A new line of servers from IBM could help make the enterprise data center more efficient, while providing additional computer power for deep learning and high performance data analytics.
IBM is releasing a new line of Linux-based servers that are designed to handle computer-intensive workloads and make data centers more efficient, the company announced Thursday. The servers could make it easier to run workloads for AI, deep learning, and big data analytics.
The new servers are a part of IBM's LC line, and they include the S822LC for High Performance Computing, the S822LC for Big Data, and the S821LC. According to an IBM press release, the "systems deliver average of 80% more performance per dollar than latest x86-based servers."
All three systems are two socket, and are configurable with up to 20 cores. The high performance computing model has 1 TB Memory with 230GB/sec memory bandwidth, while the other two options have 512 GB Memory with 115GB/sec memory bandwidth. The high performance model uses the NVIDIA Pascal GPU, and the other two work with the NVIDIA K80.
According to a press release, Chinese ISP Tencent began testing the new servers, and was able to run their workloads up to three times faster with fewer overall servers. Currently, Tencent is working on integrating the LC servers into one of its data centers.
SEE: Power checklist: Managing and troubleshooting servers (Tech Pro Research)
"The user insights and the business value you can deliver with advanced analytics, machine learning and artificial intelligence is increasingly gated by performance. Accelerated computing that can really drive big data workloads will become foundational in the cognitive era," IBM's Doug Balog said in a press release.
Much of the innovation present in the new line came through collaborations with the OpenPOWER Foundation. This is seen in the IBM Power System S822LC for High Performance Computing server, which utilizes a new IBM POWER8 chip with NVIDIA NVLink to get a 5X performance boost over its X86 counterparts, the press release said.
"NVIDIA NVLink provides tight integration between the POWER CPU and NVIDIA Pascal GPUs and improved GPU-to-GPU link bandwidth to accelerate time to insight for many of today's most critical applications like advanced analytics, deep learning and AI," Ian Buck, vice president of Accelerated Computing at NVIDIA, said in a press release.
The US Department of Energy's Oak Ridge National Laboratory and Lawrence Livermore National Laboratory will be among the first to use the high performance computing variant, and IBM noted that these systems will likely build a foundation for new supercomputers IBM is currently developing.
The S822LC for Big Data and the S821LC are available now, but the S822LC for High Performance Computing will ship on September 26, 2016.
The 3 big takeaways for TechRepublic readers
- IBM recently announced a new line of LC servers that could improve enterprise efforts around compute tasks relative to big data and artificial intelligence.
- The servers were developed through collaborations with the OpenPOWER Foundation, and IBM said they can deliver better performance than X86 competitors.
- Two of the servers are available now, but the high performance variant won't be available until late September.
- How to use darkstat to monitor Linux server network traffic (TechRepublic)
- Linux Mint 18: The best desktop -- period (ZDNet)
- Why AI could destroy more jobs than it creates, and how to save them (TechRepublic)
- Salesforce tips its AI master plan, previews Einstein (ZDNet)
- Can AI and big data improve how you get news? Cheetah Mobile is making a $57M bet that it can (TechRepublic)