The SuperCluster, along with Exadata and Exalogic, is based on parallel processing, in which different compute instructions are routed through multiple threads within processors at the same time. Parallel computing also occurs between the database and storage. But Oracle's technology goes beyond that, said John Fowler, executive VP of systems at Oracle, who held a similar title at Sun before it was acquired.
"We're not just playing a processor game here. We're not just tuning one little part of the system," said Fowler in his presentation. "The architecture of the Exadata storage cells is that instead of a server asking for a block of data, we're actually constructing queries that are executed in parallel across the different storage devices. It is not possible for any conventional storage system to equal this."
While some of Oracle's claims may seem strong, specification groups and people at other tech companies test systems to verify such claims, says Nathan Brookwood, a research fellow at Insight 64, a technology research firm. Brookwood said he considers the T-4 processor "a very strong product."
"They used to be able to make all these interesting claims about throughput, but if you had applications that didn't parallelize well, then you saw very unimpressive performance," Brookwood says. "Now, it looks like they've done a much better job on the single-thread aspect, and that should, all other things being equal, result in dramatic performance improvements."
Oracle also introduced a series of servers running the T-4 processor and the Solaris 11 operating system. The company did not reveal pricing information on the SuperCluster.
See more on this topic by subscribing to Network Computing Pro Reports Research: Storage & File Virtualization (subscription required).
