Atrix SRM Grant Proposal

Hello Serum community! kaiba from Atrix here with a SRM grant proposal.

I’ll start with a brief introduction of Atrix since we’re such a new project in the Serum ecosystem.

Atrix is a fast growing AMM and farming platform that allocates the vast majority of its liquidity to Serum without any internal pools — the first of its kind in the Serum ecosystem. Currently, almost 95% of liquidity in Atrix is locked in Serum, with further improvements to increase this currently in development.

We chose to go all-in on Serum because we truly believe that fragmented liquidity simply doesn’t make sense in today’s decentralized, composable world, and that liquidity should instead be shared in one location. We imagine a future where all assets are traded on a orderbook, with fully orderbook-based AMMs powering initial price discovery and liquidity for the long-tail of assets.

Atrix has been a key component in Serum’s recent steep rise on the DeFi TVL leaderboards, with almost $500M being contributed by Atrix to Serum at the peak, which deservedly garnered amazing support for Serum from the broader DeFi community.

We’ve also seen rapid adoption of Atrix’s new permissionless pools and no-code farms, with eight pools and farms being launched permissionlessly by the community less than a day after launch. We envision Atrix as the platform where most Solana projects launch and incentivize initial liquidity for their token, which then increases the network effect and liquidity of Serum.

To add, Atrix’s swap interface does not charge any extra fees on top of Serum’s default fees and is fully Serum based, driving 100% of Atrix’s swap volume to Serum. With further product improvements and ecosystem growth, we believe swapping on Serum through Atrix will provide the best price for users on all available pairs, eventually making it one of the top volume drivers in the ecosystem.

With this in mind, we propose a grant of 1.5M SRM out of the 5M SRM tokens available for the first 30 days of the grant cycle. Using SRM as liquidity incentives, Atrix will increase awareness of Serum and grow its TVL to $1B and beyond, contributing almost 95% of it to Serum as ecosystem liquidity and TVL. We think this is a very achievable target, since Atrix currently has over $350M TVL without any direct incentives from Atrix.

We propose the following Atrix pools to receive SRM emissions at a target APY of 15-20%:


With other partnerships in the works and multiple Atrix ecosystem projects currently in stealth, we think Atrix is in an extremely strategic position to build long-term value for the Serum and Solana ecosystem.

1 Like

Yo here cus i saw a twitter thread on this grant and clicked in to see what was up… looked a few posts and have a couple of immediate questions

TVL metrics look decent , especially if the they are all going to serum, but about 150m out 350m are in msol-sol and usdt-usdc, right? If I’m not mistaken, based on your docs, you run a K = XY equation… based at stable swap numbers, wouldn’t that mean that the majority of your current TVL is passive liquidity and basically doing nothing? And one of the pools you’re recommending would be just that.

Circling back on this cus I just saw the Raydium post… I’ve been using Solana AMM’s for a bit and know you’re a new protocol, hats off to you, but a little suspect of the claims your making having only been live for a few weeks. Not trying to hate, but provided some LP to your SRM rewards pool… was that a Serum grant?

And if what this says is true… prices way beyond the mid price… it seems like you really are just looking for rewards at this point. Any order that fills would probably just fill into another market maker IMO

1 Like

Thanks for the great questions!

We’re working on developing and adding the stableswap equation that Curve and Saber use to our stable pools, as we definitely agree we could be more capital efficient in that regard.

We are not directly incentivizing any pools/farms yet at Atrix. All farms you see currently are permissionless farms created by the community. For example, of course Atrix is not the one providing 151,601 MNDE / day for the mSOL-USDC pool. That was created by the Marinade Finance team themselves, permissionlessly.

Screen Shot 2021-10-15 at 6.21.30 PM

I just posted a response to Raydium, might help clarify your questions regarding this - Raydium SRM Grant Proposal - #13 by kaiba

Wanted to give an update that we created Serum’s first governance proposal which is currently live here - Governance | Solana

We’ve increased our grant request to 2.5M SRM as we’ve gotten a lot more traction from the community through new and upcoming partnerships and farms since when we first posted this proposal. We’d also like to propose SRM rewards for a wildcard pool that we didn’t mention in the initial proposal, but which was offered in Serum’s initial announcement. We think LINK-USDC could be a great pool to bring liquidity into the Serum/Solana ecosystem from other chains.

