Persönliche Beratung

+49 421 40 966 0 oder per Mail

SCROLL TO PRODUCTS

Produkt Filter Produkt Filter

devCUBE™

sysGen devCUBE™ – Deep Learning ready to go

High end deep learning power with up to 4x NVIDIA® GTX 1080 GPUs or Pascal Titan X and completely configured and installed software package

Deep Learning is one of the fastest growing and maybe the most exciting area in computing. Advanced neural networks and learning mechanism need absolute high end performance, but not everyone is interested in a loud and big server-rack system. The devCUBE is one of the best deskside solution for deep learning. As a complete system with all required software preinstalled you just need to plug in and start you learning network

With latest hardware and pre-installed software options the devCUBE stands for an easy entrance and ready-to-go solution for deep learning.



Key Features:

Up to 4 NVIDIA Geforce GTX 1080 or Titan X with Pascal Chipset

NVIDIA’s new flagship GeForce GTX 1080 is one of the most advanced graphics card ever created. Discover unprecedented performance and power efficiency driven by the new NVIDIA Pascal™ architecture.

  • 2560 CUDA® Cores
  • 8 GB GDDR5X
  • 320 GB/sec Memory Bandwidth
  • 1733MHz Boost Clock


Best prepared for Deep Learning

To enable the best possible start in the world of deep learning, we configure and install all relevant software. For an easy user-interface we use Ubuntu as OS. Next to all the frameworks you need, we pre-install and configure the whole NVIDIA package for deep learning and high performance computing. Don't waste time with the software, start researching now.

  • Ubuntu OS
  • NVIDIA DIGITS – Interactive Deep Learning GPU Training System
  • NVIDIA CUDA Deep Learning Neural Network library (cuDNN)*
  • NVIDIA CUDA Toolkit
  • Deep learning frameworks – e.g. Caffe, Torch, Theano and BIDMach

*Before sysGen can install cuDNN , you have to successfully register on the NVIDIA website as a system developer for "Deep Learning" and sign the NVIDIA "CUDA Software Distribution Agreement for Toolkit cuDNN".



High End Hardware

GPUs alone don't make a workstation. Newest Intel® Core™ i7 CPUs with up to 40 PCI-Lanes, 128GB ultra-fast DDR4 memory and modern M.2 socket for fast NVMe SSDs are just a few highlights of the devCUBE.

 
  • Intel® Core™ X Processor with 44 PCI-Lanes for Multi-GPU support
  • #
  • Up to 128GB DDR4 ECC 2666MHz memory (UDIMM)
  • Modern X299 Mainboard with true 4x PCIe x16 Support
  • Up to 8x hot swap SSD drives and one ultra fast M.2. Socket for NVMe SSDs
  • Titanium quality 1500 Watt PSU in a high-end desktop-case

Datasheet
X99-E Mainboard

sysGen Workstation devCUBE Workstation, consisting of:
Mainboard X99-E WS / Carbide Air 540 Desk-side housing with 1500 Watt 80+ Titanium Power Supply


If you choose TitanV or XP cards, we install the following Software, completely configured, Ready to Go:
  • NVIDIA DIGITS, providing powerful design, training, and visualization of deep neural networks for image classification
  • Pre-installed Ubuntu LTS 16.04. LTS with Caffe, Torch, Theano, BIDMach, cuDNN and CUDA *

System Expansion:
  1. Desk-Side Workstation enclosure with 1500W power supply in Titanium Quality
  2. Mainboard X99-E WS with 4x PCIe x16
  3. Core™ i7 or Xeon E5 16xx/26xx Processor Socket LGA2011-v3
  4. 128GB DDR4 Memory (8x 16GB DDR4 2400 MHz UDIMM or ECC UDIMM/RDIMM, 288-Pin)
  5. Up to four GeForce GTX 1080 Ti GPUs
  6. There are a lot of other cards you can choose from
  7. Storage-bays:
    1. Internal (swap-Frame): 2x 3.5" HDD and 4x 2.5" HDD / SSD drives
    2. External (hot swap): 8x 2.5" HDD / SSD drives or 3x 3.5" HDD drives
    3. or
    4. External (hot swap): 4x 2.5" HDD / SSD drives and optional 1x 3.5" HDD drives

*Pre-requisite: Before sysGen can install cuDNN v2, you have to successfully register on the NVIDIA website as a system developer for "Deep Learning" and sign the NVIDIA "CUDA Software Distribution Agreement for Toolkit cuDNN".
/ Overview Deep-Learning
€ 1.454,39 (€ 1.730,72 inkl. MwSt.)
Configure