TitanMIMO-4D Distributed Processing Option

Kermode-XV6
Kermode-XV6 Distributed FPGA Cluster Block Diagram

The aggregation of all 100×100 channels to a common central processing unit is a must for the initial computational steps of Massive MIMO algorithms. Implementing a Massive MIMO waveform on a set of 100x channels without actual knowledge of the total FPGA resources needed can add significant risk to a fixed, single FPGA baseband processing testbed design.

To give developers a growth path beyond a single central baseband processing element Nutaq has developed the TitanMIMO-4D as a simple extension to the capabilities of the TitanMIMO-4S.

The TitanMIMO-4D is a distributed central baseband processing version where the single FPGA processing of the Perseus 6113 within the TitanMIMO-4S is replaced by a powerful 8xVirtex-6 SX475T FPGA farm, housed on the Kermode-XV6. This is the most powerful AdvancedTCA computing blade ever built, specifically designed to tackle the most demanding signal processing applications.

The Kermode-XV6 is comprised of eight Xilinx Virtex-6 SX475T FPGAs, delivering an outstanding 8.8 TeraMACs solely from their DSP48 dedicated multiply-accumulate engines. Each FPGA interfaces with two DDR-3 SDRAM SODIMM modules, capable of supporting up to 4 GBytes, for an aggregate memory capacity of 64 GBytes.

The Kermode-XV6 supports a mesh ATCA backplane zone 2 and zone 3, with an advanced mesh topology that enables a unique very large virtual FPGA targeting.