# Dynamics GP AI Assistant Sizing and Value Calculator

> Source: https://netray.co/tools/dynamics-gp-ai-assistant-sizing-and-value-calculator
> Publisher: Netray (netray.co), AI and integration engineering for ERP and manufacturing systems.
> Section: Free Tools

## Short answer

Size a private AI assistant for Dynamics GP: GPUs, index and monthly value from users and questions per day.

## Category

on-prem-ai

## Erp

Microsoft Dynamics GP

## Audience

- CIO
- IT manager
- Controller
- GP administrator

## Tool

**Kind:** calculator

**Currency:** $

**Inputs:**

#### Named GP users

**Key:** users

**Type:** number

**Default:** 60

**Min:** 1

**Step:** 5

**Unit:** users

#### Share active on a given day

**Key:** dau_pct

**Type:** slider

**Default:** 40

**Min:** 5

**Max:** 100

**Step:** 5

**Unit:** %

#### Questions per active user per day

**Key:** q_per_user

**Type:** number

**Default:** 7

**Min:** 1

**Step:** 1

**Unit:** questions

#### Prompt plus answer tokens per question

**Key:** tokens_per_q

**Type:** number

**Default:** 2500

**Min:** 200

**Step:** 100

**Unit:** tokens

#### Documents to index

**Key:** docs

**Type:** number

**Default:** 40000

**Min:** 0

**Step:** 5000

**Unit:** documents

#### Average pages per document

**Key:** pages

**Type:** number

**Default:** 2

**Min:** 1

**Step:** 1

**Unit:** pages

#### Minutes saved per question

**Key:** mins_saved

**Type:** number

**Default:** 6

**Min:** 0

**Step:** 1

**Unit:** min

#### Loaded hourly cost

**Key:** hourly

**Type:** number

**Default:** 40

**Min:** 0

**Step:** 1

**Unit:** per hour

#### GPU monthly cost

**Key:** gpu_month

**Type:** number

**Default:** 1500

**Min:** 0

**Step:** 100

**Unit:** per GPU

#### Sustained tokens per second per GPU

**Key:** tps_gpu

**Type:** number

**Default:** 900

**Min:** 100

**Step:** 100

**Unit:** tokens/s

**Help:** Lower default for a 24 GB class card running a mid-size quantised model. Measure yours.

#### One-time build cost

**Key:** build_cost

**Type:** number

**Default:** 35000

**Min:** 0

**Step:** 5000

**Unit:** one-time

**Outputs:**

#### Daily active users

**Key:** active

**Expr:** users * dau_pct / 100

**Format:** decimal

#### Questions per day

**Key:** q_day

**Expr:** active * q_per_user

**Format:** integer

#### Tokens per day

**Key:** tokens_day

**Expr:** q_day * tokens_per_q

**Format:** integer

#### Peak concurrent requests

**Key:** peak_conc

**Expr:** q_day * 0.2 / 3600 * 20

**Format:** decimal

**Description:** Assumes 20% of questions in the busiest hour and 20 seconds per request.

#### GPUs needed (before redundancy)

**Key:** gpus

**Expr:** max(1, ceil(q_day * 0.2 * tokens_per_q / 3600 / tps_gpu))

**Format:** integer

**Primary:** true

#### Retrieval index size

**Key:** index_gb

**Expr:** docs * pages * 2 * 4096 * 1.5 / 1000000000

**Format:** decimal

#### Hours saved per month

**Key:** hours_month

**Expr:** q_day * 22 * mins_saved / 60

**Format:** hours

#### Monthly value of time saved

**Key:** value

**Expr:** hours_month * hourly

**Format:** currency

#### Monthly net value after GPU cost

**Key:** net

**Expr:** value - gpus * gpu_month

**Format:** currency

**Primary:** true

#### Simple payback

**Key:** payback

**Expr:** if(gt(net, 0), build_cost / net, 0)

**Format:** months

**Chart:**

#### Peak concurrent requests vs named users

**Type:** line

**X:** users

**From:** 10

**To:** 500

**Steps:** 25

**Series:**

- peak_conc

**Disclaimer:** Estimate for planning. Measure real throughput, include redundancy and confirm time savings with a pilot before committing budget.

## Content

**Intro:** Small and mid-size GP sites rarely need large GPU clusters. This calculator shows what a private AI assistant needs for your user count and question volume, and what time saved is worth against the cost.

**Sections:**

#### How to use it

**Body:** Enter named GP users and the portion who will use the assistant daily; finance teams often have higher usage than the rest. Count questions per user based on current requests for reports and lookups. Set tokens per question higher if you plan to summarise documents. Use a measured throughput from your own GPU.

**Bullets:**

- Finance users usually ask more
- Keep minutes saved conservative
- Compare one and two GPU scenarios
- Include the build cost for payback

#### The formulas

**Body:** Questions per day equal active users times questions each. Peak concurrency equals 20% of daily questions divided by 3600 and multiplied by 20 seconds of processing per request, an application of Little's law. GPUs equal peak tokens per second divided by GPU throughput, rounded up. Monthly value equals hours saved times hourly cost with 22 working days.

#### Reading the result

**Body:** At GP scale the GPU count is often one, and concurrency is a handful of requests. That means latency and quality matter more than raw capacity. If value is marginal, test whether SmartList, SSRS or Power BI already covers the highest volume questions before building anything.

**Cta:** If you want a second pair of eyes on these numbers for your GP estate, Netray can review the inputs with your IT team.

**Faq:**

#### Why show peak concurrency?

**Answer:** It tells you how many requests overlap at the busiest moment, which drives latency. Even small teams can see a handful of overlapping requests.

#### Is a 24 GB GPU enough?

**Answer:** For a small team and a quantised mid-size model, often yes. Quality needs and context length decide, so run your own test questions.

#### What does the chart show?

**Answer:** Peak concurrent requests as named users grow, with other inputs fixed. Use it to see when you might need a second GPU.

#### Does it include power and cooling?

**Answer:** Only if you include them in the GPU monthly cost. Enter an all-in amortised figure for a fair comparison.

## Search terms

Dynamics GP AI sizing, GP AI assistant ROI, private LLM sizing small team, GPU requirement for 50 users LLM, on-prem AI payback, LLM peak concurrency calculator

---
Talk to an engineer: https://netray.co/contact?intent=erp_ai | Machine catalog: https://netray.co/agent/catalog.json