Start-Up Profile: Smart IOPS

Location
HQs in Milpitas, CA, office in Bangalore, India

Date founded
November 2013

Financial funding
Private funding and angel investors (no figures revealed)

Main executives:

Ashutosh Das, CEO and co-founder, previously president and CEO of Picogem during six years, director of advanced technology at Juniper Networks, after working at Intel, Sun Microsystems and Nexsi as senior technology architect

Dr. Manuel D’Abreu, chief scientist and SVP, co-founder, IEEE Life Fellow, formerly senior Fellow at AMD and SanDisk

Radhakrishnan (RK) Nair, COO and co-founder, coming from SanDisk, Micron, HAL, S3 and C-DoT

Number of employees
30

Technology
Company’s architecture, TruRandom, minimizes the overhead associated with internal SSD housekeeping, garbage collection and program/erase to unleash the raw I/O processing capability in the SSD. Additional patent-pending logic enhance data pattern recognition heuristics for data access such that random and sequential I/O requests become indistinguishable, a feat not achieved in current generation of high-end competing SSDs. This advantage translates to IO/s limits only within the reach Smart IOPS Data Engine SSDs, nearly 3x faster than the closest competitor with sub-10µs latency, according to the start-up.

Products

Data Engine NVMe SSD Add-In Card

Specs of Data Engine SSDs Add-In Cards

Smart IOPS Shared Flash Appliance

Specs of Smart IOPS Shared Flash Appliance

* Effective capacity assumes HA, RAID, and metadata overhead, GB-to-GB conversion, and includes de-dupe, compression and pattern removal.

Availability
Sampling to select OEM and enterprise customers

Price
$4,000 to $20,000 for SSDs

Roadmap
2017

Data Engine T2 High Performance, up to 10TB
Data Engine T2 High Capacity HC, up to 20TB
Data Engine T2 Ultra-Low Latency (ULL)

2018-2019

Data Engine T4, up to 16TB
Data Engine T4 (HC), up to 24TB
Data Engine T4 ULL

Partners
Gigabyte, Wistron, Xilinx, Micron, Toshiba

Customers
OEM/ODM, tier-1 cloud service providers, US Federal agencies, HPC and high end enterprise data centers

Applications
Mixed-use workloads including High Frequency Trading (HFT), cloud and hyperscale data center applications, OLTP and online analytical processing, big data analytics, SQL and NoSQL for structured and unstructured data, VDI, etc.

Target market
HPC and high end enterprise data centers compute and storage nodes

Competitors
All PCIe NVMe SSD vendors