Code Llama 34B Instruct on AMD EPYC 7002 / 7003 series NAS (512 GB)

Can the AMD EPYC 7002 / 7003 series NAS (512 GB) run Code Llama 34B Instruct?

Yes, but slowly

Slow. At Q4_K_M it needs about 60.5 GB of 471 GB usable, with an estimated speed of 4.1 tok/s.

Best quant: Q4_K_M
Generation speed: 4.1 tok/s
Prompt processing: 11.9 tok/s
First token (1K prompt): 86.4 s
Largest safe context: 16K

Best app to run it on the AMD EPYC 7002 / 7003 series NAS (512 GB)

On Linux

Ollama

Command line + local API · runs on llama.cpp (ggml), MLX (Apple silicon, preview)

One command with Ollama’s own build of this model.
Serves an OpenAI-compatible API that other apps and coding tools can use.

ollama run codellama:34b-instruct-q4_K_M

llama.cpp Usually the first to support new model architectures.
KoboldCpp

Want an agent? Code Llama 34B Instruct supports tool use, so you can pair a llama.cpp or MLX LM server with one of these. Apps that run on this computer come first.

Coding agents

Personal assistants

How we choose: the files published for this model, which apps run in a container on a NAS without a graphics card (models run on its processor), and whether the model fits fully in the memory models can use. Details.

Download Code Llama 34B Instruct

Recommended for the AMD EPYC 7002 / 7003 series NAS (512 GB): Q4_K_M. Every other quantization has its own download link in the table below.

Quickest · Ollama

One command downloads a ready-to-run version and starts a chat.

Get Ollama at ollama.com/download (opens in a new tab) · model page (opens in a new tab)

ollama run codellama:34b-instruct-q4_K_M

Quantized files · GGUF

Works in LM Studio, llama.cpp, Jan and Ollama. No sign-in needed.

Download the Q4_K_M file (opens in a new tab)

huggingface.co/TheBloke/CodeLlama-34B-Instruct-GGUF (opens in a new tab)

ollama run hf.co/TheBloke/CodeLlama-34B-Instruct-GGUF:Q4_K_M

Original weights · Hugging Face

Full-precision files from Meta, for fine-tuning, vLLM, Transformers.

huggingface.co/meta-llama/CodeLlama-34b-Instruct-hf (opens in a new tab)

Free Hugging Face account needed: request access on the model page, and Meta approves it (usually within minutes to a few hours).

Official repo · GitHub

Reference code, inference scripts and docs from Meta. The weights themselves download from Hugging Face.

github.com/meta-llama/codellama (opens in a new tab)

Meta’s official Code Llama repo.

git clone https://github.com/meta-llama/codellama

Links checked 1 day ago against Hugging Face, Ollama and GitHub.

How it’s rated

Quality 36Basic quality (D) · 36 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 49/100 (quality, fit, speed and memory use) sets the ranking.: How capable the model itself is at Q4_K_M, from its editorial quality tier scaled by the quantization or precision. It’s the same on any hardware that runs this variant. Scale: A+ 85+ · A 70+ · B 55+ · C 40+ · D under 40.
Runs Slow: How well it runs on the AMD EPYC 7002 / 7003 series NAS (512 GB): speed, capped by how well it fits in memory. Runs great: 30+ tokens/s (or near the image, video or decision speed target) · Runs well: 10+ tokens/s · Slow: 3+ tokens/s · Very slow: under 3 · Won’t run: doesn’t fit even with offload. A tight fit or RAM offload caps it at Runs well; heavy offload caps it at Slow.

Match score

To order lists, each result also gets a match score out of 100 from four parts. It favors the most capable model that still runs well.

Model qualityEditorial quality tier of the model, scaled by the quantization or precision used.	16.2 / 45
Memory fitFull marks when the model fits in the share of the NAS’s memory that models can use (the rest stays with the NAS system and its storage services); fewer when it goes past that limit.	25.0 / 25
SpeedGeneration speed against a 40 tokens/s target for LLMs, or render time against a target for image and video.	4.8 / 15
Memory useRewards using the memory you have, so a bigger, better variant beats a tiny one that leaves most memory idle.	3.0 / 15
Match score	49.0 / 100

The match score orders results; it is not shown as a grade. Quality and Runs are the two ratings in the tables.

Every quantization on the AMD EPYC 7002 / 7003 series NAS (512 GB)

Quant	Memory needed	Available	Speed	Runs	Quality	Download
`Q2_K`	53.4 GB	471 GB	6.8 tok/s	Slow	26Basic quality (D) · 26 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 45/100 (quality, fit, speed and memory use) sets the ranking.	GGUF (opens in a new tab)
`Q3_K_M`	56.4 GB	471 GB	5.4 tok/s	Slow	31Basic quality (D) · 31 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 47/100 (quality, fit, speed and memory use) sets the ranking.	GGUF (opens in a new tab)
`Q4_K_M`	60.5 GB	471 GB	4.1 tok/s	Slow	36Basic quality (D) · 36 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 49/100 (quality, fit, speed and memory use) sets the ranking.	GGUF (opens in a new tab)
`Q5_K_M`	63.8 GB	471 GB	3.4 tok/s	Slow	36Basic quality (D) · 36 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 49/100 (quality, fit, speed and memory use) sets the ranking.	GGUF (opens in a new tab)
`Q6_K`	67.6 GB	471 GB	2.9 tok/s	Very slow	37Basic quality (D) · 37 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 49/100 (quality, fit, speed and memory use) sets the ranking.	GGUF (opens in a new tab)
`Q8_0`	75.7 GB	471 GB	2.1 tok/s	Very slow	38Basic quality (D) · 38 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 49/100 (quality, fit, speed and memory use) sets the ranking.	GGUF (opens in a new tab)
`F16`	107 GB	471 GB	1.0 tok/s	Very slow	38Basic quality (D) · 38 of 100: how capable the model is at F16, whatever it runs on. Match score 49/100 (quality, fit, speed and memory use) sets the ranking.	Original weights (opens in a new tab)

Memory figures assume 8K of context. Longer conversations or documents grow the KV cache.

Other hardware that runs it

36Basic quality (D) · 36 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. NVIDIA GeForce RTX 5090 Q4_K_M · 65.2 tok/s · Runs great
36Basic quality (D) · 36 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. NVIDIA GeForce RTX 4090 Q4_K_M · 36.7 tok/s · Runs well (tight fit)
31Basic quality (D) · 31 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. NVIDIA GeForce RTX 3090 Q3_K_M · 44.0 tok/s · Runs great
31Basic quality (D) · 31 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. AMD Radeon RX 7900 XTX Q3_K_M · 38.9 tok/s · Runs great
37Basic quality (D) · 37 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. NVIDIA RTX PRO 6000 Blackwell Q6_K · 45.2 tok/s · Runs great
36Basic quality (D) · 36 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. Apple M4 Max (64 GB) Q4_K_M · 17.1 tok/s · Runs well

Estimates come from published hardware specs and our open formulas, not from running the model on your machine. Real speed varies with runtime, drivers and settings. Numbers marked “Measured” come from a linked benchmark.