Edit: We’ve created a new on-chain proposal to reset the timer and allow community members to read our updated proposal

The Raydium team is strongly opposed to giving this grant to Atrix. We feel that this team has been deceptive to the community and it should be discussed.

  1. Atrix boasts about it’s TVL on the orderbook and AMM but they have not been helpful to Serum with their usable liquidity as we stated before. After their market making tactics were pointed out, they changed the mSOL-USDC pool to be more effective and made a post to show it but they didn’t make updates on any other pools for some reason. As of today all the pools aside from mSOL-USDC still only have 3 orders on each side at 0.5%, 10% and 1000x spread from the mid price. If they can change one pool to be helpful to Serum in a timely manner, there’s no reason they would not do it for every other pool unless there is something being hidden.

  2. Launching a second token (Almond) for the sole purpose of bolstering TVL. This is usually a pretty big red flag when doing due diligence. Almond has no roadmap, no team members listed, no whitepaper or description of what the tokens do except boost TVL in Atrix even though Atrix is also planning to launch their own token. They say they have close ties to each other and Almond was launched days apart from Atrix and only supports their LP tokens. For a staking pool, supporting other LP tokens is just entering a token address. Since launch, Almond has lost over 80% of its value and cost its buyers millions.

  3. They tried to pull a fast one on the community by putting in the DAO proposal for 2.5M SRM when they originally asked for 1.5M in the proposal. After being called out for it, they cancelled the proposal, added another reply to their forum post and put up a new DAO proposal for 2.5M tokens.

These actions put together point to a dishonest team who are trying to loot the Serum treasury for their own means. To start with, 2.5M tokens (19M USD value) is a lot to give to a new project with no proven track record. It is even more for a team that has been shown to be deceitful to the community. Passing this vote would set a dangerous precedent for governance that any project could come in and take millions of SRM without doing anything to help the foundation.


The reason Atrix attracted that much liquidity is Almond. The reason people locked the amount of liquidity in Almond is alm rewards. I don’t think that Msol is a decent example of cooperation as Marinade is heavily promoting itself on every solana amm/lending platform. Being in both Discords I suspect that Atrix and Almonds teams are in tight connection. No proofs tho.

  1. Each pool has unique characteristics depending on the activity of the Serum market, liquidity available in the pool, and any wrapping/unwrapping of wrapped SOL accounts, so it’s not as simple as deploying the same change the change to other pools (though we did improve mSOL/USDT’s and SRM/USDC’s orders many days ago). For example, canceling all orders for the SRM/USDC pool requires more compute than mSOL/USDC, since the market for SRM/USDC has much more orders. This means that we have to create a new method for placing more orders instead of plugging in our default one.
    To comment again on the bit about 1000x spread, those orders are according to x*y=k and are not due to any manipulation on our end. If tail-end liquidity was not counted as TVL, Uniswap V2 and Sushiswap would have 10x less TVL than they claim. To clarify, those orders have NO benefit to Atrix, and are SOLELY benefitting the TVL of Serum. To add we are also working on more capital efficient ways to allocate this liquidity, while also providing it to Serum TVL. But if the Serum community decides that they would rather have 2 more orders closer to mid rather than 100s of millions in extra TVL, we can easily make that change.
    I agree that we should be moving towards this in faster fashion, but since our team is very small, our development bandwidth is very limited. As the one with the most experience with order placement, I was personally caught up with putting out fires due to our very fast growth, fixing bugs with our frontend and encountering issues with our transactions, as evidenced below. Since these issues directly affect our users’ experience and our current Serum placement works (though could be better) we decided to focus on the immediate issues at hand before improving Serum orders. If the Serum community wants, we could drop any UI/UX/QOL related issues and focus solely on order placement.

