Graphcore Intelligence Processing Unit (IPU)

A New Compute Workload, designed for machine intelligance

More Image(s)

The Intelligence Processing Unit (IPU) is completely different from today’s CPU and GPU processors. It is a highly flexible, easy to use, parallel processor that has been designed from the ground up to deliver state of the art performance on current machine intelligence models for both training and inference. But more importantly, the IPU has been designed to allow new and emerging machine intelligence workloads to be realized.

Quickspecs (Download PDF)

Availability: In stock

Details

Graphcore IPU

Graphcore is an Intelligent Processing Unit (IPU), which is designed specifically to run machine learning workloads.

Graphcore systems will plug into traditional X86 servers via PCIe interfaces, have more than 100x the memory bandwidth than scalar CPUs, and will outperform both CPUs and vector GPU for emerging machine learning workloads for both training and scoring stages. Poplar, a software framework, has been developed, which will abstract the machine learning application development process from the underlying IPU-based hardware. Poplar was written in C++ and will be able to take applications written in other frameworks, like TensorFlow and MXNet, and compile them into optimized code to execute on IPU-boosted hardware. It will feature C++ and Python interfaces. All modern machine learning frameworks like TensorFlow, MxNet, Caffe, Theano, and Torch use the concept of a computational graph as an abstraction..

Arithmetic efficiency

The IPU delivers much better arithmetic efficiency on small batch sizes for both training and inference.

This results in faster model convergence in training, models that generalise better, the ability to parallelize over many more IPU processors to reduce training time for a given batch size, and also delivers much higher throughput at lower latencies for inference.

Available as an appliance

The Dell DSS8440, the first Graphcore IPU server, is now available to order from Dell.

The Graphcore IPU is unique in keeping the entire machine learning knowledge model inside the processor. With 16 IPU processors, all connected with IPU-Link™ technology in a server, an IPU system has over 100,000 completely independent programs, all working in parallel on the machine intelligence knowledge model.p>

Part No.	Graphcore-IPU
Manufacturer	Graphcore
End of Life?	Yes
Form Factor	PCIe Full Height/Length
Maximum Power Consumption	250 W
Thermal Solution	Passive
ECC Protection	No
GPU Memory	32GB HBM2
Memory Interface	ECC
Memory Bandwidth	1134 GB/sec
PCI Slot(s)	PCIe Gen3

Dell EMC Graphcore IPU, a revolutionary compute platform.

Intelligence Processing Unit (IPU)