Calculate the TPS of the Radeon Instinct MI200 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
Largest model it holds
Qwen3.5-122B-A10B
122B · Q3_K_M · 66.6 tok/s
Fastest model
Gemma 3 QAT 1B
542 tok/s · 1B
What AI models can a Radeon Instinct MI200 run?
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.
595 models match
Calculating| Quantisation | Fit | ||||||
|---|---|---|---|---|---|---|---|
|
542
tok/s
325–867 · low confidence |
Gemma 3 1B | 1B | Mar 2025 | 1.8 GB | 33k tokens | Q8_0 | Comfortable |
|
542
tok/s
325–867 · low confidence |
Gemma 3 QAT 1B | 1B | Apr 2025 | 1.8 GB | 33k tokens | Q8_0 | Comfortable |
|
542
tok/s
325–867 · low confidence |
HGRN 1B (WT 103) ≈ | 1B | Nov 2023 | 1.8 GB | 131k tokens ? | Q8_0 | Comfortable |
|
542
tok/s
325–867 · low confidence |
LLama 3..2 Typhoon 2 1B ≈ | 1B | Dec 2024 | 1.8 GB | 131k tokens ? | Q8_0 | Comfortable |
|
542
tok/s
325–867 · low confidence |
OLMo-1B ≈ | 1B | Feb 2024 | 1.8 GB | 131k tokens ? | Q8_0 | Comfortable |
|
542
tok/s
325–867 · low confidence |
Pythia-1b ≈ | 1B | Apr 2023 | 1.8 GB | 131k tokens ? | Q8_0 | Comfortable |
|
502
tok/s
301–803 · low confidence |
OpenELM-1.1B ≈ | 1.1B | May 2024 | 1.9 GB | 131k tokens ? | Q8_0 | Comfortable |
|
493
tok/s
296–788 · low confidence |
DeciCoder-1B ≈ | 1.1B | Aug 2023 | 1.9 GB | 131k tokens ? | Q8_0 | Comfortable |
|
493
tok/s
296–788 · low confidence |
SantaCoder ≈ | 1.1B | Jan 2023 | 1.9 GB | 131k tokens ? | Q8_0 | Comfortable |
|
493
tok/s
296–788 · low confidence |
TinyLlama-1.1B (1T token checkpoint) ≈ | 1.1B | Oct 2023 | 1.9 GB | 131k tokens ? | Q8_0 | Comfortable |
|
493
tok/s
296–788 · low confidence |
TinyLlama-1.1B (3T token checkpoint) ≈ | 1.1B | Oct 2023 | 1.9 GB | 131k tokens ? | Q8_0 | Comfortable |
|
451
tok/s
271–722 · low confidence |
EXAONE 4.0 (1.2B) ≈ | 1.2B | Jul 2025 | 2.0 GB | 131k tokens ? | Q8_0 | Comfortable |
|
451
tok/s
271–722 · low confidence |
MinerU2.5 ≈ | 1.2B | Sep 2025 | 2.0 GB | 131k tokens ? | Q8_0 | Comfortable |
|
451
tok/s
271–722 · low confidence |
Pleias 1.0 1.2B ≈ | 1.2B | Dec 2024 | 2.0 GB | 131k tokens ? | Q8_0 | Comfortable |
|
451
tok/s
271–722 · low confidence |
Pleias-RAG-1B ≈ | 1.2B | Apr 2025 | 2.0 GB | 131k tokens ? | Q8_0 | Comfortable |
|
440
tok/s
264–705 · low confidence |
Llama 3.2 1B | 1.2B | Sep 2024 | 2.2 GB | 131k tokens | Q8_0 | Comfortable |
|
434
tok/s
261–695 · low confidence |
MiniCPM-1.2B ≈ | 1.2B | Jun 2024 | 2.0 GB | 131k tokens ? | Q8_0 | Comfortable |
|
417
tok/s
250–667 · low confidence |
DeepSeek Coder 1.3B ≈ | 1.3B | Jan 2024 | 2.1 GB | 131k tokens ? | Q8_0 | Comfortable |
|
417
tok/s
250–667 · low confidence |
DeepSeek-VL-1.3B ≈ | 1.3B | Mar 2024 | 2.1 GB | 131k tokens ? | Q8_0 | Comfortable |
|
417
tok/s
250–667 · low confidence |
DigiRL ≈ | 1.3B | Jun 2024 | 2.1 GB | 131k tokens ? | Q8_0 | Comfortable |
|
417
tok/s
250–667 · low confidence |
GLA Transformer 1.3B ≈ | 1.3B | Aug 2024 | 2.1 GB | 131k tokens ? | Q8_0 | Comfortable |
|
417
tok/s
250–667 · low confidence |
Janus 1.3B ≈ | 1.3B | Oct 2024 | 2.1 GB | 131k tokens ? | Q8_0 | Comfortable |
|
417
tok/s
250–667 · low confidence |
Kosmos-2.5 ≈ | 1.3B | Aug 2024 | 2.1 GB | 131k tokens ? | Q8_0 | Comfortable |
|
417
tok/s
250–667 · low confidence |
Otter ≈ | 1.3B | May 2023 | 2.1 GB | 131k tokens ? | Q8_0 | Comfortable |
|
417
tok/s
250–667 · 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 MI200 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
- 64 GB
- Memory bandwidth
- 1,640 GB/s
- Memory type
- HBM2e
- Memory bus width
- 4,096 bit
- Memory clock
- 1.6 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
- Aldebaran
- Architecture
- CDNA 2.0
- Generation
- Radeon Instinct(MIx)
- Foundry
- TSMC
- Process size
- 6 nm
- Transistors
- 58.2 billion
- Transistor density
- 80,400 K/mm²
- Die size
- 724 mm²
- Released
- 1 December 2021
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
- 1.7 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
- 6,656
- Texture mapping units
- 416
- L1 cache
- 16 KB
- L2 cache
- 16 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)
- 181 TFLOPS
- Single precision (FP32)
- 22.6 TFLOPS
- Double precision (FP64)
- 22.6 TFLOPS
- Texture rate
- 707 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
- 1x 8-pin
- Bus interface
- PCIe 4.0 x16
- Slot width
- OAM Module
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
- 3.0
Listings
Where to buy a Radeon Instinct MI200
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
64 GB
Bandwidth
1,640 GB/s
Largest model
Qwen3.5-122B-A10B
The Radeon Instinct MI200 carries 64 GB of HBM2e, which covers the mid-sized models most people actually run — about 57.6 GB of it after the runtime and driver reserve their working space.
