Comments (2)
Hi @meve01 ,
I think connecting the simulator to STT-RAM is feasible, but you will require to perform some changes.
I have another version of the simulator which will push soon, and it will include the connection with DRAMSimv3. However, if you want to incorporate your own connection with STT-RAM this is what you should do, depending on the scenario you want to model:
- If you plan to implement an explicit decoupled memory hierarchy you will need to change quite a lot the module SDMemory. Here, you will have to separate these two modules into two: A first module will control the accesses to the DRAM and will feed the SRAM. Another one, will feed the multipliers based on the data in the SRAM. You could use a double buffer scheme. Eventtually, your model will connect with your non-volalite technology in DRAM. This scheme is quite complicated because it requires to synchronize both controllers to feed both the SRAM and the multipliers.
- If your plan is to use an implicit memory hierarchy then it is easier. You will have to create your memory hierarchy model based on cache for example, having again your STT-RAM technology in the back-end. Here, the only thing you will have to do is to send the memory requests in SDMemory.cpp to the cache, and modify slightly the controller to manage the memory requests.
If you decide to go for it and need help, please let me know it. Also, do not hesitate to email me if you prefer so. This is my email: [email protected]
from stonne.
Thank you for your response @francisco-munoz!
I think our goal would be to use an implicit architecture that we'd make an STT-RAM.cpp and .h files for. This would use the second approach based on your response. I have the architectural model already done for STT-RAM but it exists in an NVSim simulator right now, so it will mainly be a matter of re-configuring that code to work with the STONNE simulator. My email is [email protected] and I will reach out to you soon if you're able to help me out at some point :).
Thanks again, Francisco!
from stonne.
Related Issues (20)
- stonne_connect compile fail when pytorch compilered with clang++ HOT 1
- `weights_tensorflow.pb` is in owner's trash HOT 4
- `NLP/BERT/` directory is missing HOT 3
- Missing building blocks of SIGMA HOT 2
- Error in installation of vision HOT 1
- Three questions about Utilization HOT 1
- Steps to Install STONNE successfully HOT 3
- Parameter stats_path='.' missing for the SimulatedConv2d HOT 2
- Python-frontend installation errors HOT 3
- Memory leaks during big simulation runs - Destructors not being called HOT 2
- Pytorch Frontend Install Error HOT 5
- Question about the 'simulated_matmul()' HOT 7
- Questions about list of simulated functions HOT 4
- Printing individual time breakup values for the Simulator HOT 3
- Can i change stonne for our own accelerator architecture HOT 4
- Can we implement training on STONNE? HOT 2
- DRAM modeling HOT 4
- The unit of data and frequency of the stonne HOT 6
- PyTorch frontend HOT 1
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from stonne.