Midnight Rose 70B v1.0 — Hardware Requirements & GPU Compatibility
Chat
Midnight Rose 70B v1.0 is a 69.0B-parameter open language model from sophosympatheia. It supports a context window of up to 4,096 tokens. At FP16 it needs about 138.92 GB of VRAM — see which GPUs and Macs can run it below.
198 downloads
30 likesSpecifications
- Publisher
- sophosympatheia
- Parameters
- 69.0B
- Architecture
- LlamaForCausalLM
- Context Length
- 4,096 tokens
- Vocabulary Size
- 32,000
- Release Date
- 2024-02-11
- License
- Llama 2 Community
Get Started
HuggingFace
How Much VRAM Does Midnight Rose 70B v1.0 Need?
Select a quantization to see compatible GPUs below.
| Quantization | Bits | VRAM | + Context | File Size | Quality |
|---|---|---|---|---|---|
| FP16 | 16.00 | 138.9 GB | 139.6 GB | 137.95 GB | Full half-precision — baseline for inference |
Which GPUs Can Run Midnight Rose 70B v1.0?
FP16 · 138.9 GBMidnight Rose 70B v1.0 (FP16) requires 138.9 GB of VRAM to load the model weights. For comfortable inference with headroom for KV cache and system overhead, 181+ GB is recommended. Using the full 4K context window can add up to 0.7 GB, bringing total usage to 139.6 GB. No single GPU has enough memory — multi-GPU or cluster setups are needed.
Which Devices Can Run Midnight Rose 70B v1.0?
FP16 · 138.9 GB4 devices with unified memory can run Midnight Rose 70B v1.0, including NVIDIA DGX H100, NVIDIA DGX A100 640GB, Mac Pro M2 Ultra (192 GB).
Runs great
— Plenty of headroom
Decent
— Enough memory, may be tightRelated Models
Frequently Asked Questions
- How much VRAM does Midnight Rose 70B v1.0 need?
Midnight Rose 70B v1.0 requires 138.9 GB of VRAM at FP16. Full 4K context adds up to 0.7 GB (139.6 GB total).
VRAM = Weights + KV Cache + Overhead
Weights = 69.0B × 16 bits ÷ 8 = 138 GB
KV Cache + Overhead ≈ 0.9 GB (at 2K context + ~0.3 GB framework)
KV Cache + Overhead ≈ 1.6 GB (at full 4K context)
VRAM usage by quantization
FP16138.9 GBFP16 + full context139.6 GB- Can NVIDIA GeForce RTX 5090 run Midnight Rose 70B v1.0?
No — Midnight Rose 70B v1.0 requires at least 138.9 GB at FP16, which exceeds the NVIDIA GeForce RTX 5090's 32 GB of VRAM.
- Can I run Midnight Rose 70B v1.0 on a Mac?
Midnight Rose 70B v1.0 requires at least 138.9 GB at FP16, 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 Midnight Rose 70B v1.0 locally?
Yes — Midnight Rose 70B v1.0 can run locally on consumer hardware. At FP16 quantization it needs 138.9 GB of VRAM. Popular tools include Ollama, LM Studio, and llama.cpp.
- How fast is Midnight Rose 70B v1.0?
At FP16, Midnight Rose 70B v1.0 can reach ~21 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 ÷ 138.9 × 0.55 = ~21 tok/s
Estimated speed at FP16 (138.9 GB)
~21 tok/sReal-world results typically within ±20%. Speed depends on batch size, quantization kernel, and software stack.
- What's the download size of Midnight Rose 70B v1.0?
At FP16, the download is about 137.95 GB.
- Which GPUs can run Midnight Rose 70B v1.0?
No single consumer GPU has enough VRAM to run Midnight Rose 70B v1.0 at FP16 (138.9 GB). Multi-GPU or professional hardware is required.
- Which devices can run Midnight Rose 70B v1.0?
4 devices with unified memory can run Midnight Rose 70B v1.0 at FP16 (138.9 GB), including Mac Pro M2 Ultra (192 GB), Mac Studio M2 Ultra (192 GB), NVIDIA DGX A100 640GB, NVIDIA DGX H100. Apple Silicon Macs use unified memory shared between CPU and GPU, making them well-suited for local LLM inference.