Let’s refocus the spotlight on Big Blue for uncovering a new kind of memory chip that can record data 100 times faster than the contemporary flash memory. This breakthrough is a peep into the future of the universal memory chip; one that is fast, non-volatile and with incredible storage.
Phase change memory is an alloy of various unnamed elements. It leverages with resistance by changing from crystal phase to an amorous phase through electrical voltage or pulses application in a range of one to zero (basic bits for digital memory systems), and with a latency of 10 milliseconds. The crystalline phase has low resistance while the amorphous phase has higher, enabling for multiple bits to be stored in a single memory cell. IBM managed to put 4 bits in one cell, pretty much like putting 4 people in a single family home, subsequently decreasing the cost per person.
If this new multi-bit phase-change memory gets pushed through, they are predicted to change the face of enterprise computing by 2016. IBM’s new chips would be faster, cheaper and more durable in storing applications, fitted for handheld devices, as well as cloud and enterprise storage. IBM researchers have successfully demonstrated the reliability of phase-change memory in storing multiple bits of data per cell within the expected timeframe.
Universal memory in computers will enable an impressive storage capacity and superb performance, such as Apple’s flash-based MacBook Air, but phase-change memory can reportedly retrieve and write data a hundred times faster than flash, as well as offer higher storage capacity and retention of data even if the power goes down. These are some big claims against the flash storage industry, but we shouldn’t get all worked up about this technology until IBM says it can mass produce the chip.
Since IBM’s breakthrough chips are not hitting the market anytime soon, let’s also talk about current storage happenings right now. We’ve got Flashsoft, an all-software solution for freeing up server computing resources, leveraging flash as cache. It is designed to eliminate IO bottleneck while enhancing enterprise application performance. They offer support to companies like Fusion-io, which are pressing on flash storage for enterprise-level utilization. Flashsoft resides at the SSD level within the software, closest to the server in order to reduce processing and data transfer time. They seek to modify inherent hierarchy of existing server stack and are operating on “tier minus one.” They were recently funded with a $3 million round, led by Thomvest Ventures.
Also pushing the flash storage niche, SolidFire launches a faster cloud SSD storage solution. These SSD devices hope to decrease providers’ storage footprints, while extensively boosting performance and improving speed of cloud storage. Fusion-io’s recent IPO, launching them to $1.4 billion valuation and one of the strongest contender for information technology in the stock market, has inspired many to take on the flash memory solutions business.
Another piece of big news is from the SAN array vendor; Pillar Data’s acquisition by Oracle. This isn’t much of a surprise, as Pillar has always been connected to Oracle, having been funded by company CEO Larry Ellison and his investment arm, Tako Ventures. The deal isn’t in cash or stock, but rather an offering of a buyout based on Pillar’s performance in three years. So far, the company has sold only 1500 systems to some 600 customers. They have to work harder.