 Key Insights

• Faster: Close to 4X the wireless datalink performance, delivering over 36Gbps.
• More Reliable: New Multiple Resource Unit (MRU) lowers multiple user latency by 25%.
• Always-on Connected: New Multi-Link Operation (MLO) improves single user latency by 80% and throughput gains of up to 300%.

Major Features
Let’s take a look at the major features that have been approved as of Draft 3.0:

4096-QAM
Wi-Fi 7 enables significantly faster speeds by packing more data into each transmission. 320 MHz channels are twice the size of previous Wi-Fi generations. 4096-QAM (Quadrature Amplitude Modulation, 4K-QAM) enables each signal to more densely embed greater amounts of data compared to the 1K QAM with Wi-Fi 6/6E, resulting in 20% higher theoretical transmission rates.

MRU and Puncturing
OFDMA (Orthogonal Frequency Division Multiple Access) was introduced in Wi-Fi 6. It divides the radio channel into smaller frequency allocations called RU (Resource Units). Wi-Fi 7 builds on this foundation with a new MRU (Multiple RU) feature. By partitioning the channel, smaller data packets can be transmitted to multiple users simultaneously, which increases throughput and reduces latency in a dense environment. With legacy Wi-Fi 6/6E, when any part of a large high-speed channel is being used by another device, the entire channel is unavailable—and a different channel must be used. With Multi-RU Puncturing, Wi-Fi 7 devices will be able to use other parts of the same high-speed channel not in use to utilize very large channels.

MLO
Multi-Link Operation (MLO) is an aggregation of multiple bands or channels. With MLO, multi-link devices can simultaneously use the 2.4GHz, 5GHz and 6GHz bands under different circumstances, including load balancing according to traffic needs, or data aggregation across multiple bands, which significantly improves overall speeds and greatly reduces connection latency for all connected users highlighting network traffic can still flow seamlessly even if there is interference or congestion. MLO technology will be critical for delivering faster and more reliable video streaming and anything else that requires constant, sustained and real-time throughput.

Conclusion
Development of the Wi-Fi 7 amendment is ongoing with a final version expected by Q1 2024, so the key new technologies introduced in this blog do not cover all its new features. Given these short-listed new features, the Wi-Fi 7 significant enhancements are already expected to allow Wi-Fi to retain the scalability required for the future, increasing adoption of use cases, even when there are strict latency and reliability requirements. AEWIN as a professional provider of network appliance, industrial server and Edge AI solutions, we are constantly embracing new technologies and developing various platforms that deliver huge performance and IOPS, connectivity, scalability, reliability and data management to be able to handle the applications and growth a business need. Please let us know if you have any questions or comments about AEWIN’s platforms. Feel free to reach out to our friendly sales!

這篇文章 Wi-Fi 7 Unleashed 最早出現於 AEWIN。

Spinning drives and fans are the 2 major sources of potential failures inside an otherwise reliable system. Replacing the traditional spinning drives will move the overall system reliability up another grade. Solid-state drives (SSD) are not only more reliable, but generally faster in both ultimate read/write performance as well as access latencies which results to more responsive systems. There is a number of different SSDs, and we’ll quickly go through the major interfaces and form factors. 

What Is a Solid-State Drive (SSD)?
A solid-state drive (SSD) is a device for storing data on non-volatile memory. SSDs have no moving parts and are known to be smaller, more reliable, often have lower power consumption, and typically have much higher input/output performance than hard disk drives (HDDs) and less latency. The majority of SSDs today use NAND flash as the non-volatile memory to store data. NAND memory is stacked into packages and connected to a controller across various channels to improve performance. SSDs come in multiple capacities, media types, interfaces, form factors, and segments to address the large market for data storage.

Storage Interfaces
There are 3 more popular interfaces for attaching SSDs to host systems: SATA, SAS, and NVMe. SATA and SAS have long history of being used for storage to support legacy compatibility with HDDs. NVMe SSDs using the PCIe interface are much faster than SATA and SAS SSDs, due to the scalable nature of PCIe lanes and generational speed increases. NVMe is a logical interface and can be carried in many form factors.

Major Form Factors

EDSFF
The Enterprise and Data Center Standard Form Factor or EDSFF, previously known as the Enterprise and Data Center SSD Form Factor, is a family of SSD form factors for use in data center servers that is being developed by a group of 15 companies working together to address the concerns of data center storage, and is now maintained by SNIA as part of the SFF Technology Affiliate Technical Work Group (SFF TA TWG). ESDFF provides a pure NVMe over PCIe interface. One common way to provide ESDFF connections on the motherboard is through MCIO connectors. Today all the EDSFF family of form factors share the same protocol (NVMe), the same interface (PCIe), the same edge connector (SFF-TA-1002), and the same pinout and functions (SFF-TA-1009).

