diff --git a/sig-providers/meetings/039-2026-04-22.md b/sig-providers/meetings/039-2026-04-22.md new file mode 100644 index 00000000..3fd80ad1 --- /dev/null +++ b/sig-providers/meetings/039-2026-04-22.md @@ -0,0 +1,249 @@ +# Akash Network - Providers Special Interest Group (SIG) - Meeting #39 + +## Agenda + +* Provider Services Release Updates +* Gateway API Migration and Breaking Changes +* Provider Console Improvements +* Akash Home Node Alpha Testing +* Provider Release Cadence Discussions +* Provider Verification and Audit Improvements +* Provider Experience and Documentation Improvements +* Demand Growth and Provider Incentive Strategy +* VM Support and Confidential Computing Discussions +* Open Discussion and Community Feedback + +## Meeting Details + +* Date: Wednesday, April 22, 2026 +* Time: 07:54 AM PDT +* [Recording](https://4zvi6qdyy3my6utk2z6akf5djktl4dqty4nyznpx42ez337mgfla.arweave.net/5mqPQHjG2Y9SatZ8BRejSqa-DhPHG4y19-aJne_sMVY) +* [Transcript](#Transcript) + +## Participants + +* Tyler Wright +* Zeke Ezagui +* Jigar Patel +* Deval Patel +* Rodri R +* Sandeep +* Joe Deng +* Scott Carruthers +* Kenny R +* Austin Hardnett +* Damir Simpovic + +## Meeting Notes + +### Opening Remarks & SIG Providers Overview + +* Tyler Wright opened the meeting and welcomed all attendees to the Providers Special Interest Group session +* Explained that SIG Providers focuses on all provider-related topics across the Akash ecosystem +* Topics highlighted included provider upgrades, provider support, node-related issues, deployment tooling, and infrastructure improvements +* Tyler Wright reminded attendees that provider and node issues are also actively tracked through the SIG Support process and Akash Network support repositories +* Acknowledged the ongoing efforts of the Akash Core Engineering team in triaging and resolving provider-related issues +* Community members encouraged to raise additional agenda items or provider concerns during the meeting + +### Provider Services Release & Gateway API Migration + +* Tyler Wright handed the discussion over to Zeke Ezagui to provide updates regarding the latest provider release +* Zeke Ezagui announced that provider-services version `0.12.0` is nearing final release +* Major feature highlighted was the implementation of the Gateway API feature +* Gateway API migration described as a major breaking change affecting provider ingress handling +* Existing ingress implementation deprecated because Kubernetes officially discontinued support for the current implementation +* Zeke explained that future security issues and maintenance challenges would no longer be addressed under the deprecated ingress approach +* Migration documentation currently being finalized to assist providers during the transition process +* Official upgrade announcement and migration instructions expected to be shared through Discord provider announcement channels +* Tyler Wright strongly encouraged all providers to ensure they have the provider role in Discord to receive critical release announcements and updates + +### Provider Console Updates + +* Jigar Patel shared updates regarding Provider Console improvements since the previous meeting +* Provider Console UI updated to better align with the Akash Console design system +* Multiple bugs addressed including pricing-related issues and provider usability improvements +* Provider Console upgraded to support the latest provider release version +* Ongoing work initiated to support migration away from NGINX ingress in alignment with the Gateway API changes +* Additional fixes underway for Calico-related networking issues identified during testing and deployment + +### Akash Home Node Alpha Program + +* Jigar Patel announced the launch of the early alpha testing phase for Akash Home Node +* Early access participants with RTX 4090 and 5090 GPUs have already begun onboarding into the program +* Alpha testing focused on validating end-to-end architecture, latency, GPU performance, encryption, and deployment reliability +* Tyler Wright shared the Home Node registration link with participants during the meeting +* Community members with compatible GPU hardware encouraged to join the alpha testing program +* Dedicated private Discord channels established for Home Node testers to coordinate support and provide feedback +* Tyler Wright encouraged community members to direct interested users toward Home Node onboarding resources and support channels + +### Provider Release Cadence Improvements + +* Tyler Wright shared internal updates regarding efforts to improve provider release workflows and engineering cadence +* Zeke Ezagui and Scott Carruthers acknowledged for leading efforts to improve release coordination and development efficiency +* Current provider release process described as inconsistent due to varying implementation timelines and testing cycles +* Core Engineering team working toward establishing a predictable and structured provider release cadence +* Goal is to provide providers and Provider Console operators with more visibility and consistency around upcoming releases +* Tyler Wright encouraged providers to continue