Calculate the TPS of the RTX 3500 Embedded Ada Generation 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
ERNIE-4.5-21B-A3B
21B · Q3_K_M · 131 tok/s
Fastest model
Gemma 3 QAT 1B
183 tok/s · 1B
Which AI models can run on a RTX 3500 Embedded Ada Generation?
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.
396 models match
Calculating| Quantisation | Fit | ||||||
|---|---|---|---|---|---|---|---|
|
183
tok/s
156–220 |
Gemma 3 1B | 1B | Mar 2025 | 1.8 GB | 33k tokens | Q8_0 | Comfortable |
|
183
tok/s
156–220 |
Gemma 3 QAT 1B | 1B | Apr 2025 | 1.8 GB | 33k tokens | Q8_0 | Comfortable |
|
183
tok/s
110–293 · low confidence |
HGRN 1B (WT 103) ≈ | 1B | Nov 2023 | 1.8 GB | 131k tokens ? | Q8_0 | Comfortable |
|
183
tok/s
110–293 · low confidence |
LLama 3..2 Typhoon 2 1B ≈ | 1B | Dec 2024 | 1.8 GB | 131k tokens ? | Q8_0 | Comfortable |
|
183
tok/s
110–293 · low confidence |
OLMo-1B ≈ | 1B | Feb 2024 | 1.8 GB | 131k tokens ? | Q8_0 | Comfortable |
|
183
tok/s
110–293 · low confidence |
Pythia-1b ≈ | 1B | Apr 2023 | 1.8 GB | 131k tokens ? | Q8_0 | Comfortable |
|
169
tok/s
102–271 · low confidence |
OpenELM-1.1B ≈ | 1.1B | May 2024 | 1.9 GB | 131k tokens ? | Q8_0 | Comfortable |
|
166
tok/s
100–266 · low confidence |
DeciCoder-1B ≈ | 1.1B | Aug 2023 | 1.9 GB | 131k tokens ? | Q8_0 | Comfortable |
|
166
tok/s
100–266 · low confidence |
SantaCoder ≈ | 1.1B | Jan 2023 | 1.9 GB | 131k tokens ? | Q8_0 | Comfortable |
|
166
tok/s
100–266 · low confidence |
TinyLlama-1.1B (1T token checkpoint) ≈ | 1.1B | Oct 2023 | 1.9 GB | 131k tokens ? | Q8_0 | Comfortable |
|
166
tok/s
100–266 · low confidence |
TinyLlama-1.1B (3T token checkpoint) ≈ | 1.1B | Oct 2023 | 1.9 GB | 131k tokens ? | Q8_0 | Comfortable |
|
152
tok/s
91–244 · low confidence |
EXAONE 4.0 (1.2B) ≈ | 1.2B | Jul 2025 | 2.0 GB | 131k tokens ? | Q8_0 | Comfortable |
|
152
tok/s
91–244 · low confidence |
MinerU2.5 ≈ | 1.2B | Sep 2025 | 2.0 GB | 131k tokens ? | Q8_0 | Comfortable |
|
152
tok/s
91–244 · low confidence |
Pleias 1.0 1.2B ≈ | 1.2B | Dec 2024 | 2.0 GB | 131k tokens ? | Q8_0 | Comfortable |
|
152
tok/s
91–244 · low confidence |
Pleias-RAG-1B ≈ | 1.2B | Apr 2025 | 2.0 GB | 131k tokens ? | Q8_0 | Comfortable |
|
149
tok/s
126–179 |
Llama 3.2 1B | 1.2B | Sep 2024 | 2.2 GB | 131k tokens | Q8_0 | Comfortable |
|
147
tok/s
88–235 · low confidence |
DeepSeekMoE-16B | 16B | Jan 2024 | 10.0 GB | 4k tokens | Q4_K_M | Tight |
|
147
tok/s
88–235 · low confidence |
MiniCPM-1.2B ≈ | 1.2B | Jun 2024 | 2.0 GB | 131k tokens ? | Q8_0 | Comfortable |
|
141
tok/s
84–225 · low confidence |
DeepSeek Coder 1.3B ≈ | 1.3B | Jan 2024 | 2.1 GB | 131k tokens ? | Q8_0 | Comfortable |
|
141
tok/s
84–225 · low confidence |
DeepSeek-VL-1.3B ≈ | 1.3B | Mar 2024 | 2.1 GB | 131k tokens ? | Q8_0 | Comfortable |
|
141
tok/s
84–225 · low confidence |
DigiRL ≈ | 1.3B | Jun 2024 | 2.1 GB | 131k tokens ? | Q8_0 | Comfortable |
|
141
tok/s
84–225 · low confidence |
GLA Transformer 1.3B ≈ | 1.3B | Aug 2024 | 2.1 GB | 131k tokens ? | Q8_0 | Comfortable |
|
141
tok/s
84–225 · low confidence |
Janus 1.3B ≈ | 1.3B | Oct 2024 | 2.1 GB | 131k tokens ? | Q8_0 | Comfortable |
|
141
tok/s
84–225 · low confidence |
Kosmos-2.5 ≈ | 1.3B | Aug 2024 | 2.1 GB | 131k tokens ? | Q8_0 | Comfortable |
|
141
tok/s
84–225 · low confidence |
Otter ≈ | 1.3B | May 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
RTX 3500 Embedded Ada Generation 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
- 432 GB/s
- Memory type
- GDDR6
- Memory bus width
- 192 bit
- Memory clock
- 2.25 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
- AD104
- Architecture
- Ada Lovelace
- Generation
- Ada-MW(x000A)
- Foundry
- TSMC
- Process size
