Radeon RX 9060 XT (8GB)
AMDconsumer| Overview | |
|---|---|
| Architecture | RDNA 4 (2025) |
| Transistor count | 29.7B |
| GPU target (gfx) | gfx1200 |
| Memory | |
| VRAM | 8 GB GDDR6 |
| Memory bandwidth | 320 GB/s |
| Cache | 32 MB |
| Compute — vector | |
| FP32 | 25.6 TFLOPS |
| FP16 | 25.6 TFLOPS |
| Compute — matrix / tensor | |
| FP16 (dense / sparse) | 103 / 205 TFLOPS |
| FP8 (dense / sparse) | 205 / 410 TFLOPS |
| INT8 (dense / sparse) | 205 / 410 TOPS |
| INT4 | 410 TOPS |
| Cores & clocks | |
| Compute Units | 32 |
| Shader cores | 2,048 |
| Matrix / Tensor cores | 64 |
| Peak clock | 3,130 MHz |
| Board & system | |
| TDP | 150 W |
| PCIe | PCIe 5.0 x16 |
| Interconnect | PCIe only (no Infinity Fabric multi-GPU link) |
Compatible ROCm versions
fp16TFLOPS (103, dense) and fp8TFLOPS (205, dense) are both directly published on AMD's product page ("Peak Half Precision (FP16 Matrix) Performance: 103 TFLOPs", 205 with sparsity; "Peak 8-bit Precision (FP8 Matrix) Performance: 205 TFLOPs", 410 with sparsity) — RDNA4 is AMD's first Radeon architecture with native FP8 matrix support and 2:4 structured sparsity. int8TOPS/int8TOPSSparse mirror the FP8 figures exactly (205/410); int4TOPS listed but its structured-sparsity variant (821 TOPs) isn't captured in a separate schema field. matrixCoreCount uses the page's "AI Accelerators" count (64).
AMD also sells a 16GB variant of this card under a separate SKU/URL that returned a 404 during an earlier data pull — only the 8GB variant (confirmed real, distinct VRAM/possibly binned clocks) is represented here; treat the 16GB variant as a coverage gap, not confirmed identical to this entry. No Launch Date field is published on this page, so launchDate is omitted; releaseYear (2025) is unaffected. Source: amd.com/en/products/graphics/desktops/radeon/9000-series/amd-radeon-rx-9060xt-8gb.html ("Expand All" spec accordion), pulled 2026-09-01.