We take cryptocurrency — so the coins you hold can buy real hardware.Pay in crypto, spend it on real hardware. Free insured shipping worldwide over $399. Plain, unbranded boxes, sent from Germany.Free insured shipping over $399. Your rate is locked for 30 minutes once checkout opens.Rate locked 30 minutes. Every item sealed, serial-checked and covered by a 24-month warranty.Sealed, 24-month warranty.

Saved items Sign in

GIGABYTE

GIGABYTE 25FC9-RTX003-P2R RTX PRO 6000 Blackwell Max-Q 96GB GDDR7 ai accelerator card

SKU CH-58CH-SYM79 MPN 25FC9-RTX003-P2R

AI accelerator card with 96 GB GDDR7 ECC memory, 24,064 CUDA cores, four DisplayPort 2.1 outputs and a 300 W PCIe 5.0 x16 interface

  • Graphics processorRTX PRO 6000 Blackwell Max-Q
  • Memory capacity96 GB
  • Interface typePCI Express 5.0 x16
  • CUDA core count24,064
  • Type of memoryGDDR7 with ECC
  • Chipset makerNVIDIA
  • 96 GB GDDR7 memory with ECC handles massive datasets
  • 24,064 CUDA cores accelerate AI and simulation workloads
  • Blackwell architecture delivers next-generation professional performance
  • PCI Express 5.0 x16 interface provides high-speed data throughput
  • Four DisplayPort 2.1 outputs drive multi-monitor visualisation setups

AI Accelerator Card 96GB GDDR7

The GIGABYTE 25FC9-RTX003-P2R is an NVIDIA RTX PRO 6000 Blackwell Max-Q AI accelerator card built for data science, AI training, and complex visualization workloads. It delivers 24,064 CUDA cores and 96GB of GDDR7 memory with ECC for accuracy at scale. Professionals handling massive datasets or multi-GPU deployments will find the memory capacity and compute density suited to mission-critical pipelines.

PCI Express 5.0 x16 Interface 300W Power

The card installs into a PCI Express 5.0 x16 slot and draws up to 300 W, so the platform must supply adequate power and cooling for sustained workloads. Four DisplayPort 2.1 outputs drive multiple high-resolution displays without extra adapters. RTX PRO SYNC compatibility lets the GPU synchronize with supported displays and external sync sources for frame-accurate multi-GPU visualization.

96GB GDDR7 ECC Memory Headroom

The 96GB GDDR7 frame buffer with ECC accommodates large language models, high-resolution simulation data, and multi-application GPU sharing without paging to system memory. Workloads exceeding 96GB or requiring higher double-precision throughput will need additional cards or a different accelerator class. Buyers who need more memory per GPU or lower power envelopes should evaluate alternative configurations.

How it compares

Measured against everything we stock in GPUs for AI & Data Centers, at today’s prices. Recalculated every night.

Price $11,080.99
Pricier than about 77% of the 36 comparable items (median $5,369.99).
Memory capacity 96 GB
Higher than about 80% of the 12 comparable items (median 79 GB).

New to this? Buying server hardware with crypto — what to check, what it costs to pay, and what is never asked for.

Specifications

Product
Brand namePNY
Model nameRTX PRO 6000 Blackwell Max-Q
Part numberVCNRTXPRO6000BQ-PB
Memory
Type of memoryGDDR7 with ECC
Graphics
Graphics processorRTX PRO 6000 Blackwell Max-Q
Memory capacity96 GB
CUDA core count24,064
Chipset makerNVIDIA
Ports & slots
Interface typePCI Express 5.0 x16
DisplayPort connectors4 x Display Port 2.1
Size & build
Shipping weight1.52 kg
Package size358 × 216 × 76 mm

Questions about this item

When does the 96 GB GDDR7 memory capacity become the limiting factor during sustained workloads?

The 96 GB GDDR7 with ECC becomes the limit when datasets, simulations or models exceed that capacity, forcing data to spill to system memory and reducing throughput.

How much heat and power does the 300 W design produce when running continuously?

The card draws up to 300 W under full load, so sustained operation requires chassis airflow and cooling capable of dissipating that heat continuously.

Which motherboard slot and power connectors must already be present for this PCI Express 5.0 x16 card?

A PCI Express 5.0 x16 slot and sufficient auxiliary power connectors for a 300 W board are required before installation.

More in GPUs for AI & Data Centers

Similar items