  2. We’ve said this before, but Atrix and Almond are not the same team. There was also speculation in prior weeks that Saber and Atrix was created by the same team, which Saber team members had to publicly deny. We understand why this speculation could come up, but to put it as a fact is very disingenuous. Almond wanted to partner with a new and unique project in the space for a while after launch, and Atrix seemed like a good fit for that description. There’s nothing we can say on our end about Almond’s token price, other than this is crypto and token price is very volatile, especially with yield farming projects.

  3. This was definitely a mistake on our end to not update the forum post when we created the governance proposal, though we did update the text in the governance proposal when we submitted it. Subsequently, we resubmitted the proposal with an update on the forum to reset the proposal timer so users had time to digest it, and we resubmitted again after the Serum team suggested we do since they were hosting a Twitter space to kick off governance. This was a learning experience from us and in retrospect, there were definitely things we could have done better.

Overall, we would very much disagree with the fact that we have been deceitful to the community, if anything, have simply been overloaded with our fast growth and relative inexperience with governance processes. As I said, if the Serum community would like us to commit in one direction rather than another, we are very open to that, to the extent that we would even be open to contractual legal commitments if decided on by the community.

It’s really hard to instill faith in this discussion when Almond says something very different.


The wording I posted above might have been unclear – I meant that Almond started by partnering with only a single project for a duration of time after they launched (“for a while after launch”), but they also eventually planned on supporting other projects. This is also supported in their article, where they say they will be supporting other tokens in the future. The initial singular relationship between Atrix and Almond made sense for Atrix too, so we supported them where we could. We’ve also been very transparent about the relationship between Atrix and Almond in our Discord server.

I am against the proposal for the following reason:
1.) This does not need to be a year grant. Atrix is a new protocol, and the metrics of how much the TVL is beneficial to Serum are still unknown. 2.5M SRM($20m) is a massive ask for a relatively unknown protocol. Others have pointed out as well that Atrix Amm methods can use a lot of improvement. It is better to have a bi-weekly or a monthly proposal grant instead of a yearly grant. We can revisit every month to have more metrics like how much volume Atrix contributed to Serum and average price spread to make a better decision.

2.) Atrix does not have any incentive for users to provide liquidity on its platform except for the Serum referral GUI fee, which is minimal ~0.06%.
This means the only incentives are Almond and MNDE. If Atrix receives this grant, the primary motivation for people to use Atrix will be SRM emission which means people will farm SRM until it runs dry and move on to other farms. This approach is not a sustainable strategy long term. Thus it is not suitable for the Serum ecosystem.

3.) Instead of a 15% - 20% APY target, There should be a more specific SRM emissions number such as 1SRM per $10,000 TVL. 15% - 20% APY is not detailed enough and can be hard to track how the funds are actually being used.


Some interesting user activity in this thread. Anyways:

  1. It’s our understanding that this grant will be revisited every month based on various metrics. We’re also already updating many pools’ order placement to be very competitive, like SRM-USDC. We encourage the community to keep track of Atrix pools’ improving order placement over the coming weeks.
  2. Atrix currently charges a 0.20% spread fee on our orders, which accrues to Atrix LPs. Any orders that are hit generate revenue for our users, just like a normal market maker would generate revenue on an orderbook through the spread. As we continue to improve our order placement, we plan on reducing this fee further, so Atrix orders are hit more often on the orderbook, leading to more revenue for LPs. We’re working on the pipelines to aggregate and display this data from the Serum orderbook.
  3. As TVL fluctuates we’d like the pools to stay competitive, so APY seems like the best metric to keep constant. But are definitely open to other metrics if the broader community would like.

I am still against the funding as of now.Let us revisit this in a month, which seems like most of the people here agree. Currently, Atrix’s track record of 2 weeks of TVL boost caused by questionable circumstances is not enough evidence for the Serum’s community to make good decisions.
It seems like many things still need to be monitored on Atrix’s side, such as how much volume Atrix attributed to Serum, how much Atrix TVL will retain after one to two months.
It is not uncommon in crypto to see a protocol TVL drop drastically once the farms dry out like many of the farms on Binance Smart Chain.
Waiting aligns which what you said as well: “We encourage the community to keep track of Atrix pools’ improving order placement over the coming weeks.”

