SAN FRANCISCO—IBM Corp. said Wednesday (Sept. 1) that it will begin shipping Sept. 10 a new mainframe computer computer capable of 50 billion instructions per second, powered by 96 microprocessors with clock speeds up to 5.2 gigahertz.
IBM (Armonk, N.Y.) said the z196 processor is a four-core chip that contains 1.4 billion transistors on a 512-square millimeter surface. The chip was designed by IBM engineers in Poughkeepsie, N.Y., and was manufactured using IBM's 45-nm silicon-on-insulator process at the company's 300-mm fab in East Fishkill, N.Y., IBM said.
The mainframe processor makes use of IBM's embedded DRAM technology, which allows IBM to place dense DRAM caches, or components, on the same chips as high-speed microprocessors, resulting in improved performance, according to the company.
A spokesperson for IBM said the company was not disclosing any other z196 benchmark data at this time.
The new mainframe, the zEnterprise System, is the most powerful commercial IBM system ever, according to the company, capable of executing roughly 17,000 times more instructions per second than the most advanced system available in 1970, according to the company.
The z196 processor features new software to optimize performance of data-heavy workloads, including up to a 60 percent improvement in data intensive and Java workloads, according to IBM.
An IBM technician tests the z196 processor, claimed to be the world's fastest with speeds up to 5.2 GHz. (Photo courtesy of IBM).
Last week, IBM engineers at the Hot Chips conference sketched out plans for a petaflops-class supercomputer built from as many as 64,000 Power7 processors.
The zEnterprise System offers 60 percent more capacity than its predecessor, the System z10, but uses about the same amount of electricity, IBM said. Energy efficiencies were achieved through advances in microprocessor design, 45-nm silicon technology, more efficient power conversion and distribution and advanced sensors and cooling control firmware that monitors and makes adjustments based on environmental factors such as temperature and humidity levels and air density, according to IBM.
IBM said the speed of the zEnterprise System is needed to support businesses with large workloads, such as banks and retailers. Citing a study conducted by Berg Insight, IBM said the number of active users of mobile banking and related financial services worldwide is forecasted to increase from 55 million in 2009 to 894 million in 2015.
Pricing information for the zEnterprise System was not disclosed.
IBM has invested more than $1.5 billion in R&D on the zEnterprise line, including more than three years of collaboration with some of the company's top clients around the world, IBM said.
IBM labs in Austin, Texas, Germany, Israel and India made major contributions to the development of the z196 processor, the company said.
The information provided in the article doesn't really show how good this processor is. Comparing the speed of execution with 1970 makes it really jeopardize to understand clear details. But definitely IBM put a lot of money into it and we just need to wait and see for more technical details regarding the processor and its performance. Is there any other processor with dense DRAM populated on the processor chip just like this. This is the one point sounds interesting ...
I would love to see the performance and the benchmarking results of the server made up of using this processor, it would be giving remarkable performance provided OS is being able to take the power of the CPU.
Completely agree. I had the same reaction when I saw the 50 billion instructions per second figure. Actually, even the 50 billion instructions per second per processor is not that impressive by today's Intel microprocessor standards. Unless we see real benchmark results, we can't say much....
I looked it up - the 50 billion does refer to a single processor, so 96 of them would be approx 5 trillion. I really hate that kind of useless comparison, which is almost as bad as tabloids measuring power by the number of cups of coffee that can be boiled by a given generator.
Is this a joke?
"... capable of executing roughly 17,000 times more instructions per second than the most advanced system available in 1970, according to the company."
"...capable of 50 billion instructions per second, powered by 96 microprocessors..."
The numbers are ludicrous.
David Patterson, known for his pioneering research that led to RAID, clusters and more, is part of a team at UC Berkeley that recently made its RISC-V processor architecture an open source hardware offering. We talk with Patterson and one of his colleagues behind the effort about the opportunities they see, what new kinds of designs they hope to enable and what it means for today’s commercial processor giants such as Intel, ARM and Imagination Technologies.