The SignalPop products are designed to seamlessly work with the MyCaffe AI Platform. See us on GitHub for more information on MyCaffe! To easily integrate MyCaffe into your products, just search for MyCaffe on Nuget!
Sign up for a free account to get access to the free SignalPop AI Designer beta download or to purchase a full-featured version!
- SignalPop AI Designer Personal Edition 1 year subscription – enables all general product features for one year (does not included Enhanced features).
- SignalPop AI Designer Beta Edition 3 month subscription – enables all general product features for three months (does not included Enhanced features).
Free Product Downloads
- SignalPop AI Designer Download – version 0.10.2.124
- SignalPop Universal Miner Download – version 0.10.2.124
- SignalPop AI Designer Download – version 0.10.2.38
- SignalPop Universal Miner Download – version 0.10.2.38
Universal Miner Note for Symantec 360 Users
Using the SignalPop Universal Miner with the Symantec Norton 360 firewall may trigger a false positive for the ‘Miner.Bitcoinminer Activity 22’ alert, which appears to detect the Stratum protocol used by the WPE.NET.MINER.ETHEREUM.10.0.EXE SignalPop Universal Miner executable to communicate with the Ethermine.org pool site. This Symantec alert will stop the SignalPop Universal Miner from being able to mine. To disable this alert, follow the steps below:
- Open the Symantec 360 Product and select Settings.
- Next, select Firewall and then Intrusion and Browser Protection.
- Select the Configure link from the Intrusion Signatures item.
- Disable the Miner.Bitcoinminer Activity 22 item with signature ID 31152.
After accepting these settings, re-run the SignalPop Universal Miner and mining will run and correctly mine Ethereum.
Are you looking for a way to design, build, train and debug your AI models on your local Windows computer and easily integrate those models into your C# based product? The SignalPop AI Designer is specifically designed to help you do just that!
The SignalPop AI Designer is an AI development environment that helps develop, build, train, test and debug your MyCaffe AI models locally on your own Windows-based development PC. Even with the low-cost NVIDIA 1050ti GPU (under $200), you can start deep learning with GoogleNet or AlexNet on your local computer or laptop – without cloud charges!
Once you have the SignalPop AI Designer installed and running, here are some of the cool things that you can do:
Easily create and manage datasets via the Dataset Creators.
Visual dataset analysis via Iterative PCA or t-SNE algorithms.
Dynamic dataset gyms supported.
Visual editing of your model via drag-n-drop operations.
Visual blob debugging to view the data and diff contents of each blob flowing between layers.
Visual layer debugging to view how a given layer behaves.
Real-time debugging which allows you to view the data flowing through your network as you train.
Visual debugging via deconvolution and un-pooling.
Deep draw and deep dream generation over multiple octaves.
Deep dream generation
Neural Style Transfer (limited functionality) allows you to learn a style from one photograph and apply it to another ‘content‘ photograph to produce a new art piece.
The following Neural Style Transfer limitations apply to the various product editions.
NOTE: Enhanced features are only available to SignalPop partners at this time.
The General and Enhanced product editions have unlimited training time, including multi-gpu based training on headless GPU’s using NCCL 1.
And much more!
Minimum System Requirements
The SignalPop AI Designer is built for Windows developers running on either Windows 10 (recommended) or Windows 7 PC’s that have at least one NVIDIA CUDA based GPU installed. With that in mind the minimum requirements for the SignalPop AI Designer are as follows:
Operating System: 64-bit Windows 10 or Windows 7
Hard Disk Space: 10 GB free disk space (SSD recommended)
SignalPop AI Designer Specific Requirements
System Memory (PC side): at least 8 GB
GPU Model: NVIDIA 1050TI or above with CUDA 10.1 (NVIDIA 750TI or above with CUDA 10.2). CUDA support required with NVIDIA driver version 441.28 or later.
GPU Compute: by default, compute 5.3 and above (Maxwell) are used when using CUDA 10.2 (compute 3.5 is also provided, but is not the default).
GPU Memory: 2 GB (4 GB or more recommended)
Multi-GPU Training: requires at least two headless TCC capable GPU’s.
Database: Microsoft SQL or SQL Express 2008 R2 or above (Microsoft SQL or SQL Express 2016 and above is recommended on Windows 10) running in Windows Authentication Mode.
SignalPop Universal Miner Specific Requirements
System Memory (PC side): at least 4 GB
GPU Model: NVIDIA 1050TI or above (requires CUDA support and the latest NVIDIA driver version 441.28 or later). We recommend using the NVIDIA 1080 TI, 1070 TI, 1070, or 1060 GPUs for they provide cooling control.
GPU Memory: 4 GB (6 GB or more recommended)