Calculate the TPS of the Radeon Instinct MI100 on local AI models
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
679 models in our catalogue altogether
Largest model it holds
Phi-3.5-MoE
60.8B · Q3_K_M · 100 tok/s
Fastest model
Gemma 3 QAT 1B
406 tok/s · 1B
Which AI models can run on a Radeon Instinct MI100?
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 | ||||||
|---|---|---|---|---|---|---|---|
|
406
tok/s
244–650 · low confidence |
Gemma 3 1B | 1B | Mar 2025 | 1.8 GB | 33k tokens | Q8_0 | Comfortable |
|
406
tok/s
244–650 · low confidence |
Gemma 3 QAT 1B | 1B | Apr 2025 | 1.8 GB | 33k tokens | Q8_0 | Comfortable |
|
406
tok/s
244–650 · low confidence |
HGRN 1B (WT 103) ≈ | 1B | Nov 2023 | 1.8 GB | 131k tokens ? | Q8_0 | Comfortable |
|
406
tok/s
244–650 · low confidence |
LLama 3..2 Typhoon 2 1B ≈ | 1B | Dec 2024 | 1.8 GB | 131k tokens ? | Q8_0 | Comfortable |
|
406
tok/s
244–650 · low confidence |
OLMo-1B ≈ | 1B | Feb 2024 | 1.8 GB | 131k tokens ? | Q8_0 | Comfortable |
|
406
tok/s
244–650 · low confidence |
Pythia-1b ≈ | 1B | Apr 2023 | 1.8 GB | 131k tokens ? | Q8_0 | Comfortable |
|
376
tok/s
226–602 · low confidence |
OpenELM-1.1B ≈ | 1.1B | May 2024 | 1.9 GB | 131k tokens ? | Q8_0 | Comfortable |
|
369
tok/s
222–591 · low confidence |
DeciCoder-1B ≈ | 1.1B | Aug 2023 | 1.9 GB | 131k tokens ? | Q8_0 | Comfortable |
|
369
tok/s
222–591 · low confidence |
SantaCoder ≈ | 1.1B | Jan 2023 | 1.9 GB | 131k tokens ? | Q8_0 | Comfortable |
|
369
tok/s
222–591 · low confidence |
TinyLlama-1.1B (1T token checkpoint) ≈ | 1.1B | Oct 2023 | 1.9 GB | 131k tokens ? | Q8_0 | Comfortable |
|
369
tok/s
222–591 · low confidence |
TinyLlama-1.1B (3T token checkpoint) ≈ | 1.1B | Oct 2023 | 1.9 GB | 131k tokens ? | Q8_0 | Comfortable |
|
339
tok/s
203–542 · low confidence |
EXAONE 4.0 (1.2B) ≈ | 1.2B | Jul 2025 | 2.0 GB | 131k tokens ? | Q8_0 | Comfortable |
|
339
tok/s
203–542 · low confidence |
MinerU2.5 ≈ | 1.2B | Sep 2025 | 2.0 GB | 131k tokens ? | Q8_0 | Comfortable |
|
339
tok/s
203–542 · low confidence |
Pleias 1.0 1.2B ≈ | 1.2B | Dec 2024 | 2.0 GB | 131k tokens ? | Q8_0 | Comfortable |
|
339
tok/s
203–542 · low confidence |
Pleias-RAG-1B ≈ | 1.2B | Apr 2025 | 2.0 GB | 131k tokens ? | Q8_0 | Comfortable |
|
330
tok/s
198–529 · low confidence |
Llama 3.2 1B | 1.2B | Sep 2024 | 2.2 GB | 131k tokens | Q8_0 | Comfortable |
|
326
tok/s
195–521 · low confidence |
MiniCPM-1.2B ≈ | 1.2B | Jun 2024 | 2.0 GB | 131k tokens ? | Q8_0 | Comfortable |
|
313
tok/s
188–500 · low confidence |
DeepSeek Coder 1.3B ≈ | 1.3B | Jan 2024 | 2.1 GB | 131k tokens ? | Q8_0 | Comfortable |
|
313
tok/s
188–500 · low confidence |
DeepSeek-VL-1.3B ≈ | 1.3B | Mar 2024 | 2.1 GB | 131k tokens ? | Q8_0 | Comfortable |
|
313
tok/s
188–500 · low confidence |
DigiRL ≈ | 1.3B | Jun 2024 | 2.1 GB | 131k tokens ? | Q8_0 | Comfortable |
|
313
tok/s
188–500 · low confidence |
GLA Transformer 1.3B ≈ | 1.3B | Aug 2024 | 2.1 GB | 131k tokens ? | Q8_0 | Comfortable |
|
313
tok/s
188–500 · low confidence |
Janus 1.3B ≈ | 1.3B | Oct 2024 | 2.1 GB | 131k tokens ? | Q8_0 | Comfortable |
|
313
tok/s
188–500 · low confidence |
Kosmos-2.5 ≈ | 1.3B | Aug 2024 | 2.1 GB | 131k tokens ? | Q8_0 | Comfortable |
