Calculate the TPS of the FirePro W4300 on local AI models

AMD 4 GB GDDR5 96 GB/s December 2015

Every model in our catalogue assessed against this card at the context length and minimum quality you choose. Speed is an estimate for a single request, calculated from this card's memory bandwidth and the size of each model once compressed.

Calculated for this card

97 models it can run

679 models in our catalogue altogether

Largest model it holds

DeciLM 6B

5.7B · Q3_K_M · 15.0 tok/s

Fastest model

Gemma 3 QAT 1B

31.7 tok/s · 1B

Which AI models can run on a FirePro W4300?

Set the inputs, read the answer

More context means more memory for the conversation cache. Speed is for a fresh conversation and does not change with this setting.

Hides models that would only fit by being compressed below this point.

97 models match

Calculating
Quantisation Fit
31.7 tok/s

19–51 · low confidence

Gemma 3 1B 1B Mar 2025 1.8 GB 33k tokens Q8_0 Comfortable
31.7 tok/s

19–51 · low confidence

Gemma 3 QAT 1B 1B Apr 2025 1.8 GB 33k tokens Q8_0 Comfortable
31.7 tok/s

19–51 · low confidence

HGRN 1B (WT 103) ≈ 1B Nov 2023 1.8 GB 131k tokens ? Q8_0 Comfortable
31.7 tok/s

19–51 · low confidence

LLama 3..2 Typhoon 2 1B ≈ 1B Dec 2024 1.8 GB 131k tokens ? Q8_0 Comfortable
31.7 tok/s

19–51 · low confidence

OLMo-1B ≈ 1B Feb 2024 1.8 GB 131k tokens ? Q8_0 Comfortable
31.7 tok/s

19–51 · low confidence

Pythia-1b ≈ 1B Apr 2023 1.8 GB 131k tokens ? Q8_0 Comfortable
29.4 tok/s

18–47 · low confidence

OpenELM-1.1B ≈ 1.1B May 2024 1.9 GB 131k tokens ? Q8_0 Comfortable
28.8 tok/s

17–46 · low confidence

DeciCoder-1B ≈ 1.1B Aug 2023 1.9 GB 131k tokens ? Q8_0 Comfortable
28.8 tok/s

17–46 · low confidence

SantaCoder ≈ 1.1B Jan 2023 1.9 GB 131k tokens ? Q8_0 Comfortable
28.8 tok/s

17–46 · low confidence

TinyLlama-1.1B (1T token checkpoint) ≈ 1.1B Oct 2023 1.9 GB 131k tokens ? Q8_0 Comfortable
28.8 tok/s

17–46 · low confidence

TinyLlama-1.1B (3T token checkpoint) ≈ 1.1B Oct 2023 1.9 GB 131k tokens ? Q8_0 Comfortable
26.4 tok/s

16–42 · low confidence

EXAONE 4.0 (1.2B) ≈ 1.2B Jul 2025 2.0 GB 131k tokens ? Q8_0 Comfortable
26.4 tok/s

16–42 · low confidence

MinerU2.5 ≈ 1.2B Sep 2025 2.0 GB 131k tokens ? Q8_0 Comfortable
26.4 tok/s

16–42 · low confidence

Pleias 1.0 1.2B ≈ 1.2B Dec 2024 2.0 GB 131k tokens ? Q8_0 Comfortable
26.4 tok/s

16–42 · low confidence

Pleias-RAG-1B ≈ 1.2B Apr 2025 2.0 GB 131k tokens ? Q8_0 Comfortable
25.8 tok/s

15–41 · low confidence

Llama 3.2 1B 1.2B Sep 2024 2.2 GB 54k tokens Q8_0 Comfortable
25.4 tok/s

15–41 · low confidence

MiniCPM-1.2B ≈ 1.2B Jun 2024 2.0 GB 131k tokens ? Q8_0 Comfortable
24.4 tok/s

15–39 · low confidence

DeepSeek Coder 1.3B ≈ 1.3B Jan 2024 2.1 GB 131k tokens ? Q8_0 Comfortable
24.4 tok/s

15–39 · low confidence

DeepSeek-VL-1.3B ≈ 1.3B Mar 2024 2.1 GB 131k tokens ? Q8_0 Comfortable
24.4 tok/s

15–39 · low confidence

DigiRL ≈ 1.3B Jun 2024 2.1 GB 131k tokens ? Q8_0 Comfortable
24.4 tok/s

15–39 · low confidence

GLA Transformer 1.3B ≈ 1.3B Aug 2024 2.1 GB 131k tokens ? Q8_0 Comfortable
24.4 tok/s

