Microsoft is offering new instances on its Azure cloud platform optimised for artificial intelligence (AI) and high-intensity workloads that are powered by chips from AMD, that will allow customers to process larger AI models and vast quantities of data and thus save power and time and reduce costs.
► New Azure VMs are optimised for AI and vast amounts of data
► Each VM is powered by 8X AMD MI300X GPUs
The Azure ND MI300X v5 virtual machine (VM) offers an ‘unprecedented’ 1.5TB of high bandwidth memory (HBM) and is powered by eight AMD Instinct MI300X GPUs, interconnected via AMD’s Infinity Fabric 3.0 in each VM. This will enable customers to save power, reduce costs and accelerate solutions delivery times.
As well as industry-leading HBM capacity and bandwidth, the MI300X offers 192 GB of HBM3 memory per GPU, capable of speeds of up to 5.2 TB/s. With greater capacity and more bandwidth customers can process larger models faster, using fewer GPUs. In addition, these virtual machines use the same winning platform as Microsoft’s other ND-series VMs, which include the following technologies:
- 400 Gb/s NVIDIA Quantum-2 CX7 InfiniBand per GPU with 3.2Tb/s per VM
- 4th Gen Intel Xeon Scalable processors
- PCIE Gen5 host-to-GPU interconnect with 64GB/s bandwidth per GPU
- 16 Channels of DDR5 DIMMs
The ND MI300X v5 will be available as part of the Azure portfolio soon. For more information, contact the TD SYNNEX Azure team by clicking the link below.