As the EDSFF specification evolves, different form factors have been specified to fit various use cases. The EDSFF family includes both E1 form factor, as well as the E3 form factor. The E1.S is a small form factor that fits vertically in a 1U chassis, designed wider than an M.2 to accommodate more NAND flash for increased capacity per drive for greater density. E1.S offers improved flexibility for power, performance, scalability, and thermal efficiency. E1.S is also designed to be hot-pluggable for increased serviceability, which is another benefit over M.2.

E1 PCIe x4 connection Designed for 1U servers Short E1.s format Long E1.l format
E3 PCIe x16 connection Designed for 2U+ servers Short E3.s format Long E3.l format

mSATA
mSATA is a miniaturized SATA drive, hence the name: Mini-SATA. It is a popular format for embedded usage due to its smaller footprint. It has uses the Mini PCIe connector and form-factor, but electrically uses the ubiquitous SATA interface, allowing it to be used in wide variety of platform.

Picture originates from Intel

M.2
M.2, formerly known as the Next Generation Form Factor (NGFF), is a specification for internally mounted computer expansion cards and associated connectors. M.2 replaces the mSATA standard, which uses the Mini PCIe physical card layout and connectors. It comes in many standard lengths, such as 22110, 2280, 2242, 2230, and 2224. M.2 supports PCIe, SATA and USB interfaces and comes in various widths and lengths. It also has keying notches on the edge connector to designate various interface or PCIe lane configurations. Given the flexibility, M.2 SSDs are not quite well-suited for data center applications. M.2 SSDs lack standardization in terms of form factor, interface, and power requirements. M.2 SSDs are not hot-pluggable, This can be a major inconvenience in a data center environment where servers need to be able to be serviced without downtime.

Picture originates from Intel

2.5”
The 2.5-inch form factor is the most common deployment of an SSD, and is offered with PCIe (with NVMe), SAS or SATA interfaces. It is typically used in desktops, servers and storage systems built around hard disk drives (HDD). This form factor is commonly associated with the term U.2 and is sometimes referred to as the U.2 form factor. U.2 is defined as compliance with the PCI Express SFF-8639 Module specification. U.3 is built on the U.2 spec and uses the same SFF-8639 connector. It is a “tri-mode” standard, combining SATA, SAS and NVMe support into a single controller. U.3 drives are still backward compatible with U.2, but U.2 drives are not compatible with U.3 hosts.

Picture originates from Intel

Add-In Card (AIC)
An Add-in Card (AIC) is a solid-state device that utilizes a standard card form factor such as a PCIe card. Almost all early NVMe SSDs are HHHL (half height, half length) or FHHL (full height, half length) AIC, which are easy to insert into a PCIe slot of a server. Given the larger physical size, the AIC would typically have larger capacity and potentially higher performance. In addition, the larger size allows for the potential to add computational function to the storage device.

Picture originates from Intel

Summary
Technology advances increase and improve the choices we do certain things. Over time, changes in form factor, materials, and communications between devices, have created evolutions of using technology from one form to another. The rise of x86 systems in the 1990s over mainframe is a perfect example of how technology affects form factor choices. For many users, SATA SSDs are fast enough to meet their everyday data storage and transfer needs. At a lower price point, they remain an attractive option. On the other hand, at the enterprise level, NVMe is fast becoming the industry standard. With a focus on carrying NVMe into the PCIe 5.0, 6.0 ages and beyond, we’re seeing higher density and performance of data centers.

AEWIN as a professional provider of Network Appliance, Edge AI Systems and Industrial Servers, we are constantly developing various platforms with lowest latency that support NVMe and others. They deliver huge performance and throughput, IOPS, scalability, reliability and data management to be able to handle the applications and growth a business need. Please let us know if you have any questions about AEWIN’s platforms. Feel free to reach out to our friendly sales!

這篇文章 Introduction to SSD Form Factors 最早出現於 AEWIN。

Corporate sustainable development is one of our core values as part of the Qisda Group. We are aware that only through SDGs can we protect the Earth’ s environment and continuously create economical value for the society. In our previous blog, we introduced AEWIN’s green product design that delivers improved PUE, power efficiency, and carbon reduction. Today we will be talking about the green packaging that lowers the environmental impact of products.

Introducing Green Packaging Design
Green packaging design, also known as sustainable packaging or eco-friendly packaging, is the practice of creating packaging that is environmentally friendly and sustainable. It takes into account the entire life cycle of the packaging, from the extraction of raw materials to the disposal of the used packaging. Green packaging designers aim to reduce the environmental impact of packaging by using sustainable materials, minimizing waste, and designing for recyclability and reusability.

Key Principles of Green Packaging Design

• Consider the full life cycle: Green packaging designers consider the entire life cycle of the packaging, from the raw materials acquisition to the waste disposal and recycling of the used packaging. This helps to identify and reduce environmental impacts at all stages of the life cycle.
• Use sustainable materials: Green packaging designers use sustainable materials that are recycled, biodegradable, or compostable. This helps to reduce the use of virgin materials and conserve natural resources.
• Minimize waste: Green packaging designers aim to minimize waste at every stage of the packaging life cycle. This includes reducing the amount of packaging used, using efficient packaging techniques, and designing packaging that is easy to recycle or reuse.
• Design for recyclability: Green packaging designers design packaging that is easy to recycle. This includes using recyclable materials and designing packaging in a way that makes it easy to separate and recycle the different components.
• Design for reusability: Green packaging designers design packaging that can be reused multiple times. This helps to reduce waste and save cost.

Examples, Challenges and Opportunities
Here are some examples of green packaging design that are being adopted at AEWIN:

• Recyclable packaging: Recyclable packaging is made from materials that can be recycled into new products, such as molded paper pulp packaging that replaces traditional EPE foam. Compared to traditional EPE foam, the molded paper pulp packaging is smaller in size and can be stored in overlapping way to save space. It is also featured with lightweight and reusability while providing excellent shock resistance and cushioning. Molded paper pulp packaging is also biodegradable and compostable, so it is a good choice of sustainable packaging options.
• Minimalist packaging: Minimalist packaging uses the least amount of material necessary to protect the product, which means it produces less waste and greenhouse gas emissions during production and transportation. It also reduces the amount of waste that goes to landfills and incinerators.

There are a number of challenges and opportunities in green packaging design though. One challenge is the cost of sustainable materials. Green packaging materials can sometimes be more expensive than traditional packaging materials. However, the cost of green packaging is coming down as more sustainable materials are developed and the demand for green packaging increases.

Another challenge is the lack of recycling infrastructure in some parts of the world. This can make it difficult for users to recycle green packaging, even if it is designed to be recycled. However, recycling infrastructure is improving in many parts of the world, and businesses are working to develop green packaging solutions that can be recycled even where infrastructure is limited.

Despite the challenges, there are opportunities in green packaging design. Customers are increasingly demanding sustainable packaging. By choosing green packaging, customers support businesses that are committed to sustainability. This helps to create a more sustainable economy.

AEWIN is continuously conducting the green design measures and these measures not only benefit the Company, but also facilitates the healthier development of our customers, communities, suppliers and the global environment. There are many aspects in the technology area that are worth discussing and we definitely continue to promote more ESG initiatives.

這篇文章 AEWIN Platforms Eco Design for ESG Requirements 最早出現於 AEWIN。

In our previous blog, we announced AEWIN SCB-1932C Server has been validated as a NVIDIA-Certified System for enterprise edge. Today we will explore more on the GPU benchmark tests across different AEWIN platforms.

System Configurations
Applying AEWIN High Performance Appliances, SCB-1946C, SCB-1932C, and SCB-1937C.

GPU Status Monitor
For preparation, write a GPU monitor script “monitor.sh” in the host in case of throttling.

Input the status refresh duration as interval. input “y” to save log or “n” not to save log.

Benchmark Test
Run the script “benchmark.sh” from the host. It will redirect you to the GPU accelerated container. From the container run the script “benchmark.sh”. It will ask to choose between int8 mode or fp16 mode for the test. Input 1 to run in int8 mode.

Run the script “benchmark.sh” in the host to start the test. The picture below shows an example of the benchmark results.

For the Benchmark results, we only consider the percentile value of the GPU Compute. For example, the percentile value shown in the above figure is equal to 8.88623. To calculate the performance in img/sec for any GPU, we use the following formula: 1000/(percentile/128), where 128 is batch size for current test. Thus, the int8 (images/sec) is equivalent to 14,405.

Testing Script
1. sh script in the container

2. sh script in the host

3. burn-in script burn.sh in the container

4. burn-in script burn.sh in the host

5. GPU monitor script “monitor.sh” in the host

Summary
As shown in the benchmark results, we verified A30 on the platforms including SCB-1946C(Genoa), SCB-1932C(Ice Lake), and SCB-1937C(Milan). They share better or similar results compared to Nvidia benchmarks.

platforms range from edge AI appliances to general purpose computing systems, to high performance servers, customers can select the most suitable ones with the GPUs required for each application. Reach out to our friendly sales and discover more on AEWIN GPU Server platforms!