Bandwidth is 1,640 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 — 1.6 GHz here — multiplied by the bus width. It is why core counts predict generation speed so poorly.
In practice that combination tops out at Qwen3.5-122B-A10B — 122B, compressed to Q3_K_M, generating around 66.6 tokens per second.
The chip and how it was built
The Radeon Instinct MI200 is built on the Aldebaran graphics processor, using AMD's CDNA 2.0 architecture, as part of the Radeon Instinct(MIx) generation.
The chip is manufactured by TSMC, on a 6 nm process, with a die measuring 724 mm², holding 58.2 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 December 2021, roughly 4 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
181 TFLOPS
FP64
22.6 TFLOPS
On paper the Radeon Instinct MI200 reaches 181 TFLOPS at half precision and 22.6 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 22.6 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 1.7 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 MI200 has 16 KB of L1 cache, backed by 16 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 6,656 shading units, 416 texture mapping 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 MI200 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 oam module, 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 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 a Radeon Instinct MI200 can run
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 MI200
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 MI200
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
Find the model in the table
The table lists 595 models this Radeon Instinct MI200 can run. Search by name, or by size — typing 27b matches on the parameter count even when the name never states it.
-
02
Match the context to your work
Longer conversations cost memory on top of the weights. With 64 GB to work in, that is frequently the difference between a model fitting and not.
-
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.
-
04
Take the range as the answer
Speeds come with error bars for a reason. The best case here is 542 tok/s on Gemma 3 QAT 1B, and which inference software you use moves that by thirty to fifty per cent.
-
05
Check the headroom before you decide
A tight fit runs but leaves no room to raise the context later; comfortable has headroom. The memory column shows what each model needs against the 64 GB available.
-
06
Cross-check against other hardware
Following a model through to its own page lists all the hardware that can run it, so you can see where the Radeon Instinct MI200 sits against the alternatives.
Answers
Radeon Instinct MI200 — common questions
When was the Radeon Instinct MI200 released?
The Radeon Instinct MI200 was released in December 2021.
How much power does a Radeon Instinct MI200 use?
The Radeon Instinct MI200 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 MI200 have?
The Radeon Instinct MI200 has 16 KB of L1 cache, and 16 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 MI200?
The Radeon Instinct MI200 is rated at 181 TFLOPS at half precision and 22.6 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.
Does the Radeon Instinct MI200 support CUDA?
No. CUDA is NVIDIA-only, and the Radeon Instinct MI200 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 MI200 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 MI200 good for running local AI models?
Its memory is large enough for models most desktop hardware cannot touch and its bandwidth is high enough to generate text faster than most people read. In total it runs 595 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 MI200 run a model that does not fit in its memory?
Only partly. Layers beyond the 64 GB sit in system memory and run at a fraction of the speed, so a mostly-offloaded model is rarely worth using. Every figure here assumes it is fully resident on the card.
Would two Radeon Instinct MI200 cards be twice as fast?
No. A second Radeon Instinct MI200 doubles the memory to 128 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 MI200 run?
595 of the 679 open-weight language models we track fit on a Radeon Instinct MI200 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 MI200 can run?
The largest model in our catalogue that fits on a Radeon Instinct MI200 is Qwen3.5-122B-A10B at 122B parameters, compressed to Q3_K_M. It generates roughly 66.6 tokens per second and needs about 53.1 GB of the card's memory.
How many tokens per second does a Radeon Instinct MI200 produce?
It depends on the model. On a Radeon Instinct MI200 the fastest model we track is Gemma 3 QAT 1B at about 542 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 MI200 run a 7B model?
Yes. For example a Radeon Instinct MI200 runs Multi-Token Prediction 7B at Q8_0, using about 7.9 GB of memory and generating around 80.9 tokens per second.
Can a Radeon Instinct MI200 run a 13B model?
Yes. For example a Radeon Instinct MI200 runs DeepSeekMoE-16B at Q8_0, using about 17.5 GB of memory and generating around 188 tokens per second.
Can a Radeon Instinct MI200 run a 30B model?
Yes. For example a Radeon Instinct MI200 runs ERNIE-4.5-VL-28B-A3B at Q8_0, using about 29.2 GB of memory and generating around 108 tokens per second.
Can a Radeon Instinct MI200 run a 70B model?
Yes. For example a Radeon Instinct MI200 runs Qwen3-Coder-Next at Q5_K_M, using about 52.7 GB of memory and generating around 67.2 tokens per second.
How much memory does a Radeon Instinct MI200 have?
A Radeon Instinct MI200 has 64 GB of HBM2e memory. Around a tenth of that is reserved by the inference runtime and the driver, leaving roughly 57.6 GB available for a model and its conversation.
What is the memory bandwidth of a Radeon Instinct MI200?
The Radeon Instinct MI200 has 1,640 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 MI200 use?
It uses HBM2e clocked at 1.6 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 MI200?
The Radeon Instinct MI200 is a AMD product, with the chip manufactured by TSMC, on a 6 nm process.
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.