monitoring provider announcement channels for future updates regarding structured release scheduling + +### Provider Verification & Audit Improvements + +* Tyler Wright announced that the Core Engineering team is working on improving the provider audit and verification process +* Current provider audit process described as highly manual and resource intensive +* Zeke Ezagui, Scott Carruthers, and additional contributors acknowledged for ongoing efforts maintaining the existing process +* Engineering team currently exploring ways to automate portions of provider verification and audits +* Planned improvements expected to make provider verification more robust and scalable +* Draft specifications for the new verification process expected to be shared publicly for community feedback +* Tyler Wright emphasized that improving provider verification is important for deployer confidence and overall network reliability + +### Provider Documentation & Community Contributions + +* Tyler Wright encouraged providers and contributors to continue improving provider documentation +* Mentioned that documentation remains fully open source and community members can contribute through pull requests +* Community members encouraged to submit suggestions improving provider onboarding and usability +* Discussion included mention of Discord support bots intended to improve provider support workflows and reduce friction for new providers +* Tyler Wright emphasized that improving provider documentation, playbooks, and tooling remains a major ecosystem priority + +### Demand Growth & Provider Incentive Strategy + +* Sandeep raised a question regarding long-term strategies for attracting data centers and providers onto the network without heavy reliance on community pool incentives +* Tyler Wright explained that the Akash ecosystem is increasingly seeing demand-side growth from deployers and enterprise users +* Core Engineering and ecosystem teams focused on matching provider supply with growing demand across the network +* Tyler Wright noted that some data centers are independently discovering Akash and onboarding resources organically +* Explained that the historical Provider Incentive Program played an important role in early network growth +* Future provider growth strategy expected to rely less on direct incentive programs and more on organic demand expansion +* Emphasis placed on improving onboarding simplicity through Home Node, Provider Console, and documentation improvements +* Tyler Wright stated that long-term growth depends on making provider onboarding efficient and scalable + +### VM Support Discussions + +* Community discussion included updates regarding VM support on Akash +* Tyler Wright explained that VM implementation is currently in the specification and planning phase +* Mentioned that historical testing heavily utilized KubeVirt-based approaches +* Final implementation direction not yet finalized +* Tyler Wright emphasized that VM support is considered a high-priority initiative for the Core Engineering team +* Expected roadmap updates and implementation timelines to be shared in the near future +* Community members encouraged to participate in future specification reviews and discussions once drafts become public + +### Confidential Computing & Kata Containers + +* Discussion also covered confidential computing and trusted execution environments +* Tyler Wright mentioned that Kata Containers and confidential computing specifications are actively being explored +* Scott Carruthers acknowledged for previous testing efforts related to confidential computing +* Tyler Wright stated that many AI startups have expressed strong interest in confidential computing features +* VM support expected to arrive before confidential computing implementation +* Core Engineering team currently setting up testing environments and evaluating implementation approaches +* Tyler Wright confirmed that specifications and roadmap updates related to confidential computing will be shared publicly for feedback + +### Closing Remarks + +* Tyler Wright thanked all participants for attending the meeting +* Continued encouragement given to providers and contributors to participate actively through Discord support and provider channels +* Community support contributors and Akash Vanguard members acknowledged for assisting with provider support +* Tyler Wright emphasized that community-driven support helps free Core Engineering resources for roadmap development +* Reminder shared regarding ongoing Home Node alpha registration for users with compatible GPU hardware +* Meeting concluded with encouragement for continued ecosystem collaboration and provider feedback + +## Action Items + +### Zeke Ezagui + +* Finalize provider-services `0.12.0` release +* Publish Gateway API migration documentation +* Share official provider upgrade announcement through Discord + +### Jigar Patel & Deval Patel + +* Continue Provider Console improvements and bug fixes +* Complete migration support away from deprecated ingress implementations +* Continue Calico networking issue resolution +* Support Home Node alpha onboarding and architecture testing + +### Tyler Wright + +* Share provider release cadence updates with the community +* Publish provider verification and audit improvement drafts for feedback +* Continue coordinating provider roadmap communication +* Share VM and confidential computing specification updates when available +* Encourage provider feedback and documentation contributions + +### Community Providers & Contributors + +* Upgrade providers to latest provider-services release once announced +* Review migration documentation and Gateway API requirements +* Participate in Home Node alpha testing if hardware requirements are met +* Contribute documentation improvements and onboarding suggestions +* Provide feedback on provider verification, VM support, and confidential computing proposals +* Continue participating in provider support discussions through Discord + +## Transcript + + +Tyler Wright: All right, welcome everybody to SIG providers monthly meeting. It is Wednesday, April 22nd, 2026. As we always do during the special group for providers, the Akash community comes together to talk about all things related to the Akash provider. that includes provider upgrades. we just got out of a meeting six support where again any issues related to the provider. we can also talk about them here but during that six support meeting again all issues related to provider a node live inside the support inside the AOS network organization. + +Tyler Wright: the Akash core engineering team spends a great deal of time monthly to triage those issues, talk about those issues with members of the community. Obviously can do it async as well. but that is a public forum to talk about again provider and node related issues. During this call again we can talk about any provider related issues. I will hand the floor over to Zeke to talk about a recent provider update that all Akos providers should be updated to if they haven't already. + +Tyler Wright: And then we'll also get an update as we always have from Jigger and Dval about provider console as well as Akos home node which is a new product from the overclock labs team that is again built as an abstraction layer for those trying to build clusters of right now 5090s and 4090s but the idea is other clusters of GPUs can be brought together on the network more easily than maybe it is for some folks that are less technical. If anybody has any other items that they would like to discuss as a part of sick providers, feel free to either raise your hand or just drop a note in the chat and we'll make sure it gets All right. + +Tyler Wright: Again I will hand you over to Zeke first to talk about the provider announcement a little bit and then from there I'll hand it over to Jigger and Deval want to talk about any updates as pertains to provider console and anything else they want to talk about as pertains to the threshold note. + +Zeke Ezagui: Thanks Ty. So yeah, we are finalizing the release of provider services 0.12.0 which includes the gateway API feature. this feature is very much a breaking change. It completely changes how providers deal with ingress. And we're doing this because the Kubernetes SIG discontinued and decommissioned our current implementation. And so if there's any future security issues or anything like that, it won't be fixed. And so right now we're just doing some final finalizing the migration docs because it is a bit of a process. and then I should send out an announcement later today with instructions to upgrade + +Tyler Wright: Again, look out for announcement instructions inside Discord in the provider announcements channel. Just as a reminder, if anybody is a provider and does not have the provider role, I would suggest getting the provider role because you can make up to date on all provider related announcements and efforts. Cool. If there's no questions there, then I'm going to hand it over to Jigger and Dval to give us some updates potentially on provider console cashall note and whatever else they Yep. + +Jigar Patel: Thanks So from the last meeting we have updated our UI on the provider console so it matches the play console design. we also fixed a couple of bugs. the pricing skipped and obviously we have given the upgrade to the latest minute. there is a also work will be commence on moving away from engineering ingress as zig mentioned it will be supported on the pro console as well and yeah we are fixing couple of issues with the calico + + +### 00:05:00 + +Jigar Patel: that we found. but yeah, that's all from our console. for the home node, we launched our early alpha for testing purposes. So we are the people who register for our early access. We have sent them email if they have 40 or 50s to come and register on home not side. we are onboarding couple of users right now. They have some issues. We are fixing them. but yeah other than that the purpose of this alpha is to test the whole architecture and end to end from the u latency to GPU performance to the encryption and whatnot. So yeah that's all from one side. + +Tyler Wright: Thank I just dropped the link to home If anybody has again 5090s or 4090s and they want to participate in this alpha