Meeting the Demands of Future Data Centers.
The software infrastructure for Big Data and Analytics are often built on high performance databases, enterprise and custom line-of-business applications. An overwhelming number of these databases are deployed in legacy hardware where they are forced to work with smaller data sets, spend a significant amount of time fine-tuning storage and get less done in a typical work day. For instance, Hadoop deployments are designed for batch processing but storage performance limitations do not allow working on large data sets while real-time analytics are in process.
This problem is further exasperated by the architectural limitations of storage and compute being coupled together. By contract, enterprise-wide line-of-business applications typically do not require massive amounts of storage but need the highest performance in bandwidth, IOPS and low latency that IT can muster.
The principal goal of Smart IOPS’ TruRandom technology is to build a highly scalable, high-performance storage platform from scratch such that it takes full advantage of the performance characteristics of flash and next-generation nonvolatile memory (NVM1) technologies.
The performance boost and low latency capabilities of TruRandom are achieved by implementing a fully optimized non-blocking, asynchronous SSD controller architecture.
truRANDOM®: NOVELL CONTROLLER & FIRMWARE
Built For High Random Performance
- Highly tuned firmware for NAND management
- Specialized machine learning heuristics
- Non-blocking, out of order, fault tolerant
- Suitable for DB, AI/ML, OLTP, CDN, Analytics
Plug and Play NVMe SSD
- No external host SW based management
- Reduces support and maintenance issues across OS versions through life of product
- Host cycles spared from maintaining drive
- Our Storage Computing can further reduce CPU workload
- Adaptive performance and device management techniques
- Device reliability to improve BER
- Full power loss protection
- Robust, enterprise class
truRANDOM®: THE ULTIMATE RANDOMIZER
Intelligent adaptation of work loads
- Read, write, mix, random, sequential
Fine regulation of external/internal traffic
- Optimizes system efficiency
- Host, Garage Collection, Wear
- Results in a smooth IO operations
Highly optimized NAND operation
- Produces best latency and IOPS results
- Optimize writes for lower energy to improve endurance
- Unique data-reshaping, extreme ECC (400+bits)
- Optional compression and De-Dupe
We have done extensive work to interoperate and perform efficiently with different plug and play NVMe driver software that come pre-installed with operating systems like Windows, Mac OS, and Linux. This removes the complexity of maintaining dedicated driver software every time there is a new OS upgrade and simplifies the deployment and maintenance of our product significantly.