15–39 · low confidence

Janus 1.3B ≈ 1.3B Oct 2024 2.1 GB 131k tokens ? Q8_0 Comfortable
24.4 tok/s

15–39 · low confidence

Kosmos-2.5 ≈ 1.3B Aug 2024 2.1 GB 131k tokens ? Q8_0 Comfortable
24.4 tok/s

15–39 · low confidence

Otter ≈ 1.3B May 2023 2.1 GB 131k tokens ? Q8_0 Comfortable
24.4 tok/s

15–39 · low confidence

Phi-1 ≈ 1.3B Oct 2023 2.1 GB 131k tokens ? Q8_0 Comfortable

Speeds are estimates for a single request — one conversation at a time — calculated from memory bandwidth, model size and quantisation. Real throughput varies with the inference runtime and its version. Figures published by hardware vendors measure many simultaneous requests and are much higher.

On record

FirePro W4300 full specification

Everything on record for this board, ordered by how much it bears on running a language model rather than by how a spec sheet would list it. Memory comes first because it decides the outcome; the rest is context.

Memory

The two specifications that decide what this card can run and how quickly. Capacity sets which models fit; bandwidth sets how many tokens per second they produce once they do.

Memory size
4 GB
Memory bandwidth
96 GB/s
Memory type
GDDR5
Memory bus width
128 bit
Memory clock
1.5 GHz

The chip

Which processor is on the board and how it was manufactured. A smaller process size generally means more performance for the same power.

Graphics processor
Bonaire
Architecture
GCN 2.0
Generation
FirePro GCN(Wx300)
Foundry
TSMC
Process size
28 nm
Transistors
2.1 billion
Transistor density
13,000 K/mm²
Die size
160 mm²
Package
FCBGA-1093
Released
1 December 2015

Clock speeds

How fast the processor runs. Worth far less here than on a gaming benchmark: generating text is limited by memory bandwidth, so a higher clock barely moves the result.

Base clock
930 MHz
Boost clock
930 MHz

Processing units

What the chip contains. These drive graphics performance and matter mainly for processing a long prompt rather than for producing the answer.

Shading units
768
Texture mapping units
48
Render output units
16
L1 cache
16 KB
L2 cache
0.25 MB

Theoretical performance

Peak arithmetic rates published for the board. These are ceilings that no real workload reaches, and generating text reaches a small fraction of them because it is limited by memory rather than arithmetic.

Single precision (FP32)
1.4 TFLOPS
Double precision (FP64)
89.3 GFLOPS
Pixel rate
15 GPixel/s
Texture rate
45 GTexel/s

The board

What it takes to physically install and power the card — the practical constraints that decide whether it fits the machine you already own.

Power draw (TDP)
50 W
Suggested power supply
250 W
Power connectors
None
Bus interface
PCIe 3.0 x16
Slot width
Single-slot
Dimensions
171 mm
Display outputs
4x mini-DisplayPort 1.2

Software support

Which graphics and compute interfaces the card supports. CUDA compute capability is the one that bears on inference: below 7.0 there are no tensor cores, and modern inference software falls back to slower code paths.

DirectX
12.0
OpenGL
4.6
Vulkan
1.2
OpenCL
2.1
Shader model
6.5

Listings

Where to buy a FirePro W4300

No vendor is currently listing this card. Listings come from vendors who publish them here directly — browse the vendor directory to see who is selling what.

What the numbers mean

Memory: the specification that decides everything

Memory

4 GB

Bandwidth

96 GB/s

Largest model

DeciLM 6B

At 4 GB of GDDR5 the FirePro W4300 is limited to the smaller end of the catalogue. About 3.6 GB is actually available to a runtime, and a model has to fit entirely inside it before generating anything at all.

At 96 GB/s across a 128-bit bus, bandwidth is this card's real constraint. Every token requires reading the entire model out of memory, so a large model will feel slow here even when it fits.

The figure is the memory clock — 1.5 GHz here — multiplied by the bus width. It is why core counts predict generation speed so poorly.

The biggest thing it holds is DeciLM 6B (5.7B) at Q3_K_M compression, for about 15.0 tokens per second.

The chip and how it was built

The FirePro W4300 is built on the Bonaire graphics processor, using AMD's GCN 2.0 architecture, as part of the FirePro GCN(Wx300) generation.

The chip is manufactured by TSMC, on a 28 nm process, with a die measuring 160 mm², holding 2.1 billion transistors. A smaller process generally means more performance for the same power, though for language models it matters far less than the memory subsystem.