testing experience please feel free to sign up there. There will be some emails sent out. There is a private channel in Discord for home node u kind of beta testers during this time. So again, if you see anybody in Discord asking about a home node and not that channel, feel free to reach out to me or Shimpa or a Kash Vanguard and we can make sure that they're added to that channel for again continued conversation. Cool. Is there any other topics that anybody wants to make sure it gets discussed? + +Tyler Wright: If not, I'm going to give two pieces of alpha that probably people don't want me to give, but who cares? I'm my own man. as we look forward, I know that one thing that the core engineering team from Overclock Labs has been trying to do is improve obviously development life implementation cycles speed them up and also start to create a cadence around some efforts. big shout out to Zeke Scott who are on this call who have been really leading this effort but this may not be happening in the next couple of months but just again this is why their role is so important. The intent is for there to be a consistent cadence around provider releases. + +Tyler Wright: So again right now as things get worked on and issues get resolved and imple and provider gets tested releases come out every so often. The idea is to again provide a little bit more structure around that so that people know when releases this will be better for provider console again for folks that are managing the providers. So again, over the next couple of months, please look out for make sure that you're looking at provider announcements as again the core engineering team again with support from folks in the community will look to get out a consistent cadence schedule for provider releases. + +Tyler Wright: Another thing that the Akash core engineering team is working on is a spec around improving the audits provider verification process so that it's more robust geared in some way shape and form there's a number of ideas that again the engineering team has been exploring that'll be shared as a draft soon + +Tyler Wright: So really looking for community feedback there. Again, as things stand right now, it's a very manual process. Big shout out to Zeke and others, for all that they've been doing. Shimpa Scott has been a part of the process in the past. been very manual. U, but again, as we continue looking forward, that process will be more automated and a lot more robust. for the sake of deployers on the network head providers. So, I just want to bring that to people's attention for something to look out for, something to, share with folks in the community if they're like, "Hey, running into issues. + +Tyler Wright: Hey, I don't know when release these things are being actively thought about, worked on by the core team amongst many other efforts like bare metal support, that again we discussing six support, but are all on the road map that are being worked on that are also Cool. Is there anything else that anybody wants to make sure gets discussed as it pertains to the Akash provider that hasn't been talked about already? All right. If not, then certainly we can end this call early. again, if anybody has any issues, please feel free to put them in the use providers channel. + + +### 00:10:00 + +Tyler Wright: If anybody has any documentation improvements also feel free to create a PR documentation is open source so anybody can go on the website and add to the documentation again there's a number of bots u or there's a bot that's being added to discord and again its intent is for all sides of the marketplace but really if there's any questions or anything as it pertains to the provider experience that can be made + +Tyler Wright: more easily understood whether it's with provider console their playbooks whatever the case may be please feel free to make those suggestions so that those things can be added Cindy if I saw a question I was just curious how are we going to data centers and ask them to join AOS and stay longterm with us as we don't have much funds in community pool one of the things that I can say right now is There are a number of folks on the deployer side on demand side that are really looking for a lot of resources and have tested out looking to use it more at scale. + +Tyler Wright: those kind of opportunities where there is significant demand make it easier to bring on data centers whether we have relationships with them or as Zeke can testify there are now data centers that have heard about Akash and are just bringing on resources on their own in small scales and as demand continues to increase will scale those resources up. So right now I know the overclock labs core team is really without the community pool trying to do some matching on the demand and provider side. So that again the kind of provider incentives program. + +Tyler Wright: While it did a great deal to bring on a lot of resources and again bring on a lot of new customers that are using a lot of these resources to the network as we continue to scale the ideas that the PIP program does not have to exist as it has historically. and so again