HPE gives three completely different blades for the GX5000. First is the GX440n Accelerated Blade for mixed-precision computing, with 4 Nvidia Vera CPUs and eight Nvidia Rubin GPUs. As much as 24 of those blades will be configured per compute rack, which interprets to 192 Rubin GPUs per rack. Vera Rubin is because of ship late subsequent yr.
Second is the GX350a Accelerated Blade, designed for purchasers who need a common compute engine for mixed-precision computing with AMD CPUs and GPUs. The blade consists of one subsequent era “Venice” CPU and 4 AMD Intuition MI430X GPUs. As much as 28 of those blades will be configured per compute rack, offering as much as 112 AMD MI430X GPUs per rack.
Lastly, there’s the GX250 Compute Blade for purchasers who need a CPU-only partition for double-precision workloads with eight next-generation Venice CPUs per rack. As much as 40 of these blades will be configured per compute rack, delivering industry-leading flagship x86 CPU core density.
