“AMD Radeon Instinct MI200 accelerator gets 128GB HBM2e”

AMD’s Radeon Instinct MI200 data center accelerator appears to be getting 128GB of HMB2e. That is four times as much as the current MI100. The accelerator is also probably AMD’s first GPU that, like the Ryzen processors, is made up of several chiplets.

Australia’s Pawsey Supercomputing Center is working on a new supercomputer called Setonix that will feature ‘MI-next’ GPUs with 128GB of vram per GPU, writes HPC Wire. AMD has not yet announced its Instinct MI200 card, but more information has already appeared about it.

Twitter user Locuza published last week a diagram of the gpu, which is based on AMD’s CDNA2 architecture. It would be an MCM design, or a Multi-Chip Module. Multiple chips are combined, just like with the Ryzen processors with chiplets containing the CPU cores.

According to the diagram, AMD would use a total of eight stacks of 16GB HBM2e and the MI200 would consist of two chiplets. The accelerator would contain 128 compute units and that could account for 16,384 stream processors, but it’s not yet clear whether all of those have been activated.

The MI200 is the first accelerator based on the CDNA2 architecture, which was previously on the roadmap for 2022. The GPU would be codenamed Aldebaran and is the intended successor to Arcturus, or the Radeon Instinct MI100. AMD’s Instinct accelerators are used in data centers and supercomputers. The accelerators cannot be used as a video card.

AMD Instinct Accelerators

Accelerator name AMD Instinct MI100 AMD Instinct MI200*

Architecture

GPU name

GPU cores

GPU speed

FP16 Compute

FP32 Compute

FP64 Compute

Vram

Memory speed

Memory bus

Memory bandwidth

Form factor

Cooling

Tdp

*Specifications not yet officially confirmed