San Jose, Calif. March 25, 2013 - Super Micro Computer, Inc. (NASDAQ: SMCI), a global leader in high-performance, high-efficiency server, storage technology and green computing highlights its expansive selection of GPU optimized server solutions at NVIDIA's GPU Technology Conference (GTC) in San Jose, CA this week. Headlining the show will be Supermicro's newest 4U 8x GPU SuperServer® (SYS-4027GR-TR). This unique high-performance server features innovative airflow optimized architecture with independent cooling zones and side-by-side component layout. This eliminates CPU/GPU preheat, enabling maximum performance and reliability of eight 300W GPUs and dual 150W CPUs. Optimal energy efficiency and power savings is also attained with Supermicro's state-of-the-art power delivery sub-systems featuring redundant 1600W Platinum Level high efficiency (94%) power supplies. This system is the platform behind the new NVIDIA® Iray Visual Computing Appliance (VCA). Unveiled today at GTC, Iray VCA delivers unprecedented rendering speed and performance using eight of NVIDIA's most powerful GPUs for extremely photo-realistic computer models of the highest fidelity. Also highlighted at the show is the new 4U FatTwin™ featuring two side-by-side dual-processor nodes each supporting 6x GPUs and 2x additional PCI-E 3.0 (x8) slots for high bandwidth I/O expansion. These systems along with Supermicro's extensive GPU solution range address compute and graphics intensive applications across diverse fields of HPC, scientific discovery, engineering, enterprise virtualization, 3D/CG visual arts, broadcast media/film and cloud-based gaming.
"Supermicro is unrivaled in the industry with the most extensive range of GPU optimized servers solutions on the market," said Charles Liang, President and CEO of Supermicro. "Innovations in our new 4U single node server architecture enables efficient, reliable performance of eight 300W GPUs and dual 150W CPUs in high density compute clusters. Dual node 4U 12x GPU FatTwin systems provide even higher density supporting six 300W GPUs paired with dual 130W CPUs per node. Our GPU solutions offer the highest performance, density, reliability and energy efficiency in their class. Combined with Supermicro's vast range of configurations, we deliver developers the most flexible, cost effective path to highly scalable, parallel computing."
"Supermicro offers HPC customers a range of versatile high-performance, cost-effective computing platforms that support our full line of Kepler architecture-based Tesla GPUs," said Sumit Gupta, general manager of the Tesla Accelerated Computing Business at NVIDIA. "Working with Supermicro, we deliver advanced solutions to accelerate the world's most demanding scientific, engineering, enterprise, and creative applications."
Supermicro has the industry's broadest line of high-performance, application-optimized GPU server solutions that support NVIDIA Kepler™ architecture-based GPUs. These include the NVIDIA Tesla® K10, K20, K20X, K40, and K40X GPU accelerators, and the NVIDIA GRID™ K1 and K2 GPUs. Supermicro GPU platforms also support the new NVIDIA Iray VCA and GRID VDI, GRID VCA and GRID Cloud Gaming servers.
Exhibits at GTC 2014 include:
- 4U SuperServer® (SYS-4027GR-TR / -TRT) 8x GPUs, dual Intel® Xeon® E5-2600 v2 series processors (up to 150W TDP), up to 1.5TB ECC DDR3, 1866MHz memory in 24x DIMMs and 24x (48x option) 2.5" hot-swap SAS2/SATA3 drive bays. 2x 10GbE LAN ports -TRT model.
- 4U FatTwin™ (SYS-F627G3-FT+ / G2-FT+) 12x GPUs (3x per node) in 4x hot-plug nodes each supporting dual Intel® Xeon® E5-2600 v2 series processors (up to 130W TDP), front I/O 2x GbE LAN and 2x 3.5" or 6x 2.5" hot-swap drive bays.
- 4U/Tower (SYS-7047GR-TRF / -TPRF) 5x GPUs (NVIDIA Maximus™ certified), dual Intel® Xeon® E5-2600 v2 series processors (up to 150W TDP), up to 1TB memory and 8x hot-swap 3.5" drive bays.
- 4U Hyper-Speed (SYS-7047AX-TRF / -72RF) 1x GPU solution optimized and balanced for low-latency and pure speed, dual Intel® Xeon® E5-2600 v2 series processors (up to 150W TDP), up to 1TB memory, 8x hot-swap 3.5" drive bays and redundant Platinum Level 1280W high-efficiency (95%) digital switching power supplies.
- 3U SuperServer® (SYS-6037R-72RFT+) 2x GPUs, dual Intel® Xeon® E5-2600 v2 series processors (up to 135W TDP), up to 1.5TB memory and 8x hot-swap 3.5" SAS2 drive bays.
- 2U SuperServer® (SYS-2027GR-TRFH) 6x GPUs, dual Intel® Xeon® E5-2600 v2 series processors (up to 130W TDP), up to 512GB memory and 10x hot-swap 2.5" drive bays.
- 1U SuperServer® (SYS-1027GR-TRF) 3x GPUs, dual Intel® Xeon® E5-2600 v2 series processors (up to 130W TDP), up to 512GB memory and 4x hot-swap 2.5" drive bays.
- 1U SuperServer® (SYS-1027GR-72R2) 3x GPUs, dual Intel® Xeon® E5-2600 v2 series processors (up to 130W TDP), up to 1TB memory and 4x hot-swap 2.5" drive bays.
- 1U SuperServer® (SYS-1017GR-TF) 2x GPUs, single Intel® Xeon® E5-2600 v2 series processors (up to 130W TDP), up to 512GB memory and 6x hot-swap 2.5" drive bays.
- 7U SuperBlade® – The all-in-one SuperBlade features redundant Platinum Level high-efficiency (94%+) power supplies, high speed connectivity through network switch modules, including 56Gb/s FDR IB (SBM-IBS-F3616M), FC/FCoE (SBM-XEM-F8X4SM), 10GbE (SBM-XEM-X10SM) and 1/10GbE (SBM-GEM-X3S+) and centralized remote management software.
- SBI-7127RG3 Blade supports 3x NVIDIA Tesla K40X SXM form factor GPUs, dual Intel® Xeon® E5-2600 v2 series processors, up to 512GB memory and onboard BMC for IPMI 2.0 support. 10x blades in 7U SuperBlade® enclosure scales to best density (180x GPUs and 120x CPUs) and performance (256 TFLOPS theoretical) per 42U rack.
- SBI-7127RG-E Blade supports 2x GPUs, dual Intel® Xeon® E5-2600 v2 series processors, up to 512GB memory, 1x SSD or 1x SATA-DOM, and onboard BMC for IPMI 2.0 support. 10x blades in 7U SuperBlade® enclosure offers high density (120x GPUs and 120x CPUs) and performance (178 TFLOPS theoretical) per 42U rack.
To see the latest GPU server innovations from Supermicro visit Booth #421 at GTC 2014 in the San Jose McEnery Convention Center, March 24-27.