Forrester Report

Azul Zing® Saves Enterprises $2M Over 3 Years.

Making the Technical Case for a Better JVM – Why Zing Matters.

If your Java-based operation is plagued by intermittent stalls, freezes, and jitter, and you suspect Java Garbage Collection (GC) problems, you have reached the right page. Whenever metrics matter, whether you have formal or informal service-level agreements for Java-based systems – Zing® is the ideal JVM for your operation. Zing solves GC problems and removes GC as a factor impeding your business.

Zing also makes it possible to eliminate almost all of the JVM tuning that is necessary with other JVMs. Tuning JVMs can be tough, and making sure you are making the right choices among dozens of flags and settings is problematic. There’s a lot of Java tuning lore out there – some of it valid, some years out of date, however, if you’ve gotten this far it is because you need to make things happen for your Java-based business.

There’s a better alternative: Zing.

1. With Zing, you can take GC tuning off the table. Permanently.

  • Zing doesn’t use Parallel, CMS, or G1 GC technology – the C4 collector can NEVER “stop the world”
  • No need to re-architect your application – use the working code you have
  • No need for additional third-party products or off-heap kludges – Zing just works
  • Plus, your peak latencies and outliers get lost in the normal “noise” of your operating system

stop tuning

2. With Zing, you can deliver consistently better latency.

  • Zing is ideal for customer-facing interactive workloads and Web-scale applications
  • You can deliver the same level of Java performance at the 99.99th percentile as at the 90th
  • Zing also meets the needs of trading systems, matching engines, and other peak-latency-sensitive apps
  • Zing’s ReadyNow! technology solves the Java warm-up problem. No more stalls at market open.
  • In tuned low latency applications, see peak latencies in the 1-3 millisecond range (with average latencies in the microseconds)
  • Keep your customers happy — move from multisecond (or worse) Java GC pauses to full service level compliance

Consistently low latency

3. Zing allows you to better capitalize on your existing infrastructure

  • One recent Zing customer could meet their service level requirements while supporting over 3X more users on their existing Cloud (AWS) infrastructure
  • Reduce deployment and management complexity — and lower capex and opex
  • Extend the lifespan of production Java applications
  • Take advantage of the latest server hardware with Zing’s Falcon JIT compiler
  • Find and fix production performance issues (including memory leaks) in 1/10th the time
  • Move Java apps to the Cloud easily and fully leverage the benefits of Cloud elasticity

Resource efficiency

4. Change how your Developers use Java

  • Pursue architectures and application designs not practical with existing Java platforms
    • What else could your application accomplish with 8 or 50 GB heaps? What might be possible with 250 GB heaps? Or 8 TB?
  • Keep important data on the heap without extra effort working on GC issues
  • Use standard Java functions and libraries instead of special-case ZeroGC development frameworks
  • Write Java that looks like Java, using any of the 9+ million Java developers worldwide


Get started making the technical case for Zing

Performance eBook

© Azul 2021 All rights reserved.