Supermicro is using this week’s NVIDIA GPU Technology Conference to debut its next-generation GPU server architecture and new GPU-optimized SuperServer, SuperBlade, and SuperWorkstation products. The company says such products are becoming increasingly important in fields such as high-performance computing, deep learning, and big data analysis.
Here’s a quick rundown of the products Supermicro has at the show:
1U 4x GPU SuperServer. Supports up to four NVIDIA Tesla GPU accelerators; dual Intel Xeon processor E5-2600 v4; up to 2TB DDR4 2,400MHz ECC DIMMs; and two hot-swap and two fixed 2.5-inch drive bays.
1U 3x GPU SuperServer. Up to three NIVIDIA Tesla GPU accelerators; dual Intel Xeon processor E5-2600 v4; up to 2TB DDR4 2,400MHz ECC DIMMs; and four hot-swap 2.5-inch SATA3 drive bays.
2U TwinPro 2x GPU SuperServer. Two hot-pluggable systems with each node supporting dual Intel Xeon processor E5-2600 v4; up to 2TB ECC LRDIMM; and 12 2.5-inch hot-swap SAS/SATA hard drive bays.
4U FatTwin 12x GPU SuperServer. Four hot-plug system nodes in 4U with each node supporting dual Intel Xeon processor E5-2600 v4 and up to 2TB ECC DDR4.
4U 8x GPU SuperServer. Supports dual Intel Xeon processor E5-2600 v4; up to 3TB ECC DDR4; and 24 2.5-inch hot-swappable drive bays.
4U/Tower 4+1 GPU SuperServer. Supports 4+1 NVIDIA GPU accelerators; dual Intel Xeon processor E5-2600 v4; up to 2TB ECC DDR4; and eight 3.5-inch hot-swap, three fixed 5.25-inch, and one fixed 3.5-inch drive bays.
7U SuperBlade. Features 120 GPUs and 120 CPUs per 42U rack; up to two NVIDIA GPU accelerators; dual Intel Xeon processor E5-2600 v4; and up to 1TB ECC DDR4.
[source :-processor]