The storage market continues to grow and new capabilities are needed to move, manage, and protect stored data. Storage processing features—such as virtualization, data protection, data security (encryption), and data compression—are essential to increase storage capacity. In addition, storage solutions are rapidly adopting PCIe-based non-volatile memory express (NVMe) while moving serial interconnects continue. Protocols such as Fiber Channel, Serial Attached SCSI (SAS), and Serial ATA (SATA) and NVMe dominate the storage protocol landscape.
With feature enhancement comes a need to do more with less. SoC solutions allow you to consolidate functions into a single silicon solution. The combination of the high-speed transceivers, dense logic, and deep memories of Intel® FPGAs coupled with the intellectual property (IP) solutions from us and our partners allow you to easily create an ideal storage solution. Our storage solutions support: online, nearline, and offline storage.
Data centers constantly use online storage for performing real-time data transactions for server applications. Online storage is solid-state-based storage that resides in or is attached to (direct or fabric) a server. With direct-attached storage, only the server attached to the storage can access it. Fabric-attached storage enables all servers attached to the fabric to share the available storage resources. Storage shared by multiple servers is commonly referred to as a storage area network (SAN). SAN protocols are usually Fiber Channel, SCSI, or Ethernet (in the case of a network attached storage (NAS) enclosure). The in-box memory is an area that is being developed in the Hyper converged market for database acceleration. For improvements in latency from the network RDMA over converged Ethernet (ROCE) is starting to appear in many different architectures.
Online storage devices, such as just a bunch of disks (JBODs) and disk arrays, allow high-speed access to the storage, while at the same time providing data protection and security. The high-speed access to the storage is achieved with the high-speed I/O for the network, system bus, and disk drive interfaces. Data protection and security are provided with algorithms such as RAID and erasure codes for data protection and encryption.
Nearline storage has many of the same features, performance, and device requirements as online storage. However, nearline storage is deployed for backup support for online storage. Demand for nearline storage is growing rapidly because more information must be archived for regulatory reasons. Nearline storage is frequently used for data backup because large volumes of data must be quickly backed up, which sometimes cannot be achieved with slower bandwidth rates to tape-based solutions. Nearline storage is built using less expensive disk drives such as SATA drives to store information that must be accessed more quickly than is possible through tape or tape libraries.
Nearline storage designers must meet several requirements in their designs:
- Flexibility with in-house I/O sequencing and RAID algorithms
- Changing data-at-rest encryption standards
- Use of specific compression algorithms
- Tape emulation for existing customer back-up software
Most key storage vendors of nearline storage use FPGAs to meet these design challenges. We provide a reference architecture and both in-house and partner IP for nearline storage designs.
Offline storage is commonly referred to as “archive” or “back up” storage and is typically a tape drive or low-end disk drive (virtual tape). Offline storage is used to back up the data stored on both the online and nearline storage devices and is designed for storage of data for long periods of time. Because data is archived, offline storage appliances focus on data accuracy, protection, and security.
New developments in storage controller architectures and continued advances in the density and speed of NAND devices are driving performance and capacity increases in the data center. One example is large data sets using Apache Spark that enable applications in Hadoop clusters to run up to 100 times faster in memory and 10 times faster on disk. By carefully managing the performance and lifetime of the NAND and moving storage closer to the processor, latency is reduced and performance increased compared to centralized and network-based storage. Combining algorithm acceleration, data correction, network acceleration and NVM control into one flexible FPGA is improving the efficiency, power and speed while transforming the way data flows through high-performance computing (HPC) and cloud computing systems.
Intel® FPGAs, IP, and ecosystem partners combine to provide an ideal solution for development of online, nearline, and offline storage solutions. Intel® Stratix® series, Arria®, Cyclone®, and MAX® devices are invaluable to storage system architects. These platforms allow you to develop extremely flexible applications quickly, efficiently, and affordably. They allow you to craft unique, scalable, and optimized application-centric processing solutions for SAN and NAS appliances. Table 1 shows the storage features available in Intel FPGAs.
Table 1. FPGA Storage Features
|Category||Network/Fabric I/O||Storage I/O||Data Protection||Data Security|
|AES Type||Key Management and Authentication|
|Online||Fibre Channel Ethernet
ASSP or Switch
|RAID 5, RAID 6||XTS-AES, AES-GCM||Authentication
Secure hash algorithm (SHA) 1, SHA 256, SHA 384/512, MD5, Multi-mode (SHA 1. SHA 2, MD5), Advanced encryption system (AES)-based (XCBC, OMAC)
RSA and Diffie-Hellman (Mod. EXP), True Random Number Generator (TRNG), Pseudo Random Number Generator (PRNG)
|Nearline||Fibre Channel Ethernet, PCIe||SATA||RAID 5, RAID 6||XTS-AES|
|Offline||Fibre Channel Ethernet||Fibre Channel||RAID 5, RAID 6||XTS-AES, AES-GCM, AES-CCM|