Legacy NUMA technology helps the aging Itanium servers move to Xeon. Credit: HPE Can two old technologies make it in the brave new world of cloud computing? Hewlett-Packard Enterprise (HPE) thinks yes with its updates to the Superdome servers featuring its recently acquired SGI technology. HPE bought SGI in late 2016, ending for good the name and company that was once a Hollywood darling. The company enjoyed almost celebrity status around 1993 when its computers were credited for creating the realistic dinosaurs in the movie Jurassic Park. After that, though, it fell on hard times as the environment zigged and poor management zagged, blowing one opportunity after another. In 2009, Rackable Systems acquired SGI’s assets and changed its name to SGI because it thought there was some value in the name — for some inexplicable reason. However, the company continued to languish before HPE finally bought it out. While it was never a major player, SGI still had considerable technological assets, which was what HPE wanted, and now they are coming to market. HPE’s new Superdome Flex HPE has announced new Superdome servers with new in-memory computing based on SGI NUMAlink technology. Superdome has always been HPE’s top-of-the-line platform with a massive scale-up architecture meant for the most mission-critical tasks. Unfortunately, they were based on Intel’s deadend Itanium processor up until recently. Seeing the writing on the wall, HPE transitioned to the Intel Xeon E7 processor, the high end of the Xeon line with many of the vital RAS (reliability, availability and scalability) features found in Itanium and RISC processors. The new product, known as the Superdome Flex, offers a modular design that can scale up to 32 CPUs and 48TB memory, making it suitable for massive in-memory workloads like Apache Spark, SAP HANA, Microsoft SQL and Oracle Database In-Memory. The servers, which come in a 5U chassis, can also hold GPUs, which are popular in high-performance computing. The secret sauce is SGI’s NUMAlink interconnect. NUMAlink is a fabric that links large numbers of CPUs and memory components. This is a very mature technology, dating back to when SGIs ran MIPS processors, that has been running on Intel for years now. HPE’s Superdome X and MC990 X servers HPE has rebranded the SGI UV 300 server as the MC990 X product and is targeting it at these massive in-memory computing jobs. The Superdome X system has 16 sockets and 48TB of memory, while the Superdome Flex has 32 sockets and 48TB of memory. One big difference is Superdome Flex is based on the newer Skylake generation of processors, while MC990 X and Superdome X use the older Broadwell generation of processors. Another advantage for the Superdome Flex is it has the memory protection features in firmware that the Superdome X has and uses the Memory-Driven Computing design HPE is developing for The Machine project. SGI may not have been a big player, but it had good tech, and HPE will benefit from it. The Superdome is making an interesting migration, from HP-UX running on Itanium to Linux on Xeon. And with the NUMA technology, it’s positioned to be a major player in in-memory applications, which are growing in popularity as companies look for more speed and performance. Related content news Omdia: AI boosts server spending but unit sales still plunge A rush to build AI capacity using expensive coprocessors is jacking up the prices of servers, says research firm Omdia. By Andy Patrizio Dec 04, 2023 4 mins CPUs and Processors Generative AI Data Center news AWS and Nvidia partner on Project Ceiba, a GPU-powered AI supercomputer The companies are extending their AI partnership, and one key initiative is a supercomputer that will be integrated with AWS services and used by Nvidia’s own R&D teams. By Andy Patrizio Nov 30, 2023 3 mins CPUs and Processors Generative AI Supercomputers news VMware stung by defections and layoffs after Broadcom close Layoffs and executive departures are expected after an acquisition, but there's also concern about VMware customer retention. By Andy Patrizio Nov 30, 2023 3 mins Virtualization Data Center Industry news AI partly to blame for spike in data center costs Low vacancies and the cost of AI have driven up colocation fees by 15%, DatacenterHawk reports. By Andy Patrizio Nov 27, 2023 4 mins Generative AI Data Center Podcasts Videos Resources Events NEWSLETTERS Newsletter Promo Module Test Description for newsletter promo module. Please enter a valid email address Subscribe