NOL8 clears independent test with 118x data-per-dollar advantage

• 4.1x to 118x more data per dollar than Google RE2 across six tests
• NOL8 costs held near US$2 per GB/s per hour as complexity increased
• Latency stayed at 3-5 milliseconds for up to 98% of the workload
• General availability targeted for late October 2026
• 52,516,667 Class C Performance Rights vesting condition satisfied

NOL8’s cost advantage widens with workload complexity

FortifAI Limited (ASX:FTI) has put a striking number behind its NOL8 AI Data Plane pitch: the platform delivered between 4.1 and 118 times more data per infrastructure dollar than Google RE2 in six tests, with the largest advantage appearing at 5,000 rules. NOL8’s cost stayed at approximately US$2 per gigabyte per second of throughput per hour across the test set, while RE2’s rose from US$8 to US$240.

Independent review tested FPGA and CPU configurations

The verification was conducted by Professor Amit Klein, head of the Federmann Cyber Security Center at The Hebrew University of Jerusalem. His assessment involved an independent review of NOL8’s testing methodology and results rather than a disclosed replication of every test. NOL8 Technology V1.0, using FPGA infrastructure on Amazon Web Services, was compared with Google RE2 running on CPUs as the number of rules increased from 128 to 5,000 under chaotic load intervals designed to reflect peaks and idle periods.

Latency and throughput remain central to the enterprise case

NOL8 held latency between 3 and 5 milliseconds for up to 98% of the workload while producing more than 950 megabytes per second of throughput. RE2’s latency rose to 10 milliseconds, while its throughput fell from about 100 megabytes per second to 2 megabytes per second across the tests. The figures are meaningful to FortifAI’s governance proposition, but they remain dependent on the hardware configurations, workload design, rule sets and cost assumptions used in the comparison.

Megaport and Equinix channels move towards commercial testing

The results add a more demanding benchmark to NOL8’s distribution strategy through Megaport and Equinix. FortifAI previously reported across Virginia, Tokyo and Sydney, where it said the platform delivered approximately 3 milliseconds of in-region latency and 10 times more governed data per dollar than legacy CPU-based software. NOL8 joined the Megaport AI Exchange in June and became available through Equinix Fabric across more than 60 locations in September. FortifAI says proof-of-concept engagements are under way ahead of general availability targeted for late October, but the announcement discloses no revenue, signed contract value or customer conversions.

Performance rights vest as commercial milestone approaches

The verification also satisfies the vesting condition for 52,516,667 Class C Performance Rights issued in connection with FortifAI’s acquisition of FastAI and shareholder-approved CEO incentives. That creates a clear equity consequence alongside the technology milestone. The next test for the NOL8 story is less about another headline benchmark than whether the platform can turn its reported economics and governance claims into paying deployments through its channel partners.

Bottom Line?

The benchmark strengthens NOL8’s commercial case, but late-October availability and proof-of-concept conversion will matter more than the headline multiple.

Questions in the middle?

• Will Megaport and Equinix proof-of-concept engagements convert into disclosed customers or revenue?
• How would NOL8’s cost advantage perform under independent testing with different hardware, workloads and rule sets?
• What will be the effect of vesting 52,516,667 Class C Performance Rights on FortifAI’s capital structure?