Calculate the TPS of the Arc B580 on local AI models

Intel 12 GB GDDR6 456 GB/s December 2024

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

411 models it can run

721 models in our catalogue altogether

Largest model it holds

ERNIE-4.5-21B-A3B

21B · Q3_K_M · 89.6 tok/s

Fastest model

Gemma 3 QAT 1B

126 tok/s · 1B

Which AI models can run on a Arc B580?

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.

411 models match

Calculating
Quantisation Fit
126 tok/s

75–201 · low confidence

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

75–201 · low confidence

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

75–201 · low confidence

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

75–201 · low confidence

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

75–201 · low confidence

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

75–201 · low confidence

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

70–186 · low confidence

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

68–183 · low confidence

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

68–183 · low confidence

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

68–183 · low confidence

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

68–183 · low confidence

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

63–167 · low confidence

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

63–167 · low confidence

LFM2-1.2B 1.2B Jul 2025 2.0 GB 131k tokens ? Q8_0 Comfortable
105 tok/s

63–167 · low confidence

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

63–167 · low confidence

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

63–167 · low confidence

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

61–163 · low confidence

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

60–161 · low confidence

DeepSeekMoE-16B 16B Jan 2024 10.0 GB 4k tokens Q4_K_M Tight
101 tok/s

60–161 · low confidence

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

58–155 · low confidence

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

58–155 · low confidence

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

58–155 · low confidence

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

58–155 · low confidence

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

58–155 · low confidence

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

58–155 · low confidence

Kosmos-2.5 1.3B Aug 2024 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

Arc B580 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
12 GB
Memory bandwidth
456 GB/s
Memory type
GDDR6
Memory bus width
192 bit
Memory clock
2.38 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
BMG-G21
Architecture
Xe2-HPG
Generation
Battlemage(Arc 5)
Foundry
TSMC
Process size
5 nm
Transistors
19.6 billion
Transistor density
72,100 K/mm²
Die size
272 mm²
Released
13 December 2024

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
2.67 GHz
Boost clock
2.67 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
2,560
Texture mapping units
160
Render output units
80
Ray tracing cores
20
L1 cache
250 KB
L2 cache
18 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)
27.3 TFLOPS
Single precision (FP32)
13.7 TFLOPS
Double precision (FP64)
1.7 TFLOPS
Pixel rate
214 GPixel/s
Texture rate
427 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)
190 W
Suggested power supply
450 W
Power connectors
1x 8-pin
Bus interface
PCIe 4.0 x8
Slot width
Dual-slot
Dimensions
272 mm × 45 mm
Display outputs
1x HDMI 2.1a, 3x DisplayPort 2.1

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
12.2
OpenGL
4.6
Vulkan
1.4
OpenCL
3.0
Shader model
6.6

Listings

Where to buy a Arc B580

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

12 GB

Bandwidth

456 GB/s

Largest model

ERNIE-4.5-21B-A3B

Arc B580 carries 12 GB of GDDR6. That reaches comfortably into small and mid-sized models, though the largest stay out of reach without splitting them. Driver overhead leaves roughly 10.8 GB.

Memory bandwidth reaches 456 GB/s across a bus of 192 bits. That is the number governing generation speed — arithmetic per byte read is small enough that the bus, not the cores, is what everything waits on.

That comes from a memory clock of 2.38 GHz. Widening the bus and raising the clock are the two levers a manufacturer has, which is why a card with unremarkable cores can still generate quickly.

In practice that combination tops out at ERNIE-4.5-21B-A3B, 21B, compressed to Q3_K_M and generating around 89.6 tokens per second.

The chip and how it was built

Arc B580 is built on the graphics processor BMG-G21, using the architecture Xe2-HPG from Intel, as part of the generation Battlemage(Arc 5).

The chip is manufactured by TSMC, on a process of 5 nm, with a die measuring 272 mm², holding 19.6 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 2024, roughly 1.7533521398286 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

27.3 TFLOPS

FP64

1.7 TFLOPS

On paper Arc B580 reaches 27.3 TFLOPS at half precision, and 13.7 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 reaches 1.7 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 a base of 2.67 GHz to a boost of 2.67 GHz. 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

Arc B580 has an L1 cache of 250 KB, backed by an L2 cache of 18 MB. 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 2,560 shading units, 160 texture mapping units, and 80 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

190 W

Arc B580 is rated at 190 W, and the suggested system power supply is 450 W. 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 dual-slot, measuring 272 mm long, and needs 1x 8-pin. 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 4.0 x8. 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 Arc B580

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 ERNIE-4.5-21B-A3B 21B · Q3_K_M · Jun 2025 89.6 tok/s
  2. 02 GigaChat Lite (GigaChat-20B-A3B) 20B · Q3_K_M · Dec 2024 94.1 tok/s
  3. 03 InternLM2.5 20B · Q3_K_M · Aug 2024 16.9 tok/s
  4. 04 Granite 20B 20B · Q3_K_M · May 2024 16.9 tok/s
  5. 05 InternLM2-20B 20B · Q3_K_M · Jan 2024 16.9 tok/s
  6. 06 CogAgent 18B · IQ4_XS · Dec 2023 17.1 tok/s
  7. 07 SPHINX (Llama 2 13B) 19.9B · Q3_K_M · Nov 2023 17.0 tok/s
  8. 08 CogVLM-17B 17B · IQ4_XS · Nov 2023 18.1 tok/s
  9. 09 Flan UL2 19.5B · Q3_K_M · Mar 2023 17.4 tok/s
  10. 10 Palmyra Large 20B 20B · Q3_K_M · Mar 2023 16.9 tok/s

