Inspur G7 Servers for Intel Xeon Sapphire Rapids Launched

On the heels of the 4th Gen Intel Xeon Scalable launch, codenamed “Sapphire Rapids,” Inspur has new servers. While some companies are launching narrow portfolios with general-purpose 1U and 2U models, Inspur is launching a complete line spanning different optimization points.

Inspur G7 Servers for Intel Xeon Sapphire Rapids Launched

We covered this some time ago, but Inspur has changed its naming convention. Previously models had the generation as the last digit, that continues with the G7. The big difference is that the old convention had “M” for Intel, “A” for AMD, “R” for Ampere and so forth. Now, everything is G7, and the processor vendor is later in the full model number. Here is an overview of the company’s new Xeon servers:

• The Inspur TS860G7 is the company’s 6U server that supports up to eight sockets. That means it also supports up to 128x DDR5 DIMM slots
• The Inspur NF8260G7 is the company’s 2U 4-socket platform. It has a 4x 2700W redundant power supply option and up to 11x PCIe slots including the two OCP NIC 3.0 slots.
• The Inspur NF5180G7 is the company’s mainstream 1U server.
  • We reviewed the previous generation Inspur NF5180M6 and NF5180M5 generations and saw the AMD G7 at SC22.

• The NF5280G7 is the company’s mainstream 2U server.
  • We reviewed the Inspur NF5280M6, and NF5280M5 generations.

• The i24G7 is Inspur’s 2U 4-node (2U4N) offering that is designed for maximum compute density with up to 8x Xeon CPUs in four nodes in only 2U of rack space. There is also a version of this called the i24G7-1S that uses single-socket nodes.
  • We recently reviewed the Inspur i24M6, which is the previous generation of this server, as well as the Inspur i24 generation before that.

• The Inspur NF5266G7 is the company’s 4U 46x 3.5″ bay storage server that is designed for GPUs. This is for organizations that want more density than standard 24-bay designs. In this generation, the new server can support up to two dual-width GPUs as well.
• The Inspur ORS3000S centralizes functions like power supplies in a more open compute form factor. It can also handle liquid cooling.
• The Inspur NF5468G7 is a 4U GPU compute server.
  • We have not reviewed this line since the M5 generation, as seen in our Inspur NF5468M5 review.
• The NF5688G7 is the company’s 6U AI training platform supporting up to 8x 700W GPUs.
  • This is the successor to the Inspur NF5488A5 and NF5488M5 servers that we reviewed. With this generation, the chassis needed to add 2U in order to cool the higher-power GPUs.

Final Words

Inspur has a number of new servers for the new Intel Xeon Scalable line. The portfolio spans several different form factors from general-purpose servers to highly specialized servers optimized for scale-up, storage, AI, density, and other applications. Hopefully, we will get to start reviewing these soon. In the meantime, if you want to learn more about the new series, check out our Sapphire Rapids launch video here: