Calculate the TPS of the Tesla K40c 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 · 74.1 tok/s
Fastest model
Gemma 3 QAT 1B
104 tok/s · 1B
Which AI models can run on a Tesla K40c?
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 | ||||||
|---|---|---|---|---|---|---|---|
|
104
tok/s
36–208 · low confidence |
Gemma 3 1B | 1B | Mar 2025 | 1.8 GB | 33k tokens | Q8_0 | Comfortable |
|
104
tok/s
36–208 · low confidence |
Gemma 3 QAT 1B | 1B | Apr 2025 | 1.8 GB | 33k tokens | Q8_0 | Comfortable |
|
104
tok/s
36–208 · low confidence |
HGRN 1B (WT 103) ≈ | 1B | Nov 2023 | 1.8 GB | 131k tokens ? | Q8_0 | Comfortable |
|
104
tok/s
36–208 · low confidence |
LLama 3..2 Typhoon 2 1B ≈ | 1B | Dec 2024 | 1.8 GB | 131k tokens ? | Q8_0 | Comfortable |
|
104
tok/s
36–208 · low confidence |
OLMo-1B ≈ | 1B | Feb 2024 | 1.8 GB | 131k tokens ? | Q8_0 | Comfortable |
|
104
tok/s
36–208 · low confidence |
Pythia-1b ≈ | 1B | Apr 2023 | 1.8 GB | 131k tokens ? | Q8_0 | Comfortable |
|
96.1
tok/s
34–192 · low confidence |
OpenELM-1.1B ≈ | 1.1B | May 2024 | 1.9 GB | 131k tokens ? | Q8_0 | Comfortable |
|
94.4
tok/s
33–189 · low confidence |
DeciCoder-1B ≈ | 1.1B | Aug 2023 | 1.9 GB | 131k tokens ? | Q8_0 | Comfortable |
|
94.4
tok/s
33–189 · low confidence |
SantaCoder ≈ | 1.1B | Jan 2023 | 1.9 GB | 131k tokens ? | Q8_0 | Comfortable |
|
94.4
tok/s
33–189 · low confidence |
TinyLlama-1.1B (1T token checkpoint) ≈ | 1.1B | Oct 2023 | 1.9 GB | 131k tokens ? | Q8_0 | Comfortable |
|
94.4
tok/s
33–189 · low confidence |
TinyLlama-1.1B (3T token checkpoint) ≈ | 1.1B | Oct 2023 | 1.9 GB | 131k tokens ? | Q8_0 | Comfortable |
|
86.5
tok/s
30–173 · low confidence |
EXAONE 4.0 (1.2B) ≈ | 1.2B | Jul 2025 | 2.0 GB | 131k tokens ? | Q8_0 | Comfortable |
|
86.5
tok/s
30–173 · low confidence |
MinerU2.5 ≈ | 1.2B | Sep 2025 | 2.0 GB | 131k tokens ? | Q8_0 | Comfortable |
|
86.5
tok/s
30–173 · low confidence |
Pleias 1.0 1.2B ≈ | 1.2B | Dec 2024 | 2.0 GB | 131k tokens ? | Q8_0 | Comfortable |
|
86.5
tok/s
30–173 · low confidence |
Pleias-RAG-1B ≈ | 1.2B | Apr 2025 | 2.0 GB | 131k tokens ? | Q8_0 | Comfortable |
|
84.4
tok/s
30–169 · low confidence |
Llama 3.2 1B | 1.2B | Sep 2024 | 2.2 GB | 131k tokens | Q8_0 | Comfortable |
|
83.2
tok/s
29–166 · low confidence |
DeepSeekMoE-16B | 16B | Jan 2024 | 10.0 GB | 4k tokens | Q4_K_M | Tight |
|
83.2
tok/s
29–166 · low confidence |
MiniCPM-1.2B ≈ | 1.2B | Jun 2024 | 2.0 GB | 131k tokens ? | Q8_0 | Comfortable |
|
79.9
tok/s
28–160 · low confidence |
DeepSeek Coder 1.3B ≈ | 1.3B | Jan 2024 | 2.1 GB | 131k tokens ? | Q8_0 | Comfortable |
|
79.9
tok/s
28–160 · low confidence |
DeepSeek-VL-1.3B ≈ | 1.3B | Mar 2024 | 2.1 GB | 131k tokens ? | Q8_0 | Comfortable |
|
79.9
tok/s
28–160 · low confidence |
DigiRL ≈ | 1.3B | Jun 2024 | 2.1 GB | 131k tokens ? | Q8_0 | Comfortable |
|
79.9
tok/s
28–160 · low confidence |
GLA Transformer 1.3B ≈ | 1.3B | Aug 2024 | 2.1 GB | 131k tokens ? | Q8_0 | Comfortable |
|
79.9
tok/s
28–160 · low confidence |
Janus 1.3B ≈ | 1.3B | Oct 2024 | 2.1 GB | 131k tokens ? | Q8_0 | Comfortable |
|
79.9
tok/s
28–160 · low confidence |
Kosmos-2.5 ≈ | 1.3B | Aug 2024 | 2.1 GB | 131k tokens ? | Q8_0 | Comfortable |
|
79.9
tok/s
28–160 · 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
Tesla K40c 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
- 288 GB/s
- Memory type
- GDDR5
- Memory bus width
- 384 bit
- Memory clock
- 1.5 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
- GK180
- Architecture
- Kepler
- Generation
- Tesla Kepler(Kxx)
- Foundry
- TSMC
- Process size
- 28 nm
- Transistors
- 7.1 billion
- Transistor density
- 12,600 K/mm²
- Die size
- 561 mm²
- Package
- BGA-2152
- Released
- 8 October 2013
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
- 745 MHz
- Boost clock
- 876 MHz
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
- 2,880
- Texture mapping units
- 240
- Render output units
- 48
- L1 cache
- 16 KB
- L2 cache
- 1.5 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.