The fastest AI models on a Arc B580

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 126 tok/s
  2. 02 Gemma 3 1B 1B · Q8_0 · 1.8 GB 126 tok/s
  3. 03 LLama 3..2 Typhoon 2 1B 1B · Q8_0 · 1.8 GB 126 tok/s
  4. 04 OLMo-1B 1B · Q8_0 · 1.8 GB 126 tok/s
  5. 05 HGRN 1B (WT 103) 1B · Q8_0 · 1.8 GB 126 tok/s
  6. 06 Pythia-1b 1B · Q8_0 · 1.8 GB 126 tok/s
  7. 07 OpenELM-1.1B 1.1B · Q8_0 · 1.9 GB 116 tok/s
  8. 08 TinyLlama-1.1B (1T token checkpoint) 1.1B · Q8_0 · 1.9 GB 114 tok/s
  9. 09 TinyLlama-1.1B (3T token checkpoint) 1.1B · Q8_0 · 1.9 GB 114 tok/s
  10. 10 DeciCoder-1B 1.1B · Q8_0 · 1.9 GB 114 tok/s

Step by step

How to work out the tokens per second of a Arc B580

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

    Search for the model you want

    The table lists 411 models this card 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, but a long document can consume a large share of 12 GB that is frequently the difference between a model fitting and not.

  3. 03

    Pin the comparison to one quality level

    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

    Speeds come with error bars for a reason. The best case here is 126 tok/s on Gemma 3 QAT 1B. The same card and model vary by thirty to fifty per cent between inference runtimes.

  5. 05

    Read the fit verdict last

    Tight means it works today; comfortable means it still works when the conversation grows. Compare what each model needs against an available 12 GB.

  6. 06

    Check the same model from the other side

    Each model page repeats this calculation for the whole catalogue. Worth a look before deciding: it shows what else runs the same model, alongside Arc B580.

Answers

Arc B580 — common questions

01

Arc B580— what is its memory bandwidth?

Memory bandwidth reaches 456 GB/s across a bus of 192 bits. 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.

02

Arc B580— what type of memory does it use?

It uses GDDR6 clocked at 2.38 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.

03

Arc B580— who makes it?

This is a product of Intel, with the chip manufactured by TSMC, on a process of 5 nm.

04

Arc B580— when was it released?

It was released in December 2024.

05

Arc B580— how much power does it use?

Rated board power is 190 W, and the suggested system power supply is 450 W. Generating text draws hard in bursts and idles between requests, so average consumption over a working session is normally well below the rated figure.

06

Arc B580— how much cache does it have?

The L1 cache is 250 KB, and the L2 cache is 18 MB. 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.

07

Arc B580— what are its TFLOPS?

It is rated at 27.3 TFLOPS at half precision and 13.7 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.

08

Arc B580— does it support CUDA?

No. CUDA is NVIDIA-only, and this is a card from Intel. 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

Arc B580— what bus interface does it use?

It uses PCIe 4.0 x8. 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

Arc B580— is it good for running local AI models?

Its memory covers small and mid-sized models, though the largest are out of reach and its bandwidth gives usable, if unspectacular, generation speeds. In total it runs 411 of the models we track. Whether that is enough depends entirely on which model you want — the table above answers that directly.

11

Arc B580— can it run a model that does not fit in its memory?

Offloading past the card's 12 GB drags the whole thing down, and none of the figures on this page assume it.

12

Would two Arc B580 cards be twice as fast?

No. A second card doubles the memory to 24 GB to work with rather than twice the tokens per second — every figure here is for a single card.

13

Arc B580— which AI models can it run?

411 of the 721 open-weight language models we track fit on this card and can be run locally. The table on this page lists every one, with the memory it needs, the quantisation it runs at and an estimated generation speed.

14

Arc B580— what is the largest AI model it can run?

The largest model in our catalogue that fits is ERNIE-4.5-21B-A3B at 21B parameters, compressed to Q3_K_M. It generates roughly 89.6 tokens per second and needs about 10.1 GB of the card's memory.

15

Arc B580— how many tokens per second does it produce?

It depends on the model. The fastest model we track here is Gemma 3 QAT 1B at about 126 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.

16

Arc B580— can it run 7B models?

Yes. For example it runs Gemma 4 E4B at Q8_0, using about 9.8 GB of memory and generating around 27.9 tokens per second.

17

Arc B580— can it run 13B models?

Yes. For example it runs DeepSeekMoE-16B at Q4_K_M, using about 10.0 GB of memory and generating around 101 tokens per second.

18

Arc B580— how much memory does it have?

This card has 12 GB of GDDR6. Around a tenth is reserved by the inference runtime and the driver, leaving roughly 10.8 GB available for a model and its conversation.

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