You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@tvm.apache.org by Chris Hoge via Apache TVM Discuss <no...@discuss.tvm.ai> on 2021/06/23 18:10:43 UTC
[Apache TVM Discuss] [Application] TVM Community Survey
Hi everyone,
I'm putting together a short survey of TVM developers and users to help us to get a general understanding of what the community is using TVM for and directions for the roadmap to take. I've put together a set of questions, and would like your feedback on them. This is an open community survey, so all of the data and results will be available for analysis and review. Thanks for taking your time to look over these. If you have comments, you can leave them here or in [this Google doc](https://docs.google.com/document/d/1Ov9LaPCYVuc5GnGHniFnVpPP1hWEPnmb-9ESVDoy1iQ/edit?usp=sharing).
Please identify your experience with using TVM in production.
- [ ] I am new to TVM.
- [ ] I am currently exploring using TVM to optimize models.
- [ ] I am actively using TVM in production.
Please identify your development experience in the TVM Community
- [ ] I am new to the TVM community.
- [ ] I have contributed to TVM through code or reviews.
- [ ] I am a TVM project Reviewer
- [ ] I am a TVM project Committer
- [ ] I am a TVM PMC
What platforms are you using TVM for?
- [ ] X86 CPU
- [ ] ARM CPU
- [ ] Other CPU
- [ ] NVidia GPU
- [ ] AMD GPU
- [ ] Other GPU
- [ ] Embedded Platform
What sort of models are you optimizing?
- [ ] Image processing models
- [ ] Audio processing models
- [ ] Language models
- [ ] Other
What platforms are you importing models from?
- [ ] ONNX
- [ ] PyTorch
- [ ] TensorFlow
- [ ] MXNet
- [ ] Other
Which APIs are you directly using with TVM?
- [ ] C++
- [ ] Python
- [ ] Rust
- [ ] Go Lang
- [ ] Other
What aspects of TVM do you find easy to use?
Where do you feel that TVM could be improved?
What are you most excited about with TVM?
Any other comments?
---
[Visit Topic](https://discuss.tvm.apache.org/t/tvm-community-survey/10305/1) to respond.
You are receiving this because you enabled mailing list mode.
To unsubscribe from these emails, [click here](https://discuss.tvm.apache.org/email/unsubscribe/2ef1423fab135d4e2c806455c9af750bd5c92ab0cbf6024d4ae1e1f70a2a96e6).
[Apache TVM Discuss] [Application] TVM Community Survey
Posted by Xqdan via Apache TVM Discuss <no...@discuss.tvm.ai>.
[quote="hogepodge, post:1, topic:10305"]
What platforms are you using TVM for?
* [ ] X86 CPU
* [ ] ARM CPU
* [ ] Other CPU
* [ ] NVidia GPU
* [ ] AMD GPU
* [ ] Other GPU
* [ ] Embedded Platform
[/quote]
We are using TVM for DSA NPU, can you add one option, thanks!
---
[Visit Topic](https://discuss.tvm.apache.org/t/tvm-community-survey/10305/2) to respond.
You are receiving this because you enabled mailing list mode.
To unsubscribe from these emails, [click here](https://discuss.tvm.apache.org/email/unsubscribe/b4ef0750a32d0e34c4c34c0cd4688e804e06895dece4d68d35294a2026e7172c).
RE: [EXT] [Apache TVM Discuss] [Application] TVM Community Survey
Posted by Joe Chou <cc...@marvell.com>.
-----Original Message-----
From: noreply@discuss.tvm.ai <no...@discuss.tvm.ai>
Sent: Wednesday, June 23, 2021 11:11 AM
To: tvm.archiver@gmail.com
Subject: [EXT] [Apache TVM Discuss] [Application] TVM Community Survey
External Email
----------------------------------------------------------------------
Hi everyone,
I'm putting together a short survey of TVM developers and users to help us to get a general understanding of what the community is using TVM for and directions for the roadmap to take. I've put together a set of questions, and would like your feedback on them. This is an open community survey, so all of the data and results will be available for analysis and review. Thanks for taking your time to look over these. If you have comments, you can leave them here or in [this Google doc](https://urldefense.proofpoint.com/v2/url?u=https-3A__docs.google.com_document_d_1Ov9LaPCYVuc5GnGHniFnVpPP1hWEPnmb-2D9ESVDoy1iQ_edit-3Fusp-3Dsharing&d=DwICaQ&c=nKjWec2b6R0mOyPaz7xtfQ&r=84zI-iH2xz28Q_xujSGbjYaq38CgcRgl_vGzCtF6TwQ&m=wA2Vu8OXbJyLS-t2l3_L9SvS05CNdjmfRvknHBkow9Y&s=UvJEpW7f0_v7wa-VjIOvCcsGfaeHajZ8nqD4QqyY2xM&e= ).
Please identify your experience with using TVM in production.
- [ ] I am new to TVM.
- [x ] I am currently exploring using TVM to optimize models.
- [ ] I am actively using TVM in production.
Please identify your development experience in the TVM Community
- [x ] I am new to the TVM community.
- [ ] I have contributed to TVM through code or reviews.
- [ ] I am a TVM project Reviewer
- [ ] I am a TVM project Committer
- [ ] I am a TVM PMC
What platforms are you using TVM for?