It was released in December 2015, roughly 10 years ago. Inference software support tends to follow hardware by a year or two, so a card of this age generally has mature, well-optimised code paths available to it.

Compute throughput, and why it matters less than it looks

FP64

89.3 GFLOPS

Double-precision throughput is 89.3 GFLOPS. It has no bearing on running a language model — no inference runtime uses it — but it separates datacentre parts from consumer ones, since the latter deliberately restrict it.

Clocks run from 930 MHz at base to 930 MHz boosted. Worth far less here than on a gaming benchmark: raising the clock speeds up the arithmetic, and the arithmetic is not what generation is waiting on.

Cache and processing units

The FirePro W4300 has 16 KB of L1 cache, backed by 0.25 MB of L2. Cache absorbs a share of the memory traffic that would otherwise hit the main bus, which is the one place on this page where a number other than bandwidth quietly affects generation speed — a large L2 lets more of the working set stay close to the cores.

There are 768 shading units, 48 texture mapping units, and 16 render output units. These drive graphics workloads and contribute to prompt processing, but they sit idle for much of the time a model spends generating a reply.

Power, size and installation

Power draw

50 W

The FirePro W4300 is rated at 50 W, with a 250 W power supply suggested for the whole system. Running a language model keeps a card busy in bursts rather than continuously — it draws hard while generating and idles between requests — so sustained draw over a working day is usually well below the rated figure.

The board occupies a single-slot, measuring 171 mm long. Worth checking against the case and power supply already in the machine, since the largest cards need considerably more of both than a typical desktop provides.

It connects over PCIe 3.0 x16. The interface governs how quickly a model is loaded from disk into the card, not how fast it runs once there, so a narrower link costs a few seconds at startup and nothing thereafter.

The extremes

The largest AI models that run on a FirePro W4300

The biggest open-weight models that fit on this card, newest first. Each is shown at the best compression the card can hold.

  1. 01 Qwen3.5-4B 4B · Q5_K_M · Feb 2026 14.2 tok/s
  2. 02 Voxtral Mini 4.7B · Q4_K_M · Jul 2025 15.6 tok/s
  3. 03 Typhoon 2.1 Gemma 4B 4B · Q4_K_M · May 2025 18.3 tok/s
  4. 04 Qwen3-4B 4B · Q3_K_M · Apr 2025 21.4 tok/s
  5. 05 Gemma 3 QAT 4B 4B · Q4_K_M · Apr 2025 18.3 tok/s
  6. 06 Gemma 3 4B 4B · Q4_K_M · Mar 2025 18.3 tok/s
  7. 07 Phi-4-Multimodal 5.6B · Q3_K_M · Mar 2025 15.3 tok/s
  8. 08 Minitron 4B 4.2B · Q4_K_M · Nov 2024 17.4 tok/s
  9. 09 XVERSE-MoE-A4.2B 4.2B · Q4_K_M · Apr 2024 17.4 tok/s
  10. 10 DeciLM 6B 5.7B · Q3_K_M · Sep 2023 15.0 tok/s

The fastest AI models on a FirePro W4300

Where this card produces tokens quickest. Smaller models dominate here, because generating each token means reading the whole model out of memory once.

  1. 01 Gemma 3 QAT 1B 1B · Q8_0 · 1.8 GB 31.7 tok/s
  2. 02 Gemma 3 1B 1B · Q8_0 · 1.8 GB 31.7 tok/s
  3. 03 LLama 3..2 Typhoon 2 1B 1B · Q8_0 · 1.8 GB 31.7 tok/s
  4. 04 OLMo-1B 1B · Q8_0 · 1.8 GB 31.7 tok/s
  5. 05 HGRN 1B (WT 103) 1B · Q8_0 · 1.8 GB 31.7 tok/s
  6. 06 Pythia-1b 1B · Q8_0 · 1.8 GB 31.7 tok/s
  7. 07 OpenELM-1.1B 1.1B · Q8_0 · 1.9 GB 29.4 tok/s
  8. 08 TinyLlama-1.1B (1T token checkpoint) 1.1B · Q8_0 · 1.9 GB 28.8 tok/s
  9. 09 TinyLlama-1.1B (3T token checkpoint) 1.1B · Q8_0 · 1.9 GB 28.8 tok/s
  10. 10 DeciCoder-1B 1.1B · Q8_0 · 1.9 GB 28.8 tok/s

Step by step

How to work out the tokens per second of a FirePro W4300

You do not have to calculate anything by hand — the gputps.com calculator on this page has already worked it out for every model this card can hold. Reading off the answer takes six steps.

