From Jito to Solayer, an in-depth analysis of how SVM reshapes the Solana scaling game

Let's talk about $LAYER@solayer_labs, which has performed very well in the secondary market recently. In particular, why has the InfiniSVM technology roadmap attracted so much attention? What are the characteristics of the hardware-accelerated SVM expansion solution? And how will the Solana expansion ecosystem industry pattern after hardware acceleration break through? Next, let me talk about my prospective observations:

1) Different from the horizontal expansion route dominated by Ethereum in the past, the Solayer team presented a completely different expansion idea in the infiniSVM white paper: deeply optimizing SVM through hardware acceleration to create a million-level TPS blockchain network, which is essentially an expansion solution that deeply integrates hardware + software.

Looking back at the entire history of blockchain expansion, the earliest on-chain expansion idea was to achieve it through parameter adjustment (larger blocks, shorter block time), but this idea easily touches the blockchain impossible triangle dilemma; the layer2 expansion idea that emerged later was a horizontal expansion, the core purpose of which was to divert transactions through layer2 (state channels, side chains, Rollup, etc.), which would undoubtedly sacrifice some global atomicity; and the hardware acceleration route explored by InfiniSVM is a new route that upgrades the expansion concept, while maintaining a single global state, to break through the performance bottleneck with the help of specialized hardware;

Simply put: InfiniSVM's expansion method does not lie in simply optimizing the algorithm, but in reconstructing the SVM execution environment through microservice architecture and hardware acceleration, completing some key tasks on dedicated hardware, and then achieving atomicity and consistency in the global state under high load conditions.

2) Following this line of thought, many friends must be wondering why Solana's SVM execution environment requires hardware acceleration? From the data provided by the Solayer white paper, it can be seen that the current Solana verification node already requires a CPU above 3.1GHz, 500GB+ high-speed memory and 2.5TB+ high-throughput NVMe storage. Even with such a high configuration, the CPU utilization rate is only about 30% under high load, and P2P communication is close to the 1Gbps bandwidth limit of consumer-grade networks.

The question is, if the CPU is not fully utilized, why do we need more powerful hardware? This actually exposes that Solana’s current performance bottleneck lies in other aspects besides CPU computing power, such as: microservice processing architecture, which can isolate different processing links and match more suitable hardware resources; dedicated accelerators, which can assign specific tasks such as some signatures to dedicated hardware, etc.

You see, InfiniSVM is not a simple hardware upgrade, but a redesign of the entire execution environment, providing more specialized hardware optimization solutions for each bottleneck link. For example, to improve the production efficiency of a workshop, it is necessary to redesign the software and hardware of the entire production line instead of simply increasing the number of workers.

3) So, what features of InfiniSVM’s hardware acceleration solution are worth sharing?

1. Distributed microservice processing architecture. Previously, Solana's single transaction processing process could be decomposed into multiple extended processing links, such as signature verification, deduplication, scheduling, and storage. In the InfiniSVM architecture, each link can be processed independently, avoiding the big problem of "one link is stuck and the whole line is waiting";

2. Intelligent transaction scheduling system. Originally, when Solana read and write transactions, they had to queue up for processing if they belonged to the same account. However, InfiniSVM has achieved that even under the same account, operations will not interfere with each other, which greatly improves its parallel processing capabilities. To put it bluntly, it further enhances the refined management capabilities;

3. RDMA low-latency communication technology. Normal communication between nodes requires at least necessary steps such as packaging, delivery, and unpacking. RDMA can directly transfer the data of one node to the memory of another node, achieving a breakthrough in communication technology from milliseconds to microseconds, which can significantly reduce state access conflicts.

4. Distributed intelligent storage network. If a single account is allowed to take over its account data, Solana used to be limited to 10MB. However, InfiniSVM adopts a distributed cloud storage solution, which allows data to be dispersed to different nodes and marked as fast lanes, slow lanes, etc. This not only breaks through capacity limitations, but also optimizes data access speed.

4) After explaining the technical upgrade route, many people may have asked, "What's the use?" In general, with the support of hardware acceleration, Solana's competitive advantage in the layer1 competition can be further enhanced. Moreover, compared with Ethereum layer2, which needs application data support in the ecosystem to show its expansion performance, this performance breakthrough of millions of TPS achieved by hardware may only require the access of very few vertical scenarios to directly prove it, and the implementation path is shorter;

Let's change our thinking. Taking @jito_sol as an example, as the Solana MEV infrastructure, its niche value in terms of transaction sorting optimization, MEV extraction, and validator benefits. If it could not be proven before the Solana MEME craze, after the MEME craze in the past year, Jito's way of optimizing the transaction system has become an indispensable part.

In fact, Solayer's current technical position is similar. Its upgrade method of making transactions perform better in the system is not obvious in simple financial transaction scenarios, but if we think about the large-scale implementation of PayFi in the future, if Solana wants to perfectly carry the high-throughput and low-latency payment settlement infrastructure functions, the performance of TPS will be clearly perceived. In addition, there are DePIN ecology, complex chain games, AI Agent application scenarios, etc.

Anyway, it is easier to see the value of a technical infra project by looking one step ahead than just looking at it from the perspective of whether it is currently useful.