Add benchmark comparison for NeuroHTTP and NGINX

Added benchmark results comparing NeuroHTTP AI Web Server and NGINX 1.29.3, including metrics on requests/sec, latency, and data transfer.

Signed-off-by: GUIAR OQBA <117593772+okba14@users.noreply.github.com>

Author: okba14

benchmark.md

@@ -0,0 +1,70 @@
+# Benchmark: NeuroHTTP AI Web Server vs NGINX 1.29.3
+
+**Test Environment:**  
+- OS: Kali Linux  
+- CPU: 4 threads  
+- Connections: 100 concurrent  
+- Test Duration: 30 seconds per server  
+- Tool: `wrk -t4 -c100 -d30s --latency`  
+- Metrics: Requests/sec, Latency (p50, p75, p90, p99, Max), Data Transfer/sec  
+
+---
+
+## 1. Summary of Results
+
+| Server   | Requests/sec | Total Requests | Avg Latency | p50   | p75     | p90      | p99       | Max Latency | Transfer/sec |
+|----------|--------------|----------------|-------------|-------|---------|----------|-----------|-------------|--------------|
+| nginx    | 6743         | 202,701        | 80ms        | 11.79ms | 13.23ms | 218.75ms | 1.12s    | 1.33s       | 1.02MB       |
+| NeuroHTTP   | 1621         | 48,762         | 61.22ms     | 43.08ms | 100.72ms | 104.62ms | 114.27ms | 135.85ms    | 4.94MB       |
+
+---
+
+## 2. Throughput Analysis
+
+- **NGINX** achieved **6743 requests/sec**, approximately 4× higher than AIONIC.  
+- **NeuroHTTP** handled **1621 requests/sec**, with each request transferring more data (higher Transfer/sec).  
+- The lower throughput in AIONIC is expected due to internal processing (AI routing, tokenization, caching, and optimization).
+
+---
+
+## 3. Latency Analysis
+
+| Server  | Avg Latency | p50   | p75     | p90      | p99       | Max Latency |
+|---------|------------|-------|---------|----------|-----------|-------------|
+| nginx   | 80ms       | 11.79ms | 13.23ms | 218.75ms | 1.12s    | 1.33s       |
+| NeuroHTTP  | 61.22ms    | 43.08ms | 100.72ms | 104.62ms | 114.27ms | 135.85ms    |
+
+**Observations:**
+- NGINX shows very low latency for most requests (p50 = 11.79ms), but has large spikes at p90/p99 (up to 1.12s).  
+- NeuroHTTP demonstrates **stable and predictable latency**, with p99 at 114.27ms and Max latency at 135.85ms.  
+- Standard deviation confirms this: NGINX (207ms) vs AIONIC (38ms), indicating AIONIC handles requests more consistently under load.
+
+---
+
+## 4. Data Transfer
+
+- **NeuroHTTP**: 4.94MB/sec → each request delivers more content.  
+- **NGINX**: 1.02MB/sec → smaller content per request.  
+- This explains why NeuroHTTP has fewer requests/sec but higher throughput in terms of data served per second.
+
+---
+
+## 5. Strengths of Each Server
+
+| Server   | Strengths |
+|----------|-----------|
+| **NGINX** | - Extremely high throughput (6743 requests/sec)<br>- Ideal for static content (HTML/CSS/JS)<br>- Handles high concurrency efficiently<br>- Lightweight and widely adopted |
+| **NeuroHTTP** | - Stable and predictable latency (p99 = 114ms)<br>- Handles dynamic/AI-processed content<br>- Higher data transfer per request (good for large payloads)<br>- Built-in cache and optimizer reduce performance fluctuations |
+
+---
+
+## 6. Conclusion
+
+- **NGINX** excels in **raw speed** and handling a massive number of requests per second. It is perfect for static content and high-concurrency environments.  
+- **NeuroHTTP** provides **predictable latency and stability**, optimized for dynamic content and AI-based processing. Although its throughput is lower, it is better suited for applications where consistent response time and heavier payloads matter.  
+
+**Recommendation:**  
+- Use **NGINX** as a front-facing server for static content or as a reverse proxy.  
+- Use **NeuroHTTP** for dynamic AI-driven endpoints where latency stability and data-rich responses are more critical than raw request throughput.
+
+---