As far as point 3 goes, APY can’t remain the same at 15-20% forever. Let’s bring up an optimistic case where $200m TVL is deposited into usdc/usdt pool, then Atrix would need around $30m or 3.75m SRM tokens to keep the 15% APY. This is not sustainable long-term and is bad for the Serum’s community. Please have more specific metrics and emission numbers.

Atrix should not rely on SRM emission to remain competitive. However, if SRM is the only thing keeping the pool competitive, then why not start emitting Atrix tokens to accomplish the same thing?

In conclusion, let’s revisit this in a month so a lot of questions can be answered and also can be sorted out. If what you said can be accomplished, I am sure the Serum’s community will be more than happy to fund you.


Jump Crypto is opposed to this proposal (and more generally to TVL-oriented grants as discussed in Liquidity Mining Program - #22 by JumpCrypto and Raydium SRM Grant Proposal - #35 by JumpCrypto). As discussed there, we think there is a high possibility for Serum to overpay for TVL, and that the bigger problem of getting AMMs to quote more tightly on Serum will remain unaddressed.

We would really like to see Atrix (or any other AMM) experiment with more aggressive formulations for order placement on Serum so that more of the TVL is concentrated at the inside than what UniV2 prescribes. For stable pairs, the stableswap formula is a known solution… but we’d also propose experimenting with an alternate formula for non-stable pairs. This need not be UniV3-style where the user has a ton of parameters to choose from; just another pool with more efficient liquidity usage.

Doing this would drive a lot of volume, allowing Atrix to benefit more from volume-oriented rewards (Volume Incentive Program) instead.

We would actually be more open to a launch-related grant (rather than TVL-related grant) to drive buzz about the partnership between Atrix and Serum, and get new yield farmers into the Serum ecosystem. For example, a grant of 1M SRM for 1 month at $8/SRM would offer 96M USD annualized, which would be a substantial boost to Atrix’s (currently 370M) APY.

After considering the feedback from the community members above, we agree that a 1M SRM grant would be more appropriate and would also leave room for other projects to apply for grants.

In response to the demand for increased liquidity closer to mid-price, we improved our order placement even further. Non-native SOL pairs now have up to 14 orders on the orderbook, with most of the orders being within 2% of mid-price, and significant liquidity being within 10% to capture and adjust for sudden large market moves. We’ve also reduced the spread fee from 0.20% per side to 0.10% per side to allow for tighter liquidity, earning more fees for LPs.

Atrix still places two constant product consistent tail-end orders to provide significant TVL to Serum, but we’re open to moving those orders to the center in sacrifice of TVL if the Serum community decides so through a proposal.

For example, Atrix SRM/USDC orders:
Screen Shot 2021-10-24 at 5.43.22 PM

For more context, here are Atrix orders are labeled with " ** " to compare to the rest of the SRM/USDC book. In this snapshot, Atrix has the best bid with huge relative size compared to the next closest order, along with other very tight orders.

We’re working on bringing these improvements to native-SOL pools, innovating further on liquidity equations for certain pools, and increasing revenue for LPs. With the imminent release of Solana 1.8 and Serum batch order placing/canceling, the future of orderbook-based AMM liquidity is looking stronger than ever.

Around 300,000 SRM is probably more appropriate considering that Atrix is a new protocol. That’s already $2.4m just for these 30 days alone, which is still a lot. Atrix can test that out for this month with the new order placement algorithm and then propose new funding next month along with more granular metrics like the total volume provided to Serum.

As far as I know, you guys just launched three weeks ago. So 1m SRM ($8m) is just too much, considering Atrix’s inability to provide specific metrics like volume, fees, etc., except for the above demo.

I like what Atrix contributes to the Serum ecosystem. Still, I have an issue with Atrix continuously lying to Serum’s community about Almond’s relationship with Atrix that they are not the same team even though they are.
I personally don’t have any problem with creating a farm as a marketing tool to boost TVL, but lying?
Atrix is asking for $8 million or 1m SRM)from the Serum’s community yet openly lies to the community it requests funding from.
Simply tracing back the address will prove that.

Here is the address of Almond’s farm owner

The address was created at this TX:

It was created by this address who coincidentally owned the Atrix program.