|
313
tok/s
188–500 · low confidence |
Otter ≈ | 1.3B | May 2023 | 2.1 GB | 131k tokens ? | Q8_0 | Comfortable |
|
313
tok/s
188–500 · 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 Instinct MI100 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
- 1,230 GB/s
- Memory type
- HBM2
- Memory bus width
- 4,096 bit
- Memory clock
- 1.2 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
- Arcturus
- Architecture
- CDNA 1.0
- Generation
- Radeon Instinct(MIx)
- Foundry
- TSMC
- Process size
- 7 nm
- Transistors
- 25.6 billion
- Transistor density
- 34,100 K/mm²
- Die size
- 750 mm²
- Released
- 16 November 2020
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 GHz
- Boost clock
- 1.5 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
- 7,680
- Texture mapping units
- 480
- Render output units
- 64
- L1 cache
- 16 KB
- L2 cache
- 8 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)
- 184.6 TFLOPS
- Single precision (FP32)
- 23.1 TFLOPS
- Double precision (FP64)
- 11.5 TFLOPS
- Pixel rate
- 96 GPixel/s
- Texture rate
- 721 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)
- 300 W
- Suggested power supply
- 700 W
- Power connectors
- 2x 8-pin
- Bus interface
- PCIe 4.0 x16
- Slot width
- Dual-slot
- Dimensions
- 267 mm
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.
- OpenCL
- 2.1
Listings
Where to buy a Radeon Instinct MI100
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
Why memory is the number that matters here
Memory
32 GB
Bandwidth
1,230 GB/s
Largest model
Phi-3.5-MoE
The Radeon Instinct MI100 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 1,230 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 — 1.2 GHz here — multiplied by the bus width. It is why core counts predict generation speed so poorly.
In practice that combination tops out at Phi-3.5-MoE — 60.8B, compressed to Q3_K_M, generating around 100 tokens per second.
The chip and how it was built
The Radeon Instinct MI100 is built on the Arcturus graphics processor, using AMD's CDNA 1.0 architecture, as part of the Radeon Instinct(MIx) generation.
The chip is manufactured by TSMC, on a 7 nm process, with a die measuring 750 mm², holding 25.6 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 November 2020, roughly 5 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
184.6 TFLOPS
FP64
11.5 TFLOPS
On paper the Radeon Instinct MI100 reaches 184.6 TFLOPS at half precision and 23.1 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 11.5 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 GHz at base to 1.5 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 Instinct MI100 has 16 KB of L1 cache, backed by 8 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 7,680 shading units, 480 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
300 W
The Radeon Instinct MI100 is rated at 300 W, with a 700 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 dual-slot, measuring 267 mm long, and needs 2x 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 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 Radeon Instinct MI100
The biggest open-weight models that fit on this card, newest first. Each is shown at the best compression the card can hold.
