Calculate the TPS of the Radeon R7 M350 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
DeciLM 6B
5.7B · Q3_K_M · 2.5 tok/s
Fastest model
Gemma 3 QAT 1B
5.3 tok/s · 1B
Which AI models can run on a Radeon R7 M350?
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 | ||||||
|---|---|---|---|---|---|---|---|
|
5.3
tok/s
3–8 · low confidence |
Gemma 3 1B | 1B | Mar 2025 | 1.8 GB | 33k tokens | Q8_0 | Comfortable |
|
5.3
tok/s
3–8 · low confidence |
Gemma 3 QAT 1B | 1B | Apr 2025 | 1.8 GB | 33k tokens | Q8_0 | Comfortable |
|
5.3
tok/s
3–8 · low confidence |
HGRN 1B (WT 103) ≈ | 1B | Nov 2023 | 1.8 GB | 131k tokens ? | Q8_0 | Comfortable |
|
5.3
tok/s
3–8 · low confidence |
LLama 3..2 Typhoon 2 1B ≈ | 1B | Dec 2024 | 1.8 GB | 131k tokens ? | Q8_0 | Comfortable |
|
5.3
tok/s
3–8 · low confidence |
OLMo-1B ≈ | 1B | Feb 2024 | 1.8 GB | 131k tokens ? | Q8_0 | Comfortable |
|
5.3
tok/s
3–8 · low confidence |
Pythia-1b ≈ | 1B | Apr 2023 | 1.8 GB | 131k tokens ? | Q8_0 | Comfortable |
|
4.9
tok/s
3–8 · low confidence |
OpenELM-1.1B ≈ | 1.1B | May 2024 | 1.9 GB | 131k tokens ? | Q8_0 | Comfortable |
|
4.8
tok/s
3–8 · low confidence |
DeciCoder-1B ≈ | 1.1B | Aug 2023 | 1.9 GB | 131k tokens ? | Q8_0 | Comfortable |
|
4.8
tok/s
3–8 · low confidence |
SantaCoder ≈ | 1.1B | Jan 2023 | 1.9 GB | 131k tokens ? | Q8_0 | Comfortable |
|
4.8
tok/s
3–8 · low confidence |
TinyLlama-1.1B (1T token checkpoint) ≈ | 1.1B | Oct 2023 | 1.9 GB | 131k tokens ? | Q8_0 | Comfortable |
|
4.8
tok/s
3–8 · low confidence |
TinyLlama-1.1B (3T token checkpoint) ≈ | 1.1B | Oct 2023 | 1.9 GB | 131k tokens ? | Q8_0 | Comfortable |
|
4.4
tok/s
3–7 · low confidence |
EXAONE 4.0 (1.2B) ≈ | 1.2B | Jul 2025 | 2.0 GB | 131k tokens ? | Q8_0 | Comfortable |
|
4.4
tok/s
3–7 · low confidence |
MinerU2.5 ≈ | 1.2B | Sep 2025 | 2.0 GB | 131k tokens ? | Q8_0 | Comfortable |
|
4.4
tok/s
3–7 · low confidence |
Pleias 1.0 1.2B ≈ | 1.2B | Dec 2024 | 2.0 GB | 131k tokens ? | Q8_0 | Comfortable |
|
4.4
tok/s
3–7 · low confidence |
Pleias-RAG-1B ≈ | 1.2B | Apr 2025 | 2.0 GB | 131k tokens ? | Q8_0 | Comfortable |
|
4.3
tok/s
3–7 · low confidence |
Llama 3.2 1B | 1.2B | Sep 2024 | 2.2 GB | 54k tokens | Q8_0 | Comfortable |
|
4.2
tok/s
3–7 · low confidence |
MiniCPM-1.2B ≈ | 1.2B | Jun 2024 | 2.0 GB | 131k tokens ? | Q8_0 | Comfortable |
|
4.1
tok/s
2–7 · low confidence |
DeepSeek Coder 1.3B ≈ | 1.3B | Jan 2024 | 2.1 GB | 131k tokens ? | Q8_0 | Comfortable |
|
4.1
tok/s
2–7 · low confidence |
DeepSeek-VL-1.3B ≈ | 1.3B | Mar 2024 | 2.1 GB | 131k tokens ? | Q8_0 | Comfortable |
|
4.1
tok/s
2–7 · low confidence |
DigiRL ≈ | 1.3B | Jun 2024 | 2.1 GB | 131k tokens ? | Q8_0 | Comfortable |
|
4.1
tok/s
2–7 · low confidence |
GLA Transformer 1.3B ≈ | 1.3B | Aug 2024 | 2.1 GB | 131k tokens ? | Q8_0 | Comfortable |
|
4.1
tok/s
2–7 · low confidence |
Janus 1.3B ≈ | 1.3B | Oct 2024 | 2.1 GB | 131k tokens ? | Q8_0 | Comfortable |
|
4.1
tok/s
2–7 · low confidence |
Kosmos-2.5 ≈ | 1.3B | Aug 2024 | 2.1 GB | 131k tokens ? | Q8_0 | Comfortable |
|
4.1
tok/s
2–7 · low confidence |
Otter ≈ | 1.3B | May 2023 | 2.1 GB | 131k tokens ? | Q8_0 | Comfortable |
|
4.1
tok/s
2–7 · 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 R7 M350 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
- 16 GB/s
- Memory type
- DDR3
- Memory bus width
- 64 bit
- Memory clock
- 1 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
- Meso
- Architecture
- GCN 3.0
- Generation
- Gem System(R7 M300)
- Foundry
- TSMC
- Process size
- 28 nm
- Transistors
- 1.6 billion
- Transistor density
- 12,400 K/mm²
- Die size
- 125 mm²
- Released
- 5 May 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
- 1 GHz
- Boost clock
- 1.02 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
- 384
- Texture mapping units
- 24
- Render output units
- 8
- L1 cache
- 16 KB
- L2 cache
- 0.13 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)
- 779.5 GFLOPS
- Single precision (FP32)
- 779.5 GFLOPS
- Double precision (FP64)
- 48.7 GFLOPS
- Pixel rate
- 8 GPixel/s
- Texture rate
- 24 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.
