AMD

AMD Instinct MI325X

128GB HBM3 — lowest-cost ROCm inference and AIGC

Reserved and on-demand, self-service access

£1.10/hr starting£800/mo from

Reserved and committed-use discounts up to 60%. All prices GBP, ex-VAT.

Compare AMD GPUs

MI325X is the cost-efficient entry into the AMD Instinct family — a single-GPU ROCm card engineered for low-volume inference, AIGC generation and small-team fine-tuning without the footprint of an 8-way bare-metal node.

Memory
128GB HBM3
1.6 TB/s bandwidth
Architecture
CDNA 3
Family · AMD Instinct
Compute (FP16)
0.65 PFLOPS (8-GPU)
BF16: 0.65 PFLOPS (8-GPU)
Interconnect
100G RoCE (no Infinity Fabric)
TDP · 300W

Highlights

  • 128GB HBM3 fits modern 7B-13B models in a single card
  • Lowest entry price for AMD ROCm inference
  • Mature RCCL + vLLM-ROCm stack

Software stack

ROCm 6.2+
PyTorch 2.x
vLLM-ROCm
RCCL

Best-fit scenarios

  • ROCm low-cost inference
  • AIGC generation
  • Small-model fine-tuning

Pricing by configuration

All available SKUs for this GPU across bare-metal, VM and container form factors. Reserved and committed-use discounts available.

On-demand starting at £1.10/h

ConfigurationGPUsVRAM totalCPUMemoryStorageNetwork£/hr£/mo
Bare Metal — 8× MI325X
mi325x-bm-8
81 TB192 vCPU1 TB DDR58× 1.92 TB NVMe800G RoCE£9.76£7,120
Virtual Machine — 1× MI325X
Popular
mi325x-vm-1
1128GB30 vCPU120 GB DDR5960 GB NVMe10Gbps£1.22£890
Container — 1× MI325X
mi325x-ctr-1
1128GB16 vCPU64 GB500 GB10Gbps£1.10£800

Prices shown are list prices in GBP (ex-VAT). Reserved capacity and committed-use discounts up to 60% available — contact sales for a custom quote.

Technova at a glance

Full-stack AI cloud

GPUs, networking, storage, managed Slurm & Kubernetes on one platform.

Manual UK deployment

Every resource deployed and configured by our UK engineering team.

UK & US data residency

UK workloads stay in Glasgow & London. US workloads in Texas. You choose the region.

Reliable

Historical uptime over 99.9% with sensible SLAs and fair compensation.

Dual-stack CUDA + ROCm

Professional full-stack expertise across both NVIDIA and AMD.

World-class support

Proactive support from ML engineers and infrastructure specialists.

Ready to deploy AMD Instinct MI325X?

Spin up a private cluster in days, or talk to our sales engineers about custom configurations.

Explore AI Cloud