The fastest AI models on a Radeon Instinct MI100
Where this card produces tokens quickest. Smaller models dominate here, because generating each token means reading the whole model out of memory once.
Step by step
How to work out the tokens per second of a Radeon Instinct MI100
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.
-
01
Start with the model, not the specification
Every one of the 513 models this Radeon Instinct MI100 runs is in the table above. Search narrows it by name or by size.
-
02
Decide how long your conversations run
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 32 GB.
-
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.
-
04
Read the speed and the range
The figures are calculated, not measured. 406 tok/s on Gemma 3 QAT 1B is the fastest result on this card, and like every row it carries a range that reflects how much the runtime matters.
-
05
Check the memory column before committing
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.
-
06
Cross-check against other hardware
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 Radeon Instinct MI100 is the right buy for it or merely a card that fits.
Answers
Radeon Instinct MI100 — common questions
Does the Radeon Instinct MI100 support CUDA?
No. CUDA is NVIDIA-only, and the Radeon Instinct MI100 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.
What bus interface does the Radeon Instinct MI100 use?
It uses PCIe 4.0 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.
Is the Radeon Instinct MI100 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.
Can a Radeon Instinct MI100 run a model that does not fit in its memory?
It can be split, with the overflow held in system memory — but that part drags the whole thing down, and none of the 32 GB figures on this page assume it.
Would two Radeon Instinct MI100 cards be twice as fast?
No. A second Radeon Instinct MI100 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.
What AI models can a Radeon Instinct MI100 run?
513 of the 679 open-weight language models we track fit on a Radeon Instinct MI100 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.
What is the largest AI model a Radeon Instinct MI100 can run?
The largest model in our catalogue that fits on a Radeon Instinct MI100 is Phi-3.5-MoE at 60.8B parameters, compressed to Q3_K_M. It generates roughly 100 tokens per second and needs about 27.2 GB of the card's memory.
How many tokens per second does a Radeon Instinct MI100 produce?
It depends on the model. On a Radeon Instinct MI100 the fastest model we track is Gemma 3 QAT 1B at about 406 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.
Can a Radeon Instinct MI100 run a 7B model?
Yes. For example a Radeon Instinct MI100 runs Multi-Token Prediction 7B at Q8_0, using about 7.9 GB of memory and generating around 60.7 tokens per second.
Can a Radeon Instinct MI100 run a 13B model?
Yes. For example a Radeon Instinct MI100 runs DeepSeekMoE-16B at Q8_0, using about 17.5 GB of memory and generating around 141 tokens per second.
Can a Radeon Instinct MI100 run a 30B model?
Yes. For example a Radeon Instinct MI100 runs ERNIE-4.5-VL-28B-A3B at Q6_K, using about 22.7 GB of memory and generating around 117 tokens per second.
How much memory does a Radeon Instinct MI100 have?
A Radeon Instinct MI100 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.
What is the memory bandwidth of a Radeon Instinct MI100?
The Radeon Instinct MI100 has 1,230 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.
What type of memory does a Radeon Instinct MI100 use?
It uses HBM2 clocked at 1.2 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.
Who makes the Radeon Instinct MI100?
The Radeon Instinct MI100 is a AMD product, with the chip manufactured by TSMC, on a 7 nm process.
When was the Radeon Instinct MI100 released?
The Radeon Instinct MI100 was released in November 2020.
How much power does a Radeon Instinct MI100 use?
The Radeon Instinct MI100 has a rated board power of 300 W, and a 700 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.
How much cache does a Radeon Instinct MI100 have?
The Radeon Instinct MI100 has 16 KB of L1 cache, and 8 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.
What are the TFLOPS of a Radeon Instinct MI100?
The Radeon Instinct MI100 is rated at 184.6 TFLOPS at half precision and 23.1 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.
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.