Deep learning is enabling many practical applications of machine or artificial intelligence and has been stunningly effective across many problem domains. In recent years, we have seen amazing successes in natural speech translation, image recognition, medical diagnosis, financial and market prediction, and so much more. Deep learning is powering a renaissance in machine/artificial intelligence and transforming many industries. Drive your deep learning workflows with solutions that deliver:
- State-of-the-art performance
- Flexible configurations
- Cost-effective efficiency
- Expert services
Penguin Computing, with decades of experience in open technologies and high-performance computing, is uniquely positioned to accelerate your most demanding, deep learning workflows.
TundraTM ES – OCP Deep Learning Solutions
Deep Learning with NVIDIA® Acceleration
Features dual Intel® Xeon® CPUs and up to four NVIDIA® Tesla® GPUs in a 1 OU, Open Compute form-factor. This is your high-density, ultra-efficient, GPU-based deep learning solution.
Features dual Intel® Xeon® CPUs with NVIDIA® Tesla® P4 GPU.
Deep Learning with Intel® Acceleration
RelionTM Phi X1030e
Features Intel® Xeon PhiTM “Knights Landing” in a TundraTM compute sled, allowing three servers per 1 OU. The Relion Phi X1030e supplies massive parallel computing architecture ideal for deep neural network training.
Features dual Intel® Xeon® processors in a TundraTM compute sled, supporting high-density scale-out.
Traditional 19-inch EIA – Deep Learning Solutions
Penguin Computing offers many traditional 19-inch EIA rack options, allowing a wide variety of solutions designed around your workflow. Some featured configurations include:
Deep Learning with NVIDIA® or Intel® Acceleration
Features dual Intel® Xeon® processors with up to 8 double-width NVIDIA® Tesla® GPUs or Intel® Xeon PhiTM processors.
Ideal for training deep, convolutional neural networks.
Features dual Intel® Xeon® processors with up to 4 double-width NVIDIA® P40 GPUs or Intel® Xeon PhiTM processors.
Deep Learning with OpenPOWERTM and NVIDIA® Acceleration
Features dual OpenPOWERTM POWER8™ processors with NVIDIA® NVLinkTM interconnect for up to 5x improvement in memory bandwidth available to up to 4 NVIDIA® Tesla® P100 GPUs.
Features dual OpenPOWERTM POWER8™ processors with up to 2 NVIDIA® Tesla® P40 GPUs.