LFE5UM-25 vs XC7A12T
A comparison of the Lattice ECP5 LFE5UM-25 and the AMD Artix-7 XC7A12T using the data from both vendors' selection guides.
Side-by-side specifications
| LFE5UM-25 | XC7A12T | |
|---|---|---|
| Vendor / family | Lattice ECP5 | AMD Artix-7 |
| Logic (LUT4-equivalent) | 24,000 | 12,800 |
| LUTs as published | 24,000 LUT4 | 8,000 LUT6 |
| Vendor logic figure | — LC | 12,800 LC |
| DSP | 28 18×18 multipliers | 40 DSP slices |
| Block RAM | 1 Mb | 0.7 Mb |
| Large RAM | — | — |
| Max. user I/O | 197 | 150 |
| Transceivers | 2 × SERDES 3.2 Gb/s | 2 × GTP 6.6 Gb/s |
| Hard PCI Express | None | Gen2 |
| Hard processor | None | None |
| Process | 40 nm | 28 nm |
| Configuration memory | SRAM | SRAM |
| Packages | 2 | 2 |
| Design tool | Lattice Diamond | AMD Vivado |
Highlighted values are higher. Logic is compared in LUT4 equivalents; see how FPGA vendors count logic.
Key differences
The LFE5UM-25 has about 88% more logic capacity (24K vs 12.8K LUT4 equivalents).
They use different logic architectures: the LFE5UM-25 is built from LUT4 and the XC7A12T from LUT6, so the raw LUT counts (24K vs 8K) are not directly comparable.
For arithmetic, the XC7A12T has 40 DSP slices against 28 18×18 multipliers on the LFE5UM-25.
The LFE5UM-25 has more on-chip memory (1 Mb vs 0.7 Mb).
Both have transceivers: 2 at up to 3.2 Gb/s on the LFE5UM-25 and 2 at up to 6.6 Gb/s on the XC7A12T.
The LFE5UM-25 is designed with Lattice Diamond and the XC7A12T with AMD Vivado.
Full specifications
See the complete spec sheets with packages and grades: LFE5UM-25 (ECP5) and XC7A12T (Artix-7).