WhiteRabbitNeo 33B V1 — Hardware Requirements & GPU Compatibility
Chat
WhiteRabbitNeo 33B V1 is a 33B-parameter open language model from WhiteRabbitNeo. It supports a context window of up to 16,384 tokens. At BF16 it needs about 66.82 GB of VRAM — see which GPUs and Macs can run it below.
1.4K downloads
93 likesSpecifications
- Publisher
- WhiteRabbitNeo
- Parameters
- 33B
- Architecture
- LlamaForCausalLM
- Context Length
- 16,384 tokens
- Vocabulary Size
- 32,256
- Release Date
- 2024-02-15
- License
- Other
Get Started
HuggingFace
How Much VRAM Does WhiteRabbitNeo 33B V1 Need?
Select a quantization to see compatible GPUs below.
| Quantization | Bits | VRAM | + Context | File Size | Quality |
|---|---|---|---|---|---|
| BF16 | 16.00 | 66.8 GB | 70.5 GB | 66.00 GB | Brain floating point 16 — preferred for training |
Which GPUs Can Run WhiteRabbitNeo 33B V1?
BF16 · 66.8 GBWhiteRabbitNeo 33B V1 (BF16) requires 66.8 GB of VRAM to load the model weights. For comfortable inference with headroom for KV cache and system overhead, 87+ GB is recommended. Using the full 16K context window can add up to 3.6 GB, bringing total usage to 70.5 GB. No single GPU has enough memory — multi-GPU or cluster setups are needed.
Which Devices Can Run WhiteRabbitNeo 33B V1?
BF16 · 66.8 GB5 devices with unified memory can run WhiteRabbitNeo 33B V1, including NVIDIA DGX H100, NVIDIA DGX A100 640GB.
Related Models
Frequently Asked Questions
- How much VRAM does WhiteRabbitNeo 33B V1 need?
WhiteRabbitNeo 33B V1 requires 66.8 GB of VRAM at BF16. Full 16K context adds up to 3.6 GB (70.5 GB total).
VRAM = Weights + KV Cache + Overhead
Weights = 33B × 16 bits ÷ 8 = 66 GB
KV Cache + Overhead ≈ 0.8 GB (at 2K context + ~0.3 GB framework)
KV Cache + Overhead ≈ 4.5 GB (at full 16K context)
VRAM usage by quantization
BF1666.8 GBBF16 + full context70.5 GB- Can NVIDIA GeForce RTX 5090 run WhiteRabbitNeo 33B V1?
No — WhiteRabbitNeo 33B V1 requires at least 66.8 GB at BF16, which exceeds the NVIDIA GeForce RTX 5090's 32 GB of VRAM.
- Can I run WhiteRabbitNeo 33B V1 on a Mac?
WhiteRabbitNeo 33B V1 requires at least 66.8 GB at BF16, which exceeds the unified memory of most consumer Macs. You would need a Mac Studio or Mac Pro with a high-memory configuration.
- Can I run WhiteRabbitNeo 33B V1 locally?
Yes — WhiteRabbitNeo 33B V1 can run locally on consumer hardware. At BF16 quantization it needs 66.8 GB of VRAM. Popular tools include Ollama, LM Studio, and llama.cpp.
- How fast is WhiteRabbitNeo 33B V1?
At BF16, WhiteRabbitNeo 33B V1 can reach ~44 tok/s on AMD Instinct MI300X. Speed depends mainly on GPU memory bandwidth. Real-world results typically within ±20%.
tok/s = (bandwidth GB/s ÷ model GB) × efficiency
Example: AMD Instinct MI300X → 5300 ÷ 66.8 × 0.55 = ~44 tok/s
Estimated speed at BF16 (66.8 GB)
~44 tok/s~33 tok/s~27 tok/sReal-world results typically within ±20%. Speed depends on batch size, quantization kernel, and software stack.
- What's the download size of WhiteRabbitNeo 33B V1?
At BF16, the download is about 66.00 GB.
- Which GPUs can run WhiteRabbitNeo 33B V1?
No single consumer GPU has enough VRAM to run WhiteRabbitNeo 33B V1 at BF16 (66.8 GB). Multi-GPU or professional hardware is required.
- Which devices can run WhiteRabbitNeo 33B V1?
5 devices with unified memory can run WhiteRabbitNeo 33B V1 at BF16 (66.8 GB), including Mac Pro M2 Ultra (192 GB), Mac Studio M2 Ultra (192 GB), Mac Studio M4 Max (128 GB), NVIDIA DGX A100 640GB. Apple Silicon Macs use unified memory shared between CPU and GPU, making them well-suited for local LLM inference.