Granite Code 20B

ibm-granite/granite-20b-code-instruct

Parameters

20B

Context

128K

Downloads

9.9K

Likes

44

Architecture

transformer

View on HuggingFace

GPU Requirements

VRAM / RAM / disk estimates by quantization and usage scenario

QuantizationMinimumRecommendedProduction
FP1616bit
52.1 GB
RAM 78.1 · Disk 97.1
70.0 GB
RAM 105.0 · Disk 123.1
169.0 GB
RAM 253.5 · Disk 162.1
FP88bit
28.5 GB
RAM 42.8 · Disk 68.1
45.3 GB
RAM 68.0 · Disk 94.1
143.3 GB
RAM 215.0 · Disk 133.1
INT88bit
28.5 GB
RAM 42.8 · Disk 68.1
45.3 GB
RAM 68.0 · Disk 94.1
143.3 GB
RAM 215.0 · Disk 133.1
AWQ4bit
17.2 GB
RAM 32.0 · Disk 54.1
33.5 GB
RAM 50.2 · Disk 80.1
130.9 GB
RAM 196.4 · Disk 119.1
GPTQ4bit
17.3 GB
RAM 32.0 · Disk 54.3
33.6 GB
RAM 50.5 · Disk 80.3
131.1 GB
RAM 196.6 · Disk 119.3
GGUF4bit
17.6 GB
RAM 32.0 · Disk 54.6
33.9 GB
RAM 50.9 · Disk 80.6
131.4 GB
RAM 197.1 · Disk 119.6

GPU Recommendations

Best ValueBEST

1× H100 SXM

80 GB VRAM

$1,170/mo

≈ 346.8 tok/s

Effective production throughput ≈ 242.8 tok/s

score 0.6012

Deploy on vast ↗
Cheapest

4× Tesla V100

64 GB VRAM

$81.2/mo

≈ 0 tok/s (N/A — benchmark unavailable)

Effective production throughput ≈ 0 tok/s

score 0.5378

Deploy on vast ↗
Performance

8× H200 SXM

1128 GB VRAM

$20,440/mo

≈ 1590.1 tok/s

Effective production throughput ≈ 1113.1 tok/s

score 0.2197

Min Complexity

1× RTX A6000

48 GB VRAM

$210/mo

≈ 79.5 tok/s

Effective production throughput ≈ 55.7 tok/s

score 0.6771

Deploy on vast ↗

💬 Community — real deployment experience

0 Articles · 0 Benchmarks

View Community →

GPUs that fit (single card)

RTX A6000

48 GB VRAM

$210

/mo · vast

L40S

48 GB VRAM

$341

/mo · vast

RTX PRO 6000 WS

48 GB VRAM

$482

/mo · vast

RTX PRO 6000 S

48 GB VRAM

$633

/mo · vast

A100 80GB

80 GB VRAM

$876

/mo · vast

AmciHub

AI Model Deployment & Compute Intelligence — analyze AI model requirements, GPU performance, cloud pricing and deployment costs to find the right deployment solution.

© 2026 AmciHub. All rights reserved. Model → Requirement → GPU → Benchmark → Cloud → Cost → Recommendation