CNN Committee (MNIST)
No estimate
No hardware requirements for this model
The weights for this model have not been published, so it cannot be downloaded or run on your own hardware at any size. It is reachable only through its provider, and no graphics card changes that.
On record
Full specification
Everything on record for this model. Most of it describes how it was trained rather than how it runs — useful context for judging how much work went into it, and how it compares with models built at a different scale.
Origin
Who built this model, where, and when it was published.
- Organisation
- IDSIA
- Organisation type
- Academia
- Country
- Switzerland
- Published
- 18 September 2011
- Authors
- D. Ciresan, U. Meier, L. Gambardella, J. Schmidhuber
What it does
The problem areas the model was built for. A model can carry several of each.
- Domain
- Vision
- Task
- Image classification
- Approach
- Supervised
Size
How large the model is and how much data it was trained on. Parameters are the figure that decides whether it fits on a given graphics card.
- Parameters
- 120.6K
- Training data
- 420,000 tokens
- Epochs
- 800
1st CNN layer: 20*4*4=320 2nd CNN layer: 20*40*9*9=64800 1st FC layer: 40*3*3*150=54000 2nd FC layer: 150*10=1500 Total: 320+64800+54000+1500=120620
[images] "Our CNNs are trained for around 800 epochs"
Training compute
The arithmetic performed to train the model, measured in floating-point operations. It is a measure of what the training run cost, not of how fast the finished model answers you.
- Training compute
- 5.2 × 10¹⁶ FLOP
- How it was established
- Hardware,Operation counting
GPU hour estimate: 800 epochs, 14 hours per network, 7 networks total 7*14*60*60*(2*1580000000000+2*1345000000000)*0.3=6.19e17 Counting estimate: 50000 examples, 800 epochs 1st CNN layer: 2*20*26*26*4*4=432640 2nd CNN layer: 2*20*6*6*9*9*40=4665600 1st FC layer: 2*40*3*3*150=108000 2nd FC layer: 2*150*10=3000 Total: 432640+4665600+108000+3000=5209240 5209240*3*50000*800*7=4375761600000000=4.4e15 Geometric mean: sqrt(4.4e15*6.19e17)=5.2e16
The training run
What it physically took to train: which chips, how many, for how long, and what that drew from the wall.
- Training hardware
- NVIDIA GeForce GTX 580,NVIDIA GeForce GTX 480
- Chips used
- 4
- Wall-clock time
- 98 hours
Availability
Whether you can obtain the model and run it on your own hardware, which is what decides if any of the graphics-card figures on this page apply.
- Weights
- Closed — provider access only
- Model access
- Unreleased
- Training code
- Unreleased
How it is classified
Labels the source dataset applies when tracking notable models, and how confident it is in the entry.
- Frontier model
- Yes
- Why it is tracked
- Historical significance
- Record confidence
- Likely
Sources
Where this record came from and when it was last checked.
- Reference
- Convolutional Neural Network Committees for Handwritten Character Classification
- Last updated
- 11 February 2026
What the numbers mean
About this model
CNN Committee (MNIST) was published by IDSIA, in Switzerland, in September 2011. The organisation is categorised as academia.
It works in Vision, and is recorded as doing image classification.
Its weights were never published, so it can only be reached through its provider. No graphics card changes that.
What went into building it
Training it took roughly 5.2 × 10¹⁶ FLOP of computation, on NVIDIA GeForce GTX 580,NVIDIA GeForce GTX 480 — a measure of what producing the model cost, not of how fast it answers.
It was trained on about 420,000 tokens of text.
It is tracked in the underlying dataset for one reason in particular: historical significance.
Answers
CNN Committee (MNIST) — common questions
Who created CNN Committee (MNIST)?
CNN Committee (MNIST) was published by IDSIA, based in Switzerland, categorised as academia.
When was CNN Committee (MNIST) released?
CNN Committee (MNIST) was published in September 2011. Capability per parameter has improved considerably since, so a newer model of the same size is often the better use of the same hardware.
What is CNN Committee (MNIST) used for?
CNN Committee (MNIST) works in Vision, and is recorded as handling image classification. A model can carry several of each, so these are the areas it was built for rather than a limit on what it will attempt.
How much compute was used to train CNN Committee (MNIST)?
Around 5.2 × 10¹⁶ FLOP, on NVIDIA GeForce GTX 580,NVIDIA GeForce GTX 480. That measures what producing the model cost and says nothing about how quickly it answers once trained — inference speed comes from memory bandwidth, not from the training budget.
What GPU do I need to run CNN Committee (MNIST)?
None. CNN Committee (MNIST) is a closed model — its weights were never published, so it cannot be downloaded or run on your own hardware at any price. It is reachable only through its provider.
Is CNN Committee (MNIST) open source?
No. CNN Committee (MNIST) has not had its weights published, so it exists only as a service controlled by its owner.
How many parameters does CNN Committee (MNIST) have?
CNN Committee (MNIST) has 120.6K parameters. 1st CNN layer: 20*4*4=320 2nd CNN layer: 20*40*9*9=64800 1st FC layer: 40*3*3*150=54000 2nd FC layer: 150*10=1500 Total: 320+64800+54000+1500=120620. That figure is the total, and it is what decides how much memory the model needs — roughly half a gigabyte per billion at the compression most people use.
The other direction
Looking at it from the other side?
This page starts from the model. If you already own a card and want to know everything it will run, start from the hardware instead.