Calculate the TPS of the Radeon Pro Vega II on local AI models

AMD 32 GB HBM2 825 GB/s June 2019

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

513 models it can run

679 models in our catalogue altogether

Largest model it holds

Phi-3.5-MoE

60.8B · Q3_K_M · 67.2 tok/s

Fastest model

Gemma 3 QAT 1B

273 tok/s · 1B

Which AI models can run on a Radeon Pro Vega II?

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.

513 models match

Calculating
Quantisation Fit
273 tok/s

164–436 · low confidence

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

164–436 · low confidence

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

164–436 · low confidence

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

164–436 · low confidence

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

164–436 · low confidence

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

164–436 · low confidence

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

151–404 · low confidence

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

149–397 · low confidence

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

149–397 · low confidence

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

149–397 · low confidence

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

149–397 · low confidence

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

136–364 · low confidence

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

136–364 · low confidence

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

136–364 · low confidence

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

136–364 · low confidence

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

133–355 · low confidence

Llama 3.2 1B 1.2B Sep 2024 2.2 GB 131k tokens Q8_0 Comfortable
219 tok/s

131–350 · low confidence

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

126–336 · low confidence

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

126–336 · low confidence

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

126–336 · low confidence

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

126–336 · low confidence

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

126–336 · low confidence

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

126–336 · low confidence

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

126–336 · low confidence

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

126–336 · 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

Radeon Pro Vega II 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
32 GB
Memory bandwidth
825 GB/s
Memory type
HBM2
Memory bus width
4,096 bit
Memory clock
806 MHz

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
Vega 20
Architecture
GCN 5.1
Generation
Radeon Pro Mac(Vega Series)
Foundry
TSMC
Process size
7 nm
Transistors
13.2 billion
Transistor density
40,000 K/mm²
Die size
331 mm²
Package
BGA-1933
Released
3 June 2019

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
1.57 GHz
Boost clock
1.72 GHz

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
4,096
Texture mapping units
256
Render output units
64
L1 cache
16 KB
L2 cache
4 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.

Half precision (FP16)
28.2 TFLOPS
Single precision (FP32)
14.1 TFLOPS
Double precision (FP64)
7 TFLOPS
Pixel rate
110 GPixel/s
Texture rate
440 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)
475 W
Suggested power supply
850 W
Bus interface
Apple MPX
Slot width
Quad-slot
Display outputs
1x HDMI 2.0b, 4x Thunderbolt

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
12.1
OpenGL
4.6
Vulkan
1.3
OpenCL
2.1
Shader model
6.7

Listings

Where to buy a Radeon Pro Vega II

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

32 GB

Bandwidth

825 GB/s

Largest model

Phi-3.5-MoE

The Radeon Pro Vega II carries 32 GB of HBM2, which covers the mid-sized models most people actually run — about 28.8 GB of it after the runtime and driver reserve their working space.

Bandwidth is 825 GB/s across a 4,096-bit bus. Generating a token means reading every weight once, so that figure sets the pace more than any other number here, and at this level text arrives faster than most people read.

The figure is the memory clock — 806 MHz here — multiplied by the bus width. It is why core counts predict generation speed so poorly.

The biggest thing it holds is Phi-3.5-MoE (60.8B) at Q3_K_M compression, for about 67.2 tokens per second.

The chip and how it was built

The Radeon Pro Vega II is built on the Vega 20 graphics processor, using AMD's GCN 5.1 architecture, as part of the Radeon Pro Mac(Vega Series) generation.

The chip is manufactured by TSMC, on a 7 nm process, with a die measuring 331 mm², holding 13.2 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 June 2019, roughly 7 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

FP16

28.2 TFLOPS

FP64

7 TFLOPS

On paper the Radeon Pro Vega II reaches 28.2 TFLOPS at half precision and 14.1 TFLOPS at single precision. These are peak figures no real workload sustains, and generating text reaches only a small fraction of them — decoding is limited by memory rather than arithmetic, which is why a card can look enormously powerful here and still produce tokens at an ordinary rate.

Double-precision throughput is 7 TFLOPS. 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 1.57 GHz at base to 1.72 GHz 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 Radeon Pro Vega II has 16 KB of L1 cache, backed by 4 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 4,096 shading units, 256 texture mapping units, and 64 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

475 W

The Radeon Pro Vega II is rated at 475 W, with a 850 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 quad-slot. 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 Apple MPX. 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 Radeon Pro Vega II

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 Kimi Linear 48B · IQ4_XS · Oct 2025 14.0 tok/s
  2. 02 Llama Nemotron Super v1.5 49B · IQ4_XS · Jul 2025 13.7 tok/s
  3. 03 Nemotron-H 56B 56B · Q3_K_M · Apr 2025 13.1 tok/s
  4. 04 Nemotron-H 47B 47B · IQ4_XS · Apr 2025 14.2 tok/s
  5. 05 Llama Nemotron Super 49B 49B · IQ4_XS · Mar 2025 13.7 tok/s
  6. 06 Jamba 1.6 Mini 52B · IQ4_XS · Mar 2025 55.8 tok/s
  7. 07 Jamba 1.5 Mini 52B · IQ4_XS · Aug 2024 55.8 tok/s
  8. 08 Qwen2-57B-A14B 57B · IQ4_XS · Jun 2024 47.8 tok/s
  9. 09 Phi-3.5-MoE 60.8B · Q3_K_M · Apr 2024 67.2 tok/s
  10. 10 Jamba 51.6B · IQ4_XS · Mar 2024 55.8 tok/s