- Bus interface
- PCIe 3.0 x8
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 Radeon R7 M350
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
Capacity and bandwidth
Memory
4 GB
Bandwidth
16 GB/s
Largest model
DeciLM 6B
Radeon R7 M350 carries only 4 GB of DDR3. That limits it to the smaller end of the catalogue, and a model has to fit entirely inside before it generates anything at all. A runtime actually gets about 3.6 GB.
Memory bandwidth reaches 16 GB/s across a bus of 64 bits. 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.
That comes from a memory clock of 1 GHz. Both halves matter, and neither is visible in a gaming benchmark.
The practical ceiling is DeciLM 6B, 5.7B, compressed to Q3_K_M and generating around 2.5 tokens per second.
The chip and how it was built
Radeon R7 M350 is built on the graphics processor Meso, using the architecture GCN 3.0 from AMD, as part of the generation Gem System(R7 M300).
The chip is manufactured by TSMC, on a process of 28 nm, with a die measuring 125 mm², holding 1.6 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 May 2015, roughly 11.241120374033 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
779.5 GFLOPS
FP64
48.7 GFLOPS
On paper Radeon R7 M350 reaches 779.5 GFLOPS at half precision, and 779.5 GFLOPS 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 48.7 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 a base of 1 GHz to a boost of 1.02 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
Radeon R7 M350 has an L1 cache of 16 KB, backed by an L2 cache of 0.13 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 384 shading units, 24 texture mapping units, and 8 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
It connects over PCIe 3.0 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 Radeon R7 M350
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 R7 M350
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 R7 M350
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
Search for the model you want
The table lists 97 models this card runs. Search narrows the list by name or by size.
-
02
Match the context to your work
Longer conversations cost memory on top of the weights. Against 4 GB that is frequently the difference between a model fitting and not.
-
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
Look at the range, not just the number
The figures are calculated, not measured. The fastest result on this card is 5.3 tok/s on Gemma 3 QAT 1B. The same card and model vary by thirty to fifty per cent between inference runtimes.
-
05
Check the headroom before you decide
The fit column separates models that just fit from those with room to spare — worth checking before settling on one, against an available 4 GB.
-
06
Cross-check against other hardware
Each model page repeats this calculation for the whole catalogue. Worth a look before deciding: it shows what else runs the same model, alongside Radeon R7 M350.
Answers
Radeon R7 M350 — common questions
Would two Radeon R7 M350 cards be twice as fast?
Capacity adds, throughput does not. Two of them give you 8 GB of combined memory, at roughly the same generation speed as one.
Radeon R7 M350— which AI models can it run?
97 of the 679 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.
Radeon R7 M350— what is the largest AI model it can run?
The largest model in our catalogue that fits is DeciLM 6B at 5.7B parameters, compressed to Q3_K_M. It generates roughly 2.5 tokens per second and needs about 3.5 GB of the card's memory.
Radeon R7 M350— 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 5.3 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.
Radeon R7 M350— how much memory does it have?
This card has 4 GB of DDR3. Around a tenth is reserved by the inference runtime and the driver, leaving roughly 3.6 GB available for a model and its conversation.
Radeon R7 M350— what is its memory bandwidth?
Memory bandwidth reaches 16 GB/s across a bus of 64 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.
Radeon R7 M350— what type of memory does it use?
It uses DDR3 clocked at 1 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.
Radeon R7 M350— who makes it?
This is a product of AMD, with the chip manufactured by TSMC, on a process of 28 nm.
Radeon R7 M350— when was it released?
It was released in May 2015.
Radeon R7 M350— how much cache does it have?
The L1 cache is 16 KB, and the L2 cache is 0.13 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.
Radeon R7 M350— what are its TFLOPS?
It is rated at 779.5 GFLOPS at half precision and 779.5 GFLOPS 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.
Radeon R7 M350— does it support CUDA?
No. CUDA is NVIDIA-only, and this is a card from AMD. 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.
Radeon R7 M350— what bus interface does it use?
It uses PCIe 3.0 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.
Radeon R7 M350— is it 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.
Radeon R7 M350— can it run a model that does not fit in its memory?
It can be split, with the overflow held in system memory beyond the card's 4 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.
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.