  1. 01

    Find the model in the table

    The table lists 97 models this FirePro W4300 can run. Search by name, or by size — typing 27b matches on the parameter count even when the name never states it.

  2. 02

    Match the context to your work

    Set the context to your real working length. Short questions cost almost nothing; a long document can consume a large share of the card's 4 GB.

  3. 03

    Pin the comparison to one quality level

    Compression is what lets bigger models fit. The quality control drops any model that needs more of it than you are willing to give.

  4. 04

    Look at the range, not just the number

    Each speed is an estimate for a single conversation, with a range beneath it — 31.7 tok/s on Gemma 3 QAT 1B at the top end here. The same card and model vary by thirty to fifty per cent between inference runtimes.

  5. 05

    Check the headroom before you decide

    The fit column separates models that just fit from those with room to spare — worth checking against the card's 4 GB before settling on one.

  6. 06

    Check the same model from the other side

    Every model name in the table links to its own page, which runs the same calculation across every card we hold. That is where you see whether the FirePro W4300 is the right buy for it or merely a card that fits.

Answers

FirePro W4300 — common questions

01

How much memory does a FirePro W4300 have?

A FirePro W4300 has 4 GB of GDDR5 memory. Around a tenth of that is reserved by the inference runtime and the driver, leaving roughly 3.6 GB available for a model and its conversation.

02

What is the memory bandwidth of a FirePro W4300?

The FirePro W4300 has 96 GB/s of memory bandwidth, across a 128-bit memory bus. This is the single best predictor of how fast it generates text, because producing each token means reading the entire model out of memory once.

03

What type of memory does a FirePro W4300 use?

It uses GDDR5 clocked at 1.5 GHz. HBM types are found on datacentre accelerators and carry far more bandwidth than the GDDR used on desktop cards, which is why they generate tokens considerably faster at the same capacity.

04

Who makes the FirePro W4300?

The FirePro W4300 is a AMD product, with the chip manufactured by TSMC, on a 28 nm process.

05

When was the FirePro W4300 released?

The FirePro W4300 was released in December 2015.

06

How much power does a FirePro W4300 use?

The FirePro W4300 has a rated board power of 50 W, and a 250 W system power supply is suggested. Generating text draws hard in bursts and idles between requests, so average consumption over a working session is normally well below the rated figure.

07

How much cache does a FirePro W4300 have?

The FirePro W4300 has 16 KB of L1 cache, and 0.25 MB of L2 cache. Cache absorbs part of the memory traffic that would otherwise reach the main bus, so a larger L2 gives a modest lift to generation speed beyond what bandwidth alone predicts.

08

Does the FirePro W4300 support CUDA?

No. CUDA is NVIDIA-only, and the FirePro W4300 is a AMD card. It runs language models through ROCm, Vulkan or Metal depending on the software, which are less mature than the CUDA path — our estimates apply a penalty for that.

09

What bus interface does the FirePro W4300 use?

It uses PCIe 3.0 x16. This governs how fast a model is loaded onto the card rather than how fast it runs once loaded, so it costs a few seconds at startup and nothing during generation.

10

Is the FirePro W4300 good for running local AI models?

Its memory limits it to smaller models though its bandwidth means generation will feel slow on larger models. In total it runs 97 of the models we track. Whether that is enough depends entirely on which model you want — the table above answers that directly.

11

Can a FirePro W4300 run a model that does not fit in its memory?

Offloading past the card's 4 GB is possible and usually a false economy: the system-memory portion is slow enough to dominate the result.

12

Would two FirePro W4300 cards be twice as fast?

No. A second FirePro W4300 doubles the memory to 8 GB, which lets you hold models neither could hold alone, but generation does not split that way. These figures describe one card.

13

What AI models can a FirePro W4300 run?

97 of the 679 open-weight language models we track fit on a FirePro W4300 and can be run locally on it. The table on this page lists every one, with the memory it needs, the quantisation it runs at and an estimated generation speed.

14

What is the largest AI model a FirePro W4300 can run?

The largest model in our catalogue that fits on a FirePro W4300 is DeciLM 6B at 5.7B parameters, compressed to Q3_K_M. It generates roughly 15.0 tokens per second and needs about 3.5 GB of the card's memory.

15

How many tokens per second does a FirePro W4300 produce?

It depends on the model. On a FirePro W4300 the fastest model we track is Gemma 3 QAT 1B at about 31.7 tokens per second, while larger models run proportionally slower because each token requires reading the whole model out of memory once. Speeds are estimates for a single conversation at a time.

The other direction

Looking at it from the other side?

This page starts from the hardware. If you already know which model you want and need to know what it takes to run it, start from the model instead.

All GPUs