Product overview
Purpose-built accelerator for enterprise AI workloads
Designed for dense inference, model adaptation, and private AI infrastructure where predictable availability, localized supply, and software compatibility are critical.
Architecture
CoreX
Memory & Performance
High-bandwidth memory profile for AI inference and training.
VRAM capacity32
Memory typeHBM2
Memory bandwidth250
Interconnect typeInter-chip Interconnect
Interconnect speed200-400
Architecture
Compute architecture and software execution model.
ArchitectureCoreX
Compute units-
Power & Thermal
Data center integration requirements.
Thermal design power250
CoolingПассивное
Form factorPCIe
Pixel Rate-
Texture Rate-
Benchmarks
Peak theoretical compute for common AI precisions.
FP64
-
FP32
32
FP16
128
TF32
-
BF16 Tensor
128
FP8 Tensor
-
INT8 Tensor
256
Compatibility
Interfaces, frameworks, and deployment environment.
PCIe interfacePCIe 4.0x16
Video encoding-
Video decoding-
Physical Dimensions
Card dimensions for server platforms.
Slots2
Length312 mm
Height111 mm
Width39 mm
Pricing
On request
Volume pricing available for cluster deployments and pilot batches.
Documentation