RNA Preps Analysis Memory Cloud

The vendor is in the final stages of completing development, and plans to announce a general availability date for the platform this month, officials say.

"The key is to expand on our 'memory visor' concept-a layer that uses a shared, distributed memory cloud service... to enhance performance in different ways," says Rod Butters, chief marketing officer and vice president of products at RNA.

The MVX cloud allows applications to use spare memory on other nearby machines, regardless of the applications and operating systems they are running, enabling developers to build applications whose memory size reflects the demands of the application rather than the constraints of the server on which it runs.

Butters says the expanded memory will appeal to business functions that need to analyze large amounts of tick data, or that want to drill down from a tranche of instruments to the individual notes, or run risk analysis in real time that might only have been performed on a daily basis.

"A lot of applications out there haven't been written to take advantage of multiple processors... and they run out of processing space pretty quickly. You can buy a bigger server, but in six months you need another one," Butters says. "This is a way to change the parameters of applications to consume more memory footprint."

He says MVX enables applications to access data held in shared memory at latency measured in microseconds-compared to disk speeds in the tens of milliseconds-when connected using high-speed layers such as 10 Gigabit Ethernet. Although not a cheap option, "compared to huge boxes with their upgrade cycles, or completely re-architecting applications, it's a very cost-effective option," Butters says.

Max Bowie