XCKU040

AMDKintex UltraScale

The XCKU040 is a Kintex UltraScale FPGA from AMD, built on a 20 nm process. It provides 242,400 LUT6 look-up tables (530,000 system logic cells), 484,800 flip-flops, 21.1 Mb of block RAM and 1,920 DSP slices. It offers up to 520 user I/O pins and 20 GTH/GTY transceivers at up to 16.3 Gb/s, with hard PCI Express Gen3 blocks. It is available in 4 packages, from the A784 (23 × 23 mm) to the A1156 (35 × 35 mm). Temperature grades: commercial, extended, industrial. Designs are built with AMD Vivado.

242,400LUTs (LUT6)

1,920DSP slices

21.1 MbBlock RAM

520Max. user I/O

20Transceivers (16.3 Gb/s)

1,024CoreScore (SERV cores)

Specifications

Vendor	AMD
Family	Kintex UltraScale
Series	UltraScale
Process	20 nm
Configuration memory	SRAM
Look-up tables (LUT6)	242,400
Logic block	6-input look-up table, fracturable into two LUT5 with shared inputs
LUT4 equivalent (approx.)	≈387,840
System logic cells	530,000
Flip-flops	484,800
Distributed RAM	7,050 Kb
Block RAM	21.1 Mb (600 × 36 Kb)
DSP (DSP slice)	1,920
Clock management tiles	10

Max. user I/O	520
HP I/O	416
Differential pairs	240
GTH/GTY transceivers	20 × 16.3 Gb/s
PCI Express	3× Gen3
Analog / monitor	System Monitor
CoreScore	1,024 SERV cores (hpc_ku)
Design tool	AMD Vivado

Packages

Package	Size (mm)	Pitch (mm)	User I/O	Transceivers
A784	23 × 23	0.8	468	8
A676	27 × 27	1.0	312	16
A900	31 × 31	1.0	468	16
A1156	35 × 35	1.0	520	20

Package names are shown as in the selection guide; the full ordering code adds speed grade and temperature (see the data sheet).

Speed and temperature grades

Temperature grade	Junction range	Speed grades
Commercial (C)	0 to 85 °C	-1
Extended (E)	0 to 100 °C	-2, -3
Industrial (I)	−40 to 100 °C	-1, -1L, -2

Alternatives to the XCKU040

Closest matches in other families and vendors by logic, DSP, memory, I/O and transceivers. Check pin compatibility and tools before migrating.

Device	LUTs	Vendor logic	DSP	Block RAM	Max I/O	Transceivers	PCIe	CoreScore
XC7K410TAMD Kintex-7	254,200 LUT6	407K LC	1,540	27.9 Mb	500	16 GTX 12.5 Gb/s	Gen2	—
XC7K420TAMD Kintex-7	260,600 LUT6	417K LC	1,680	29.4 Mb	400	32 GTX 12.5 Gb/s	Gen2	1,276 hpc_k7
XC7K355TAMD Kintex-7	222,600 LUT6	356K LC	1,440	25.1 Mb	300	24 GTX 12.5 Gb/s	Gen2	—
MPF500TMicrochip PolarFire	481,000 LUT4	481K LE	1,480	29.7 Mb	584	24 SERDES 12.7 Gb/s	Gen2	—
LAV-AT-G70Lattice Avant-G	397,000 LUT4	637K LC	1,800	35.6 Mb	554	28 SERDES 12.5 Gb/s	—	—
LAV-AT-X70Lattice Avant-X	397,000 LUT4	637K LC	1,800	35.6 Mb	554	28 SERDES 25 Gb/s	—	—

Compare side by side: XCKU040 vs XC7K410T, XCKU040 vs XC7K420T, XCKU040 vs XC7K355T.

Compare the XCKU040 with up to three other FPGAs.

XCKU040 quick facts

How many LUTs does the XCKU040 have?

The XCKU040 has 242,400 LUT6 look-up tables (530,000 system logic cells as published by AMD). Logic block: 6-input look-up table, fracturable into two LUT5 with shared inputs.

How many DSP blocks does the XCKU040 have?

The XCKU040 has 1,920 DSP slices and 21.1 Mb of block RAM.

Does the XCKU040 have transceivers?

Yes, 20 GTH/GTY transceivers at up to 16.3 Gb/s.

What is the CoreScore of the XCKU040?

The XCKU040 fits 1,024 SERV RISC-V cores in the CoreScore benchmark (board: hpc_ku). CoreScore counts how many of these bit-serial cores the device and its tools can place and route, a practical measure of usable logic.

What are the alternatives to the XCKU040?

Devices of similar size in other families include XC7K410T (AMD Kintex-7), XC7K420T (AMD Kintex-7), XC7K355T (AMD Kintex-7).

Which software is used to design with the XCKU040?

Designs for the XCKU040 are built with AMD Vivado.

Other Kintex UltraScale devices

Device	LUTs	Vendor logic	DSP	Block RAM	Max I/O	Transceivers	PCIe	CoreScore
XCKU025	145,440 LUT6	318K SLC	1,152	12.7 Mb	312	12 GTH/GTY 16.3 Gb/s	Gen3	—
XCKU035	203,128 LUT6	444K SLC	1,700	19 Mb	520	16 GTH/GTY 16.3 Gb/s	Gen3	—
XCKU040	242,400 LUT6	530K SLC	1,920	21.1 Mb	520	20 GTH/GTY 16.3 Gb/s	Gen3	1,024 hpc_ku
XCKU060	331,680 LUT6	726K SLC	2,760	38 Mb	624	32 GTH/GTY 16.3 Gb/s	Gen3	—
XCKU085	497,520 LUT6	1088K SLC	4,100	57 Mb	676	56 GTH/GTY 16.3 Gb/s	Gen3	—
XCKU095	537,600 LUT6	1176K SLC	768	59.1 Mb	702	64 GTH/GTY 16.3 Gb/s	Gen3	—
XCKU115	663,360 LUT6	1451K SLC	5,520	75.9 Mb	832	64 GTH/GTY 16.3 Gb/s	Gen3	—

Source: AMD UltraScale FPGAs Product Selection Guide, XMP102 (v1.7). Data is extracted automatically from the vendor's selection guide; verify critical values in the device data sheet. CoreScore result from the CoreScore world ranking (olofk/corescore).