• SCB-1932: 2U Dual Ice Lake-SP PCIe 4.0 Platform with short depth design, 4x PCIe Gen4 slots plus dual FHFL GPU slots or 4x PCIe Gen4 NIC, and IPMI.
• SCB-1933: 2U Ice Lake-SP PCIe 4.0 Platform with short depth design, 4x PCIe Gen4 slots plus dual FHFL GPU slots or 4x PCIe Gen4 NIC, and IPMI.
• SCB-1942: 2U Dual Sapphire Rapids-SP PCIe 5.0/CXL Platform with short depth design, 4x PCIe Gen5 slots plus dual FHFL GPU slots or 4x PCIe Gen4 NIC, and IPMI.
• SCB-1943: 2U Sapphire Rapids-SP PCIe 5.0/CXL Platform with short depth design, 4x PCIe Gen5 slots plus dual FHFL GPU slots or 4x PCIe Gen5 NIC, and IPMI.
• SCB-1946: 2U Dual EPYC-9004 (Genoa/Bergamo) PCIe 5.0/CXL Platform with short depth design, 4x PCIe Gen5 slots plus dual FHFL GPU slots or 4x PCIe Gen4 NIC, and IPMI.
• SCB-1947: 2U EPYC-8004 (Siena) PCIe 5.0/CXL Platform with short depth design, 8x PCIe Gen5 slots NIC, NVMe, and IPMI.
• BAS-6101A: 2U High-Density Edge Computing Server with AMD Bergamo/Genoa/Genoa-X processor, total 8x PCIe slots (2x dual width FHFL PCIe Gen5 x16 or 4x single width FHFL PCIe Gen5 x16, 2x single width FHHL PCIe Gen5 x16, 2x HHHL PCIe Gen4 x8) + 1x OCP 3.0 slot for NICs and Accelerators.
• BAS-6101B: 2U High-Performance Server with AMD Bergamo/Genoa/Genoa-X processor, total 8x PCIe slots (2x dual width FHFL PCIe Gen5 x16 or 4x single width FHFL PCIe Gen5 x16, 2x single width FHHL PCIe Gen5 x16, 2x HHHL PCIe Gen4 x8) for NICs and Accelerators.

這篇文章 GPU Benchmark Tests of Genoa, Milan and Ice Lake Platforms 最早出現於 AEWIN。

What is C-TPAT?
C-TPAT stands for Customs-Trade Partnership Against Terrorism. It is a voluntary supply-chain security partnership program led by U.S. Customs and Border Protection (CBP) focused on improving the security of private companies’ supply chains with respect to terrorism.

The goal of C-TPAT is to strengthen international supply chains and improve United States border security. The program works by partnering with businesses to improve their security practices and procedures. In return for their participation, C-TPAT Partners are able to better identify their own security vulnerabilities and take corrective actions to mitigate risks.

C-TPAT Benefits
Some of the benefits of the program include:

• Reduced number of CBP examinations
• Front of the line inspections
• Possible exemption from Stratified Exams
• Shorter wait times at the border
• Assignment of a Supply Chain Security Specialist to the company
• Access to the Free and Secure Trade (FAST) Lanes at the land borders
• Access to the C-TPAT web based Portal and a library of training materials
• Possibility of enjoying additional benefits by being recognized as a trusted trade Partner by foreign Customs administrations that have signed Mutual Recognition with the United States
• Eligibility for other US Government pilot programs
• Business resumption priority following a natural disaster or terrorist attack
• Importer eligibility to participate in the Importer Self-Assessment Program (ISA)
• Priority consideration at CBP’s industry-focused Centers of Excellence and Expertise

C-TPAT Security Criteria
The C-TPAT minimum security criteria can be divided into three focus areas:

• Corporate security: This includes having a written security policy, conducting background checks on employees, and training employees on security procedures.
• Physical security: This includes securing facilities and cargo, controlling access to sensitive areas, and using security seals and other tamper-evident devices.
• Transportation security: This includes conducting security assessments of carriers, using secure containers and trailers, and tracking and tracing shipments.

Conclusion
C-TPAT is a valuable program that help businesses to improve their security posture and reduce their risk of being targeted by terrorists. AEWIN is a Certified member of the US Customs Border Protection’s (CBP) C-TPAT program. The program is internationally recognized by member states of the World Customs Organization and demonstrate highest supply chain security standards within the private sector to secure trade across international borders.

Please let us know if you have any questions or comments about AEWIN’s supply chain and hardware integrity. Feel free to reach out to our friendly sales!

這篇文章 Introducing C-TPAT Program(Customs-Trade Partnership Against Terrorism) 最早出現於 AEWIN。