just a combination of matching and just people seeing that there's growing demand in the network for certain types of resources. And again really the focus around Kosh home node continue work with provider playbooks documentation is to make it as easy for data centers and individuals with these kinds of resources to bring them onto the network quickly and efficiently. Pardon me, I was just sneezing. + +Tyler Wright: But yeah so short answer to the question is the idea is as the core team is thinking about it is the PIP program as it has existed historically is not how it exists in the future and I think it served its purpose in kind of pushing demand and the idea is to be able to bring on supply a little bit more organically because of growth in demand. But great questions, Cindy. If anybody has any ideas about again ways to improve the provider experience, we talked about provider audits and provider verification improvements. + +Tyler Wright: Again the core engineering team as well as other folks are all constantly trying to improve the provider as a whole from stability to teachers. there's a lot being worked on but if anybody has any additional ideas this is the platform and place to have those conversation All If there's nothing else then again I hope everyone has a great rest of their day. There is a working group for the website and documentation tomorrow and then we'll have a steering committee next week as well as I think sik economics where maybe we'll talk a little bit more about u pipesque programs. but again we're really looking to just push on demand. Hope everyone has a great rest of their day. Again please feel free to reach out in Discord. + +Tyler Wright: And regarding VM, I think this would be the call to ask. Again, that is very much in the spec for process right now. as the core engineering team is kind of outlined, just trying to finalize the spec and implementation direction. I think implementation is not going to take that long, but this is a high priority for the core engineering team. so this is something that should be moving in progress on the progress boards for sometime very soon and is tracking I think we're going to update the timelines on the road map but really looking to get in the next couple weeks. I don't have that information. + + +### 00:15:00 + +Tyler Wright: I know that historically we were testing a lot with Cubevert, but I don't know what the final spec is directions for VMs, but again once that spec is out, I'll make sure that gets dropped in the providers channel so that if anybody has any questions they can make those things heard at that point. But as Scott mentioned, I think it's most likely going to be Cubert, but nothing gets finalized yet. I know a lot of testing has been done historically to see what's possible leveraging Cooper. So any other provider related questions again we'll continue to track this effort Sep thank you for bringing up we can continue to track this between meetings using sick providers channel. Cool. If not I hope everyone has a great rest of their day. + +Tyler Wright: Hopefully you all confidential computing. Yeah, again this is another spec yes the kata containers is being worked on again I think its hasn't been finalized yet. There was an addition original spec u but there's going to be some additional work. I know that again the core engineering team is setting up some environments to do some testing. + +Tyler Wright: Scott has done some testing historically for a period of time. and this is a feature that I know a lot of AI startups are really interested in. so again, I would say VMs, going to come before TEE, but I know that the core engineering team is looking at them both simultaneously. as we as network thinks about network upgrades, etc., This is a very very busy time for the core engineering team. I'll make sure that they're all in GitHub in a repo, but I will share whether it's for provider verification provider audits confidential computing DEM support. Make sure those are shared with the community for additional feedback, questions, comments, etc. + +Tyler Wright: And then we'll look to the core engineering side. Those are key features that part of the road map that we see continue to drive demand on the network. we'll implement those as quickly as possible. All right. for the update. Thank you, Jigger and as always, again, if anybody has 59s or 490s that they want to bring on as part of testing for Home Node, please feel free to follow the form on the Home Node site to get early access. If anybody has any questions between meetings, feel free to use the providers channel. I know there's a number of folks from the core engineering team and elsewhere Akash Vanguard is also in there providing support. + +Tyler Wright: So big shout out to those providing support and really leaning on again members of the community to continue providing support so that the core engineering team can focus on all those things that have been mentioned. Thanks Hope everyone has a great rest of the day. We'll talk soon. Bye-bye. + +Rodri R: Bye + +Deval Patel: Thank you. + + +### Meeting ended after 00:19:45 👋 + +