- 5 nm
- Transistors
- 35.8 billion
- Transistor density
- 121,800 K/mm²
- Die size
- 294 mm²
- Released
- 21 March 2023
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
- 1.73 GHz
- Boost clock
- 2.25 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
- 5,120
- Texture mapping units
- 160
- Render output units
- 64
- Streaming multiprocessors
- 40
- Tensor cores
- 160
- Ray tracing cores
- 40
- L1 cache
- 128 KB
- L2 cache
- 48 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)
- 23 TFLOPS
- Single precision (FP32)
- 23 TFLOPS
- Double precision (FP64)
- 360 GFLOPS
- Pixel rate
- 144 GPixel/s
- Texture rate
- 360 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)
- 100 W
- Suggested power supply
- 300 W
- Power connectors
- None
- Bus interface
- PCIe 4.0 x16
- Slot width
- IGP
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.
- CUDA compute capability
- 8.9
- DirectX
- 12.2
- OpenGL
- 4.6
- Vulkan
- 1.4
- OpenCL
- 3.0
- Shader model
- 6.8
Listings
Where to buy a RTX 3500 Embedded Ada Generation
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
What the memory subsystem means for AI
Memory
12 GB
Bandwidth
432 GB/s
Largest model
ERNIE-4.5-21B-A3B
12 GB of GDDR6 puts the RTX 3500 Embedded Ada Generation comfortably into small and mid-sized models, with roughly 10.8 GB usable once the driver overhead is taken out. The largest models are out of reach without splitting them.
The memory bus moves 432 GB/s across a 192-bit bus. That is the number that governs generation speed — arithmetic per byte read is small enough that the bus, not the cores, is what everything waits on.
The figure is the memory clock — 2.25 GHz here — multiplied by the bus width. It is why core counts predict generation speed so poorly.
The practical ceiling is ERNIE-4.5-21B-A3B at 21B, held at Q3_K_M and running at roughly 131 tokens per second.
The chip and how it was built
The RTX 3500 Embedded Ada Generation is built on the AD104 graphics processor, using NVIDIA's Ada Lovelace architecture, as part of the Ada-MW(x000A) generation.
The chip is manufactured by TSMC, on a 5 nm process, with a die measuring 294 mm², holding 35.8 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 March 2023, roughly 3 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
23 TFLOPS
FP64
360 GFLOPS
Tensor cores
160
On paper the RTX 3500 Embedded Ada Generation reaches 23 TFLOPS at half precision and 23 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 360 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.
The card carries 160 tensor cores across 40 streaming multiprocessors. These accelerate the matrix arithmetic at the heart of a transformer, and they are what make prompt processing — reading a long document before answering — dramatically faster than it would otherwise be.
Clocks run from 1.73 GHz at base to 2.25 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 RTX 3500 Embedded Ada Generation has 128 KB of L1 cache, backed by 48 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 5,120 shading units, 160 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
100 W
The RTX 3500 Embedded Ada Generation is rated at 100 W, with a 300 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 igp. 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 RTX 3500 Embedded Ada Generation
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 RTX 3500 Embedded Ada Generation
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 RTX 3500 Embedded Ada Generation
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
Every one of the 396 models this RTX 3500 Embedded Ada Generation runs is in the table above. Search narrows it by name or by size.
