Quadro RTX 6000 24GB

The Quadro RTX 6000 24GB is a Turing-generation NVIDIA GPU built on the TU102 architecture, manufactured on a TSMC 12nm process node with 18.6 billion transistors. Released in 2018, it delivers 130.5 TFLOPS of FP16 throughput and 130.5 TFLOPS BF16 — the two precision formats most commonly used for transformer model training and inference. The TU102 architecture represents NVIDIA's approach to balancing compute throughput, memory bandwidth, and power efficiency for data center AI workloads. At 295W TDP, the Quadro RTX 6000 24GB sits in the mid-range data center tier (295W), fitting standard GPU server form factors.

Memory capacity is 24GB of GDDR6 with 672 GB/s bandwidth. This determines which models can run without quantization: approximately 12B parameters at FP16 (2 bytes/param), 24B at INT8 (1 byte/param), or up to 48B parameters at INT4/GGUF quantization (0.5 bytes/param). These figures are theoretical maximums — actual capacity is reduced by KV cache, framework overhead, and activation memory, typically by 10–20% for inference and 30–40% for training. The arithmetic intensity ceiling is approximately 194 FLOP/byte (130.5 TFLOPS ÷ 672 GB/s). Most autoregressive LLM inference falls well below this threshold, making the 672 GB/s memory bandwidth the binding constraint on tokens-per-second throughput rather than raw TFLOPS.

Multi-GPU configurations are a first-class use case for the Quadro RTX 6000 24GB. NVLink 2.0 / PCIe 3.0 provides 100 GB/s of bidirectional NVLink bandwidth between cards, enabling tensor-parallel inference across multiple GPUs with near-linear VRAM scaling. A two-card configuration provides 48GB of pooled VRAM — sufficient for 24B parameter models at FP16 — while a four-card setup reaches 96GB. NVLink's low-latency, high-bandwidth fabric makes all-reduce operations in data-parallel training significantly faster than PCIe-based alternatives, which top out at ~64 GB/s bidirectional for PCIe 5.0 x16.

The primary workloads for the Quadro RTX 6000 24GB are Budget 13B inference, NVLink multi-GPU setups, Legacy professional workloads. High-end Turing professional GPU with NVLink 2.0. 24GB GDDR6. Two cards give 48GB unified. Good for 13B model inference at low cost. Key limitations to factor into your evaluation: Older Turing architecture — no BF16 hardware support; GDDR6 memory bandwidth far below HBM alternatives; Limited cloud availability as legacy hardware. When comparing this GPU against alternatives at similar price points, the most important metrics are memory bandwidth (for inference throughput), VRAM capacity (for model size), and FP16/BF16 TFLOPS (for training speed). Raw TFLOPS figures can be misleading for inference — a GPU with lower TFLOPS but higher memory bandwidth will often outperform a higher-TFLOPS card on tokens-per-second for memory-bound autoregressive generation.

In the broader GPU market, the Quadro RTX 6000 24GB is a 8-year-old design that is increasingly being replaced by newer architectures in cloud deployments, though it remains available at competitive rental rates. CUDA compatibility is a significant advantage: the Quadro RTX 6000 24GB benefits from the full NVIDIA software ecosystem including cuDNN, TensorRT, FlashAttention, and all major inference frameworks (vLLM, TGI, TensorRT-LLM). CUDA's maturity means optimized kernels are available for virtually every model architecture. For cloud rental, availability varies significantly by provider — some specialize in this GPU tier while others may have limited stock. Compare on-demand and spot pricing across providers using the rental comparison table on this page, and factor in region availability if latency is a concern for your inference workload.