# Spice.xyz May Update | Enterprise-Grade Performance for Web3 Data [Margaret Zhang](https://margaret-c-zhang.medium.com/?source=post_page---byline--d1604b72b3d5---------------------------------------) 6 min read · Jun 23, 2023 🔵 **Updates and improvements to Spice.xyz** — May 2023 [A few weeks ago](https://twitter.com/spice_ai/status/1653243187404603393?s=20), Spice AI completed its upgrade to **second generation platform architecture (Gen. 2)**. Today, we walk through the key performance upgrades that **empower developers building the next generation of apps** with the fastest, easiest, and smartest way to leverage web3 data. Spice.xyz now features improved developer experience in app with expandable query SQL explorer, SQL query error underline, and collapsible side bar. ## Feature of the Month: Performance Upgrade 🚀 Until now, most data platforms remain dashboard-focused and don’t prioritize infrastructure performance. Spice.xyz is **application-focused** and built specifically to address the development infrastructure gap from day one. **Developers like** [**EigenLayer**](https://www.eigenlayer.xyz/) **trust the enterprise-grade infrastructure to be** **Reliable.** >99.9% high-availability. **Scalable.** 1000+ SQL queries per second. **Fast.** Subsecond block indexing. Let’s get started **🟢🟢🟢 ↓** **🟢 Block Latency** Spice.xyz returns the most recent block accurately and in seconds. It takes just **<1 second** for core data and **2.5 seconds*** for enriched data to query availability from time we see the block — down from an already low-latency 3.6 seconds. The [next closest solution](https://www.transpose.io/blogs/how-transpose-is-different-to-dune) has a block time latency of 10 seconds, while the industry standard takes more than 30 minutes! For application developers that require real-time data, such as building transaction auditing and personalized UX, subsecond block latency is a game changer. Comparing Block Latency Performance _ *block time: average latency from when Spice sees the block in our nodes to when it is available to be queried in platform, for Ethereum, at 95% confidence level._ This also enables sophisticated DeFi models and apps to retrieve a month’s (720h) worth of ETH-BTC prices in just **1.04 seconds** using the [historical prices API](https://blog.spice.ai/spice-xyz-april-update-whats-new-21bb84646d55) at a 5 minute granularity: **🟢 Execution Speed** Querying execution time is up to **15x faster**! With Spice AI’s [multi-chain architecture](https://medium.com/spice-ai/build-multi-chain-with-spice-ai-now-with-polygon-support-9db156f36d80), queries span cross-chain and cross-dataset for ecosystem-wide insights. You can test examples in the SQL Explorer by referencing pre-written sample queries from the Spice.xyz docs: 1. Get [All NFTs from a single collection](https://docs.spice.xyz/reference/example-ethereum-sql-queries/nfts#all-nfts-from-a-single-collection) — instant, complete BAYC insights 2. Get [Cross DEX prices](https://docs.spice.xyz/reference/example-dex-queries/cross-dex) — complex, multi-source queries with ease 3. Get [Most Ethereum burned](https://docs.spice.xyz/reference/example-ethereum-sql-queries/blocks) — fast gets 10x faster 4. Get [Ethereum address of ENS domain](https://docs.spice.xyz/reference/example-ethereum-sql-queries/ens-domains) — match your data, know your users 5. Get [Sushiswap Liquidity](https://docs.spice.xyz/reference/example-dex-queries/sushiswap) — better pulse check, better moves 6. Get [Advanced Liquidity Aggregations for USD Token Pairs](https://docs.spice.xyz/reference/example-dex-queries/uniswap#advanced-liquidity-aggregations-for-usd-token-pairs) — pricing accuracy referencing min, max, average 7. Get [Confidence for Token Standards](https://docs.spice.xyz/reference/example-ethereum-sql-queries/tokens#get-confidence-for-token-standards) — simple smart contract checks 8. Get [Voluntary Exits Count](https://docs.spice.xyz/reference/example-ethereum-beacon-sql-queries/voluntary-exits) — harness the new Beacon dataset 9. Get [Last 5000 blocks](https://docs.spice.xyz/reference/example-ethereum-sql-queries/blocks) — experience the power of Spice’s Apache Arrow API 10. Get [Average Gas Used by Transactions per Block](https://docs.spice.xyz/reference/example-ethereum-sql-queries/gas-and-fees) — get smart and optimize your gas fees Accelerate accessing web3 data further by querying with SQL over high-performance [Apache Arrow APIs](https://docs.spice.xyz/api/sql-query-api/apache-arrow-flight-api). **🟢 Planet-Scale for any Enterprise** Gen. 2 handles an even larger volume of queries. Enjoy updated plan limits that can run up to **1000 requests per second and 50 concurrent requests.** These scalability improvements enable developers to retrieve data from the blockchain more efficiently and perform previously impossible analyses with complex queries. **🟢 Enabling Intelligent Applications** The rollout of Spice AI’s second generation platform architecture (Gen. 2) helps bring enterprise-scale data infrastructure to further advance the capabilities of AI/ML across chains: - more scalable platform that handles data more effectively means Spice can support **more chains, faster**. - for the first time, enhanced speed and efficiency allows developers to leverage fast AI-ready, web3 data for **inferencing.** App-builders can now ship better experiences and use cases — critical to the success of intelligent applications in web3. For more information, visit our [docs](https://docs.spice.xyz/). ## What’s New ## 1. Expandable Query SQL Explorer and Collapsible Side Bar Make the query zone larger for a complete view of complex SQL queries both in standard view and full screen mode. Simply drag the bottom of the explorer to adjust the split between your query and results. Expandable SQL Explorer and Full Screen Mode. Featured Query: Advanced Liquidity Aggregations for USD Token Pairs Expand the explorer while maintaining access to the Python, Node.js, and Go SDKs. Simply toggle views with the translucent collapse side bar tab located on the bottom left corner of the app. ## 2. SQL Query Error Underline Writing SQL just got simpler. The spice.xyz query editor underlines errors in your SQL query and suggests fixes directly in app. Compose precise, clear, mistake-free queries to get the results you need faster. You can also query in plaintext and let [Spice AI Flow](https://blog.spice.ai/introducing-spice-ai-flow-8bcfc5cc2294) autocomplete with a SQL query. Developers will receive a red error notice in Results and a red underline indicating the error location in the query. > _🔵_ **_Around the corner_** > _Two highly requested datasets, Ethereum Wallet Balances and NFT Metadata, will be released out of preview._ > _The 2023 Season 2 Flow Hackathon is_ [_LIVE_](https://flow.com/hackathon) _.