Yi 34B 200K RPMerge GPTQ — Hardware Requirements & GPU Compatibility
Chat
Yi 34B 200K RPMerge GPTQ is a 34B-parameter open language model from LoneStriker in the Yi family. It supports a context window of up to 200,000 tokens. At FP16 it needs about 68.80 GB of VRAM — see which GPUs and Macs can run it below.
11 downloads
4 likesSpecifications
- Publisher
- LoneStriker
- Family
- Yi
- Parameters
- 34B
- Architecture
- LlamaForCausalLM
- Context Length
- 200,000 tokens
- Vocabulary Size
- 64,002
- Release Date
- 2024-02-13
- License
- Other
Get Started
HuggingFace
How Much VRAM Does Yi 34B 200K RPMerge GPTQ Need?
Select a quantization to see compatible GPUs below.
| Quantization | Bits | VRAM | + Context | File Size | Quality |
|---|---|---|---|---|---|
| FP16 | 16.00 | 68.8 GB | 117.5 GB | 68.00 GB | Full half-precision — baseline for inference |
Which GPUs Can Run Yi 34B 200K RPMerge GPTQ?
FP16 · 68.8 GBYi 34B 200K RPMerge GPTQ (FP16) requires 68.8 GB of VRAM to load the model weights. For comfortable inference with headroom for KV cache and system overhead, 90+ GB is recommended. Using the full 200K context window can add up to 48.7 GB, bringing total usage to 117.5 GB. No single GPU has enough memory — multi-GPU or cluster setups are needed.
Which Devices Can Run Yi 34B 200K RPMerge GPTQ?
FP16 · 68.8 GB5 devices with unified memory can run Yi 34B 200K RPMerge GPTQ, including NVIDIA DGX H100, NVIDIA DGX A100 640GB.
Related Models
Frequently Asked Questions
- How much VRAM does Yi 34B 200K RPMerge GPTQ need?
Yi 34B 200K RPMerge GPTQ requires 68.8 GB of VRAM at FP16. Full 200K context adds up to 48.7 GB (117.5 GB total).
VRAM = Weights + KV Cache + Overhead
Weights = 34B × 16 bits ÷ 8 = 68 GB
KV Cache + Overhead ≈ 0.8 GB (at 2K context + ~0.3 GB framework)
KV Cache + Overhead ≈ 49.5 GB (at full 200K context)
VRAM usage by quantization
FP1668.8 GBFP16 + full context117.5 GB- Can NVIDIA GeForce RTX 5090 run Yi 34B 200K RPMerge GPTQ?
No — Yi 34B 200K RPMerge GPTQ requires at least 68.8 GB at FP16, which exceeds the NVIDIA GeForce RTX 5090's 32 GB of VRAM.
- Can I run Yi 34B 200K RPMerge GPTQ on a Mac?
Yi 34B 200K RPMerge GPTQ requires at least 68.8 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 Yi 34B 200K RPMerge GPTQ locally?
Yes — Yi 34B 200K RPMerge GPTQ can run locally on consumer hardware. At FP16 quantization it needs 68.8 GB of VRAM. Popular tools include Ollama, LM Studio, and llama.cpp.
- How fast is Yi 34B 200K RPMerge GPTQ?
At FP16, Yi 34B 200K RPMerge GPTQ can reach ~42 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 ÷ 68.8 × 0.55 = ~42 tok/s
Estimated speed at FP16 (68.8 GB)
~42 tok/s~32 tok/s~26 tok/sReal-world results typically within ±20%. Speed depends on batch size, quantization kernel, and software stack.
- What's the download size of Yi 34B 200K RPMerge GPTQ?
At FP16, the download is about 68.00 GB.
- Which GPUs can run Yi 34B 200K RPMerge GPTQ?
No single consumer GPU has enough VRAM to run Yi 34B 200K RPMerge GPTQ at FP16 (68.8 GB). Multi-GPU or professional hardware is required.
- Which devices can run Yi 34B 200K RPMerge GPTQ?
5 devices with unified memory can run Yi 34B 200K RPMerge GPTQ at FP16 (68.8 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.