- Single precision (FP32)
- 5 TFLOPS
- Double precision (FP64)
- 1.7 TFLOPS
- Pixel rate
- 53 GPixel/s
- Texture rate
- 210 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)
- 245 W
- Suggested power supply
- 550 W
- Power connectors
- 1x 6-pin + 1x 8-pin
- Bus interface
- PCIe 3.0 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.
- CUDA compute capability
- 3.5
- DirectX
- 11.0
- OpenGL
- 4.6
- Vulkan
- 1.2
- OpenCL
- 3.0
- Shader model
- 5.1
Listings
Where to buy a Tesla K40c
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
288 GB/s
Largest model
ERNIE-4.5-21B-A3B
12 GB of GDDR5 puts the Tesla K40c 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.
At 288 GB/s across a 384-bit bus, 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 1.5 GHz memory clock across the bus width above. 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.
Put together, the largest model that fits is ERNIE-4.5-21B-A3B at 21B, running Q3_K_M and producing around 74.1 tokens per second.
The chip and how it was built
The Tesla K40c is built on the GK180 graphics processor, using NVIDIA's Kepler architecture, as part of the Tesla Kepler(Kxx) generation.
The chip is manufactured by TSMC, on a 28 nm process, with a die measuring 561 mm², holding 7.1 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 October 2013, roughly 12 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
FP64
1.7 TFLOPS
Double-precision throughput is 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 745 MHz at base to 876 MHz 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 Tesla K40c has 16 KB of L1 cache, backed by 1.5 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 2,880 shading units, 240 texture mapping units, and 48 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
245 W
The Tesla K40c is rated at 245 W, with a 550 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 1x 6-pin + 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 3.0 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 Tesla K40c
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 Tesla K40c
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 Tesla K40c
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
All 396 models the Tesla K40c handles are already listed. The search box takes a name or a size such as 27b, which matches on parameter count.
-
02
Set the context length you will actually use
Longer conversations cost memory on top of the weights. With 12 GB to work in, 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
Speeds come with error bars for a reason. The best case here is 104 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 12 GB available.
-
06
Open the model to compare cards
Each model page repeats this calculation for the whole catalogue. Worth a look before deciding: it shows what else runs the same model, and how the Tesla K40c compares.
Answers
Tesla K40c — common questions
What is the largest AI model a Tesla K40c can run?
The largest model in our catalogue that fits on a Tesla K40c is ERNIE-4.5-21B-A3B at 21B parameters, compressed to Q3_K_M. It generates roughly 74.1 tokens per second and needs about 10.1 GB of the card's memory.
How many tokens per second does a Tesla K40c produce?
It depends on the model. On a Tesla K40c the fastest model we track is Gemma 3 QAT 1B at about 104 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 Tesla K40c run a 7B model?
Yes. For example a Tesla K40c runs DeepSeek Coder 6.7B at Q6_K, using about 9.9 GB of memory and generating around 22.5 tokens per second.
Can a Tesla K40c run a 13B model?
Yes. For example a Tesla K40c runs DeepSeekMoE-16B at Q4_K_M, using about 10.0 GB of memory and generating around 83.2 tokens per second.
How much memory does a Tesla K40c have?
A Tesla K40c has 12 GB of GDDR5 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 Tesla K40c?
The Tesla K40c has 288 GB/s of memory bandwidth, across a 384-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 Tesla K40c use?
It uses GDDR5 clocked at 1.5 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 Tesla K40c?
The Tesla K40c is a NVIDIA product, with the chip manufactured by TSMC, on a 28 nm process.
When was the Tesla K40c released?
The Tesla K40c was released in October 2013.
How much power does a Tesla K40c use?
The Tesla K40c has a rated board power of 245 W, and a 550 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 Tesla K40c have?
The Tesla K40c has 16 KB of L1 cache, and 1.5 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.
Does the Tesla K40c support CUDA?
Yes. The Tesla K40c reports CUDA compute capability 3.5, which predates tensor cores. 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 Tesla K40c use?
It uses PCIe 3.0 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 Tesla K40c good for running local AI models?
Its memory covers small and mid-sized models, though the largest are out of reach though its bandwidth means generation will feel slow on larger models. 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 Tesla K40c 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 Tesla K40c cards be twice as fast?
Capacity adds, throughput does not. Two of them give you 24 GB to work with rather than twice the tokens per second — every figure here is for a single Tesla K40c.
What AI models can a Tesla K40c run?
396 of the 679 open-weight language models we track fit on a Tesla K40c 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.
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.