← Return to FleetShop This Rig ↗
Neon Future
Status: onlineA prototype build simulating the next generation of consumer AI hardware. Focused on high-bandwidth memory and neural-specific instruction sets. **Update (02/14/26):** Upgraded from RTX 5060 (8GB) to RTX 5070 (12GB) to test higher VRAM capacity models.
GPUNVIDIA RTX 5070 (12GB)
VRAM12GB GDDR7
CPUIntel Core Ultra 5
RAM96GB DDR5
Storage2TB NVMe Gen5
OSUbuntu
Performance Log
| Date | Model | Hardware | Params | Context | Prompt Eval | Token Gen |
|---|---|---|---|---|---|---|
| 2026-08-28 | Llama-3.1-8B(CPU Only (Ultra 5)) | CPU Only (Ultra 5) | Default (Q4) | 4096 tk | 45.2 t/s | 6.9 t/s |
| 2026-02-15 | Gemma-3-4.3B(RTX 5070 (12GB)) | RTX 5070 (12GB) | Q4_K_M | 4096 tk | 8944.5 t/s | 141.8 t/s |
| 2026-02-15 | Gemma-3-4.3B (32k Context)(RTX 5070 (12GB)) | RTX 5070 (12GB) | Q4_K_M | 32768 tk | 7953.6 t/s | 122.7 t/s |
| 2026-02-15 | Gemma-3-4.3B (64k Context)(RTX 5070 (12GB)) | RTX 5070 (12GB) | Q4_K_M | 65536 tk | 6968.6 t/s | 109.5 t/s |
| 2026-02-15 | Llama-3.1-8B(RTX 5070 (12GB)) | RTX 5070 (12GB) | Default (Q4) | 4096 tk | 4992.8 t/s | 99.9 t/s |
| 2026-02-15 | Llama-3.1-8B (32k Context)(RTX 5070 (12GB)) | RTX 5070 (12GB) | Default (Q4) | 32768 tk | 2550.2 t/s | 14.5 t/s |
| 2026-02-15 | Llama-3.1-8B (64k Context)(RTX 5070 (12GB)) | RTX 5070 (12GB) | Default (Q4) | 65536 tk | 991.7 t/s | 2.0 t/s |
| 2026-01-29 | Gemma-3-4.3B(RTX 5060 (8GB)) | RTX 5060 (8GB) | Q4_K_M | 4096 tk | 5545.6 t/s | 115.4 t/s |
| 2026-01-29 | Gemma-3-4.3B (32k Context)(RTX 5060 (8GB)) | RTX 5060 (8GB) | Q4_K_M | 32768 tk | 4809.1 t/s | 95.6 t/s |
SPONSORED TESTS// AD_SLOT: 1234567890 // FORMAT: AUTO
Performance Analysis
Trend Visualization
Historical Archive
Filter:
Showing 9 of 9 runs| Date | Model | Quant | Context | VRAM | Prompt (t/s) | Gen (t/s) |
|---|---|---|---|---|---|---|
| 2026-08-28 | Llama-3.1-8B | Default (Q4) | 4,096 | - | 45.2 | 6.9 |
| 2026-02-15 | Gemma-3-4.3B | Q4_K_M | 4,096 | 3.7GB | 8944.5 | 141.8 |
| 2026-02-15 | Gemma-3-4.3B (32k Context) | Q4_K_M | 32,768 | 4.3GB | 7953.6 | 122.7 |
| 2026-02-15 | Gemma-3-4.3B (64k Context) | Q4_K_M | 65,536 | 5.1GB | 6968.6 | 109.5 |
| 2026-02-15 | Llama-3.1-8B | Default (Q4) | 4,096 | 5.3GB | 4992.8 | 99.9 |
| 2026-02-15 | Llama-3.1-8B (32k Context) | Default (Q4) | 32,768 | 12GB (Spillover) | 2550.2 | 14.5 |
| 2026-02-15 | Llama-3.1-8B (64k Context) | Default (Q4) | 65,536 | 12GB (Major Spillover) | 991.7 | 2.0 |
| 2026-01-29 | Gemma-3-4.3B | Q4_K_M | 4,096 | - | 5545.6 | 115.4 |
| 2026-01-29 | Gemma-3-4.3B (32k Context) | Q4_K_M | 32,768 | - | 4809.1 | 95.6 |