-
02
Match the context to your work
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 12 GB.
-
03
Set a minimum quality if you need one
Compression is what lets bigger models fit. The quality control drops any model that needs more of it than you are willing to give.
-
04
Take the range as the answer
The figures are calculated, not measured. 183 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
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 12 GB available.
-
06
Check the same model from the other side
Following a model through to its own page lists all the hardware that can run it, so you can see where the RTX 3500 Embedded Ada Generation sits against the alternatives.
Answers
RTX 3500 Embedded Ada Generation — common questions
What AI models can a RTX 3500 Embedded Ada Generation run?
396 of the 679 open-weight language models we track fit on a RTX 3500 Embedded Ada Generation 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 RTX 3500 Embedded Ada Generation can run?
The largest model in our catalogue that fits on a RTX 3500 Embedded Ada Generation is ERNIE-4.5-21B-A3B at 21B parameters, compressed to Q3_K_M. It generates roughly 131 tokens per second and needs about 10.1 GB of the card's memory.
How many tokens per second does a RTX 3500 Embedded Ada Generation produce?
It depends on the model. On a RTX 3500 Embedded Ada Generation the fastest model we track is Gemma 3 QAT 1B at about 183 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 RTX 3500 Embedded Ada Generation run a 7B model?
Yes. For example a RTX 3500 Embedded Ada Generation runs DeepSeek Coder 6.7B at Q6_K, using about 9.9 GB of memory and generating around 39.7 tokens per second.
Can a RTX 3500 Embedded Ada Generation run a 13B model?
Yes. For example a RTX 3500 Embedded Ada Generation runs DeepSeekMoE-16B at Q4_K_M, using about 10.0 GB of memory and generating around 147 tokens per second.
How much memory does a RTX 3500 Embedded Ada Generation have?
A RTX 3500 Embedded Ada Generation has 12 GB of GDDR6 memory. Around a tenth of that is reserved by the inference runtime and the driver, leaving roughly 10.8 GB available for a model and its conversation.
What is the memory bandwidth of a RTX 3500 Embedded Ada Generation?
The RTX 3500 Embedded Ada Generation has 432 GB/s of memory bandwidth, across a 192-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 RTX 3500 Embedded Ada Generation use?
It uses GDDR6 clocked at 2.25 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 RTX 3500 Embedded Ada Generation?
The RTX 3500 Embedded Ada Generation is a NVIDIA product, with the chip manufactured by TSMC, on a 5 nm process.
When was the RTX 3500 Embedded Ada Generation released?
The RTX 3500 Embedded Ada Generation was released in March 2023.
How much power does a RTX 3500 Embedded Ada Generation use?
The RTX 3500 Embedded Ada Generation has a rated board power of 100 W, and a 300 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 RTX 3500 Embedded Ada Generation have?
The RTX 3500 Embedded Ada Generation has 128 KB of L1 cache, and 48 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 RTX 3500 Embedded Ada Generation?
The RTX 3500 Embedded Ada Generation is rated at 23 TFLOPS at half precision and 23 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.
How many tensor cores does a RTX 3500 Embedded Ada Generation have?
The RTX 3500 Embedded Ada Generation has 160 tensor cores across 40 streaming multiprocessors. They accelerate the matrix arithmetic a transformer is built from, which mainly speeds up processing a long prompt rather than producing the reply.
Does the RTX 3500 Embedded Ada Generation support CUDA?
Yes. The RTX 3500 Embedded Ada Generation reports CUDA compute capability 8.9. Capability 7.0 and above has tensor cores, which modern inference software uses; below that it falls back to slower code paths for quantised models.
What bus interface does the RTX 3500 Embedded Ada Generation 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 RTX 3500 Embedded Ada Generation 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 396 of the models we track. Whether that is enough depends entirely on which model you want — the table above answers that directly.
Can a RTX 3500 Embedded Ada Generation run a model that does not fit in its memory?
Only partly. Layers beyond the 12 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 RTX 3500 Embedded Ada Generation cards be twice as fast?
No. A second RTX 3500 Embedded Ada Generation doubles the memory to 24 GB, which lets you hold models neither could hold alone, but generation does not split that way. These figures describe one 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.