- [x ] X86 CPU
- [x ] ARM CPU
- [ ] Other CPU
- [x ] NVidia GPU
- [x ] AMD GPU
- [ ] Other GPU
- [x ] Embedded Platform
What sort of models are you optimizing?
- [x ] Image processing models
- [ ] Audio processing models
- [ ] Language models
- [x ] Other
What platforms are you importing models from?
- [x ] ONNX
- [ ] PyTorch
- [x ] TensorFlow
- [x ] MXNet
- [ ] Other
Which APIs are you directly using with TVM?
- [x ] C++
- [x ] Python
- [ ] Rust
- [ ] Go Lang
- [ ] Other
What aspects of TVM do you find easy to use?
** Integrating user level python packages and ML python frameworks to run, say, target=llvm inference to get results as well as per-layer (or per fused op) related performance numbers
Where do you feel that TVM could be improved?
** TVM "Internal" development related documentations
** NOTE: TVM stacks and sub-components, plus, communications between them are complex. To enable user and domain-specific TVM-specialization developer to fully utilize TVM stacks & sub-components and to extend from current TVM codebase, respectively, IMO getting up to speed faster becomes the key.
What are you most excited about with TVM?
** People are really using TVM. TVM's Python integration make prototyping of domain-specific TVM specializations much easier. Once the AOT flow becomes stable, it can be used for production.
Any other comments?
** Thanks so many people make TVM available as it is now.
---
[Visit Topic](https://urldefense.proofpoint.com/v2/url?u=https-3A__discuss.tvm.apache.org_t_tvm-2Dcommunity-2Dsurvey_10305_1&d=DwICaQ&c=nKjWec2b6R0mOyPaz7xtfQ&r=84zI-iH2xz28Q_xujSGbjYaq38CgcRgl_vGzCtF6TwQ&m=wA2Vu8OXbJyLS-t2l3_L9SvS05CNdjmfRvknHBkow9Y&s=4RmR4q_yshSgBZ8e8FTTb2az4wz71PiJA5lPiw0xucQ&e= ) to respond.
You are receiving this because you enabled mailing list mode.
To unsubscribe from these emails, [click here](https://urldefense.proofpoint.com/v2/url?u=https-3A__discuss.tvm.apache.org_email_unsubscribe_2ef1423fab135d4e2c806455c9af750bd5c92ab0cbf6024d4ae1e1f70a2a96e6&d=DwICaQ&c=nKjWec2b6R0mOyPaz7xtfQ&r=84zI-iH2xz28Q_xujSGbjYaq38CgcRgl_vGzCtF6TwQ&m=wA2Vu8OXbJyLS-t2l3_L9SvS05CNdjmfRvknHBkow9Y&s=sd0cHgrEOihIJbQOU4-GpQK8KtwkFhtAv8JUQ04qa4M&e= ).
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@tvm.apache.org
For additional commands, e-mail: dev-help@tvm.apache.org
[Apache TVM Discuss] [Application] TVM Community Survey
Posted by Xqdan via Apache TVM Discuss <no...@discuss.tvm.ai>.
[quote="hogepodge, post:1, topic:10305"]
What platforms are you using TVM for?
* [ ] X86 CPU
* [ ] ARM CPU
* [ ] Other CPU
* [ ] NVidia GPU
* [ ] AMD GPU
* [ ] Other GPU
* [ ] Embedded Platform
[/quote]
We are using TVM for DSA NPU, can you add one option, thanks!
---
[Visit Topic](https://discuss.tvm.apache.org/t/tvm-community-survey/10305/2) to respond.
You are receiving this because you enabled mailing list mode.
To unsubscribe from these emails, [click here](https://discuss.tvm.apache.org/email/unsubscribe/b4ef0750a32d0e34c4c34c0cd4688e804e06895dece4d68d35294a2026e7172c).
[Apache TVM Discuss] [Application] TVM Community Survey
Posted by "Cody H. Yu via Apache TVM Discuss" <no...@discuss.tvm.ai>.
Thanks for the efforts. I left some comments in the doc.
In addition, I'm wondering if we could add a text box to all "other" options so that people can specify what they really have.
---
[Visit Topic](https://discuss.tvm.apache.org/t/tvm-community-survey/10305/3) to respond.
You are receiving this because you enabled mailing list mode.
To unsubscribe from these emails, [click here](https://discuss.tvm.apache.org/email/unsubscribe/91c6b2b2ba2f428e662f8a0d8c7609fe8bf7e4b003df5d9c99edf5828a84385c).
[Apache TVM Discuss] [Application] TVM Community Survey
Posted by "Cody H. Yu via Apache TVM Discuss" <no...@discuss.tvm.ai>.
Thanks for the efforts. I left some comments in the doc.
In addition, I'm wondering if we could add a text box to all "other" options so that people can specify what they really have.
---
[Visit Topic](https://discuss.tvm.apache.org/t/tvm-community-survey/10305/3) to respond.
You are receiving this because you enabled mailing list mode.
To unsubscribe from these emails, [click here](https://discuss.tvm.apache.org/email/unsubscribe/91c6b2b2ba2f428e662f8a0d8c7609fe8bf7e4b003df5d9c99edf5828a84385c).