Danny hosts Carlos from Blockworks research and Sam Schubert, making his first appearance on a Blockworks podcast. Both will be presenting at the upcoming Digital Asset Summit in New York (March 24-26) alongside the launch of Blockworks' new Solana Lightspeed IR platform.

The conversation centers on block building dynamics within Solana's ecosystem, specifically the ongoing competition between Jito's BAM client and Temporal's Harmonic solution. This technical battle has significant implications for market structure, with different block packing approaches creating inconsistencies that affect market makers, applications, and end users.

The discussion examines how these competing approaches impact transaction ordering, fees, and overall market microstructure, while exploring Solana's roadmap toward Multiple Concurrent Proposers (MCP) as a potential long-term solution to current fragmentation issues.

The Block Building Battle: Jito BAM vs Harmonic Architecture

Jito BAM uses a network of nodes with TEEs (Trusted Execution Environments) to keep transaction ordering private until execution, defaulting to priority fee and CU ordering while enabling future application-controlled execution through plugins.

Harmonic operates as an aggregation layer sitting on top of multiple block builders, allowing validators to choose blocks based on preferences like maximizing revenue or minimizing toxic MEV exposure.

"The fundamental thing is that block building has a direct impact on that market structure" - Danny, highlighting how different packing approaches create operational challenges for market participants.

Market Structure Impact and Performance Metrics

Jito defines optimal block building as streaming non-vote transactions continuously across 64 ticks, while some validators cram transactions at the end of slots, undermining Solana's streaming design.

Oracle update win rate analysis showed Humidify had significantly higher success rates under Harmonic blocks during the first two weeks of 2025, raising questions about transaction ordering advantages.

Block build time data reveals Harmonic blocks consistently ran above 400 milliseconds in epochs 909-919, though recent improvements have reduced this to around 400ms with a 50ms gap versus BAM.

Axiom users face network fees that are "orders of magnitude larger" than other applications, suggesting possible vertical integration concerns and side deals - Sam.

Vertical Integration Concerns and Market Dynamics

Temporal operates three potentially conflicting services: Harmonic block builder, Humidify prop AMM, and Nosami transaction landing service, creating incentive alignment concerns.

"If they own these three parts of the stack, you can imagine that incentives are sort of biased toward... Humidify can benefit from Nosami, and they can also benefit from the way that harmonic orders blocks" - Carlos.

Harmonic's market share has grown from 3-4% to approximately 16% of stake, while both Harmonic and BAM remain closed source, limiting verification of their claims.

The Path to Multiple Concurrent Proposers (MCP)

MCP requires Alpenglow as a prerequisite, targeted for August 2025, with MCP implementation expected 6-12 months later, providing consistent transaction sequencing rules across all slots.

"Having this process essentially a lot more centralized and having a centralized actor determine what is optimal for the network would be the best process" - Sam, advocating for in-protocol solutions.

MCP will use first-come-first-serve ordering with priority fee per CU for conflicting transactions, eliminating the current inconsistency where "it can be a different rule set every single block."

The end game includes Application Controlled Execution (ACE), allowing applications and market makers more control over sequencing while maintaining protocol-level consistency.

Urgency and Business Model Implications

"Solana can't afford to have the same problems with market structure for that long. Otherwise, the lead that other apps like Hyperliquid Perps have gets so big that they can't come back from it" - Carlos.

Current block building business models face existential challenges once MCP launches, though teams are "well aware of what is happening and they're going to continue to innovate" - Sam.

New perpetuals protocols are building custom sequencers or sidecar processes to bypass native Solana limitations, creating a race between protocol improvements and application workarounds.