The fastest AI models on a Radeon Pro Vega II

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 273 tok/s
  2. 02 Gemma 3 1B 1B · Q8_0 · 1.8 GB 273 tok/s
  3. 03 LLama 3..2 Typhoon 2 1B 1B · Q8_0 · 1.8 GB 273 tok/s
  4. 04 OLMo-1B 1B · Q8_0 · 1.8 GB 273 tok/s
  5. 05 HGRN 1B (WT 103) 1B · Q8_0 · 1.8 GB 273 tok/s
  6. 06 Pythia-1b 1B · Q8_0 · 1.8 GB 273 tok/s
  7. 07 OpenELM-1.1B 1.1B · Q8_0 · 1.9 GB 252 tok/s
  8. 08 TinyLlama-1.1B (1T token checkpoint) 1.1B · Q8_0 · 1.9 GB 248 tok/s
  9. 09 TinyLlama-1.1B (3T token checkpoint) 1.1B · Q8_0 · 1.9 GB 248 tok/s
  10. 10 DeciCoder-1B 1.1B · Q8_0 · 1.9 GB 248 tok/s

Step by step

How to work out the tokens per second of a Radeon Pro Vega II

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

    Start with the model, not the specification

    The table lists 513 models this Radeon Pro Vega II can run. Search by name, or by size — typing 27b matches on the parameter count even when the name never states it.

  2. 02

    Set the context length you will actually use

    Longer conversations cost memory on top of the weights. With 32 GB to work in, that is frequently the difference between a model fitting and not.

  3. 03

    Choose how far you will compress

    By default the table picks the least-compressed copy that fits. Setting a floor removes models that only qualify through heavy compression.

  4. 04

    Take the range as the answer

    Each speed is an estimate for a single conversation, with a range beneath it — 273 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

    Compare what each model needs with the 32 GB this card provides. Tight means it works today; comfortable means it still works when the conversation grows.

  6. 06

    Open the model to compare cards

    Each model page repeats this calculation for the whole catalogue. Worth a look before deciding: it shows what else runs the same model, and how the Radeon Pro Vega II compares.

Answers

Radeon Pro Vega II — common questions

01

Who makes the Radeon Pro Vega II?

The Radeon Pro Vega II is a AMD product, with the chip manufactured by TSMC, on a 7 nm process.

02

When was the Radeon Pro Vega II released?

The Radeon Pro Vega II was released in June 2019.

03

How much power does a Radeon Pro Vega II use?

The Radeon Pro Vega II has a rated board power of 475 W, and a 850 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.

04

How much cache does a Radeon Pro Vega II have?

The Radeon Pro Vega II has 16 KB of L1 cache, and 4 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.

05

What are the TFLOPS of a Radeon Pro Vega II?

The Radeon Pro Vega II is rated at 28.2 TFLOPS at half precision and 14.1 TFLOPS at single precision. These are peak arithmetic ceilings rather than achievable rates, and text generation reaches only a small fraction of them because it is limited by memory bandwidth instead.

06

Does the Radeon Pro Vega II support CUDA?

No. CUDA is NVIDIA-only, and the Radeon Pro Vega II 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.

07

What bus interface does the Radeon Pro Vega II use?

It uses Apple MPX. 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.

08

Is the Radeon Pro Vega II good for running local AI models?

Its memory comfortably covers the mid-sized models most people run locally and its bandwidth is high enough to generate text faster than most people read. In total it runs 513 of the models we track. Whether that is enough depends entirely on which model you want — the table above answers that directly.

09

Can a Radeon Pro Vega II run a model that does not fit in its memory?

Offloading past the card's 32 GB is possible and usually a false economy: the system-memory portion is slow enough to dominate the result.

10

Would two Radeon Pro Vega II cards be twice as fast?

No. A second Radeon Pro Vega II doubles the memory to 64 GB, which lets you hold models neither could hold alone, but generation does not split that way. These figures describe one card.

11

What AI models can a Radeon Pro Vega II run?

513 of the 679 open-weight language models we track fit on a Radeon Pro Vega II 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.

12

What is the largest AI model a Radeon Pro Vega II can run?

The largest model in our catalogue that fits on a Radeon Pro Vega II is Phi-3.5-MoE at 60.8B parameters, compressed to Q3_K_M. It generates roughly 67.2 tokens per second and needs about 27.2 GB of the card's memory.

13

How many tokens per second does a Radeon Pro Vega II produce?

It depends on the model. On a Radeon Pro Vega II the fastest model we track is Gemma 3 QAT 1B at about 273 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.

14

Can a Radeon Pro Vega II run a 7B model?

Yes. For example a Radeon Pro Vega II runs Multi-Token Prediction 7B at Q8_0, using about 7.9 GB of memory and generating around 40.7 tokens per second.

15

Can a Radeon Pro Vega II run a 13B model?

Yes. For example a Radeon Pro Vega II runs DeepSeekMoE-16B at Q8_0, using about 17.5 GB of memory and generating around 94.7 tokens per second.

16

Can a Radeon Pro Vega II run a 30B model?

Yes. For example a Radeon Pro Vega II runs ERNIE-4.5-VL-28B-A3B at Q6_K, using about 22.7 GB of memory and generating around 78.6 tokens per second.

17

How much memory does a Radeon Pro Vega II have?

A Radeon Pro Vega II has 32 GB of HBM2 memory. Around a tenth of that is reserved by the inference runtime and the driver, leaving roughly 28.8 GB available for a model and its conversation.

18

What is the memory bandwidth of a Radeon Pro Vega II?

The Radeon Pro Vega II has 825 GB/s of memory bandwidth, across a 4,096-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.

19

What type of memory does a Radeon Pro Vega II use?

It uses HBM2 clocked at 806 MHz. 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.

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