Just this forum alone, Atrix has lied at least 2 times, look above.
"We've said this before, but Atrix and Almond are not the same team. There was also speculation in prior weeks that Saber and Atrix was created by the same team, which Saber team members had to publicly deny. We understand why this speculation could come up, but to put it as a fact is very disingenuous. Almond wanted to partner with a new and unique project in the space for a while after launch, and Atrix seemed like a good fit for that description. There's nothing we can say on our end about Almond's token price, other than this is crypto and token price is very volatile, especially with yield farming projects. "

Even when @AlphaRay confronted, Atrix continues to double down with the lies
"The wording I posted above might have been unclear – I meant that Almond started by partnering with only a single project for a duration of time after they launched ("for a while after launch"), but they also eventually planned on supporting other projects. This is also supported in their article, where they say they will be supporting other tokens in the future. The initial singular relationship between Atrix and Almond made sense for Atrix too, so we supported them where we could. We've also been very transparent about the relationship between Atrix and Almond in our Discord server. "

There is no such thing as “creating an address at a TX”. Addresses are created offline by generating a keypair which can then be used to receive balances; Atrix never “created” Almond’s address in a TX. Atrix also never deployed the Almond team’s program for them. Literally the only thing that we did was send them some SOL to their deployment address when we were on a call with them to help them launch on mainnet. The transaction you linked confirms that we just sent them the SOL for their deployment — all of the evidence is on the blockchain on the very transaction that you linked yourself. We encourage readers of this forum to click the transaction link to view that it is just a SOL transfer themselves.

As the Almond team has said themselves, we helped Almond get started in the Solana ecosystem. We were on a call with an Almond dev to get their program deployed to mainnet, since the process is not straightforward for a new developer, and deployment costs (therefore mistakes) are very expensive. Atrix devs never deployed the program, we merely assisted in the process.

These sort of accusations cause a lot of harm. A similar situation occurred two weeks ago when a tweet went viral accusing Saber, Sunny, Atrix, and Almond being created by the same team, which was then publicly refuted by the Saber team. However the rumors still exist, and you can still find some accusations of this in replies to Atrix’s tweets.

To further clear up this misunderstanding, I’m sure the Almond team would be open to chatting with Raydium.

Hey Kaiba, I have no problem with Atrix and Almond being on the same team. But, Would you please stop lying to the community?

Do you expect people to believe that the first person to send over a transaction to a newly created KeyPair is not the owner? Literally, the first transaction.
Your story keeps changing every time people confront you.
How many lies are going to be enough?

I was doing a little more digging and found another evidence:

The same Github account, theo-almond, commits both Atrix’s and Almond’s codes on Defillama. So why would a team upload codes to calculate TVL for another team?

The pieces of evidence are stacked that you guys are the same team.

1.) The only thing that benefits from Almond is Atrix TVL, evident by exclusively listing Atrix’s pool.

2.) Almond’s address is funded by Atrix’s

3.) Both Atrix and Almond share the same GitHub account.

Atrix relies on Almond’s for its TVL, and Almond’s farm is drowning, and Almond tokens are diluted day by day. Atrix knows this, so it needs SRM emissions to survive.

Please stop lying

Hi everyone, theo from Almond here!

So here Kaiba did help with deploying the program during a call. I asked Kaiba to send me funds to an Almond dedicated wallet that I generated in Phantom to help future deployments. The seed phase I use for my wallet is different than kaiba’s or anyone else’s, so it’s not possible for us to generate the same keypair.

About the DefiLlama stuff, I was chatting with Alameda soon after launch time and they requested that Almond and Atrix be added to DefiLlama before they put in funds, so that it would be reflected in Solana’s total TVL across Almond, Atrix, and Serum. Because of wanting to onboard Alameda, I said that I could create the Atrix and Almond adapters myself asap since and they said that would be great. Almond already internally called Atrix’s api, and I was familiar with it because I worked in their codebase. I’ve asked the person I was in contact with at Alameda if I can share screenshots.

It makes sense that you see us as one team that created two projects due to the close collaboration. It’s all subjective on where you draw the line on one team or two teams working together.

I can answer anything else too :sweat_smile: