The XCKU035-1FBVA900I is a high-performance Field Programmable Gate Array (FPGA) from AMD Xilinx, belonging to the Kintex® UltraScale™ family. Built on cutting-edge 20nm process technology, this device delivers an exceptional balance of signal processing power, DSP bandwidth, and power efficiency — making it a premier choice for 100G networking, medical imaging, video processing, and wireless infrastructure applications.
Whether you are designing for industrial environments or high-throughput data centers, the XCKU035-1FBVA900I provides the flexibility, logic density, and I/O capability your design demands. For a broader selection of configurable logic solutions, explore our full range of Xilinx FPGA products.
What Is the XCKU035-1FBVA900I?
The XCKU035-1FBVA900I is a member of AMD’s Kintex UltraScale FPGA family — a mid-range device line engineered to maximize price-per-performance-per-watt at 20nm. The part number breaks down as follows:
| Part Number Segment |
Meaning |
| XC |
Xilinx Commercial/Industrial FPGA |
| KU035 |
Kintex UltraScale, device size 035 |
| -1 |
Speed grade -1 (standard performance) |
| FBVA |
Flip-Chip Ball Grid Array package variant |
| 900 |
900-pin ball count |
| I |
Industrial temperature range (–40°C to +100°C) |
Key Specifications at a Glance
| Parameter |
Value |
| Manufacturer |
AMD (Xilinx) |
| Part Number |
XCKU035-1FBVA900I |
| Family |
Kintex® UltraScale™ |
| Technology Node |
20nm |
| Logic Cells |
355,474 |
| Macrocells / CLB LUTs |
444,343 |
| User I/Os |
468 |
| Block RAM |
19,456 Kbit (19 Mb) |
| Package |
900-Pin FCBGA (FC-BGA) |
| Package Dimensions |
31mm × 31mm |
| Core Voltage (VCCINT) |
0.95V |
| Speed Grade |
-1 |
| Operating Temperature |
–40°C to +100°C (Industrial) |
| Process Technology |
TSMC 20nm HPL |
| Mounting Type |
Surface Mount |
| RoHS Compliant |
Yes |
XCKU035-1FBVA900I Architecture Overview
UltraScale Architecture
The XCKU035-1FBVA900I is built on Xilinx’s UltraScale architecture, which delivers ASIC-like performance and power management capabilities. This architecture uses a next-generation interconnect fabric that eliminates traditional routing bottlenecks found in earlier FPGA generations, enabling predictable timing closure and higher operating frequencies.
The UltraScale architecture also introduces ASIC-like clocking with fine-granular clock gating, which directly contributes to the device’s class-leading power reduction — up to 40% lower power versus previous-generation Kintex FPGAs.
Logic Resources
| Resource |
XCKU035 |
| CLB LUTs |
203,128 |
| CLB Flip-Flops |
406,256 |
| Distributed RAM (Kb) |
2,888 |
| Block RAM Tiles |
432 |
| Block RAM (Mb) |
19.1 |
| UltraRAM (Mb) |
0 |
| DSP Slices |
1,440 |
DSP and Signal Processing
With 1,440 DSP48E2 slices, the XCKU035-1FBVA900I delivers substantial signal processing throughput suited for:
- FIR Filtering in communications and radar
- FFT / IFFT operations for spectrum analysis
- Matrix multiplication in AI inference pipelines
- Wireless baseband processing in LTE/5G systems
- Medical imaging reconstruction and real-time ultrasound
Transceiver Capabilities
| Transceiver Type |
Count |
Max Line Rate |
| GTH Transceivers |
20 |
16.3 Gb/s |
| PCIE Gen3 x8 |
1 |
Integrated |
These next-generation GTH transceivers support multi-protocol operation across PCIe Gen3, CPRI, JESD204B, SRIO, and 10G/40G/100G Ethernet, making the XCKU035-1FBVA900I ideal for high-bandwidth line-card and backplane designs.
XCKU035-1FBVA900I Pin and Package Details
Package: 900-Pin FCBGA
The Flip-Chip Ball Grid Array (FCBGA) package used by the XCKU035-1FBVA900I provides a compact 31mm × 31mm footprint with 0.8mm ball pitch. This package is footprint-compatible with other Kintex UltraScale devices, allowing for straightforward design migration.
| Package Detail |
Value |
| Package Type |
FCBGA (FC-BGA) |
| Total Ball Count |
900 |
| Ball Pitch |
0.8mm |
| Package Size |
31mm × 31mm |
| User I/O Pins |
468 |
| High-Performance (HP) Banks |
4 |
| High-Range (HR) Banks |
6 |
| Configuration I/O |
Included |
I/O Standards Supported
The XCKU035-1FBVA900I supports a wide range of single-ended and differential I/O standards:
| I/O Standard Category |
Examples |
| Single-Ended |
LVCMOS 1.2V / 1.5V / 1.8V / 2.5V / 3.3V |
| Differential |
LVDS, LVDS 25, RSDS, MINI-LVDS, PPDS |
| Memory Interfaces |
DDR4, DDR3L, LPDDR4 |
| High-Speed Serial |
GTH @ up to 16.3 Gb/s |
Power Specifications
Supply Voltages
| Supply Rail |
Nominal Voltage |
Function |
| VCCINT |
0.95V |
Core logic voltage |
| VCCAUX |
1.8V |
Auxiliary circuits |
| VCCO |
1.2V – 3.3V |
I/O banks output voltage |
| VCCBRAM |
0.95V |
Block RAM supply |
| MGTAVCC |
1.0V |
Transceiver analog core |
| MGTAVTT |
1.2V |
Transceiver termination |
Power Efficiency Highlights
The -1 speed grade of the XCKU035 operates at a VCCINT of 0.95V, delivering a strong balance between performance and power consumption. The UltraScale architecture features:
- Fine-granular clock gating reducing dynamic power
- Low-power standby modes
- Up to 40% total power reduction vs. Kintex-7 FPGAs
- Optimized SSI (Stacked Silicon Interconnect) power delivery
Target Applications for XCKU035-1FBVA900I
The XCKU035-1FBVA900I is purpose-built for performance-critical, mid-range FPGA applications. Its combination of DSP density, transceiver bandwidth, and industrial temperature rating makes it a preferred solution in:
Networking and Data Center
- 100G Ethernet packet processing and line-rate forwarding
- Network Function Virtualization (NFV) offload engines
- OTN (Optical Transport Network) framing and switching
- Data center interconnect and switching fabrics
Wireless Infrastructure
- 4G LTE / 5G NR Remote Radio Head (RRH) digital front-end
- Massive MIMO beamforming and signal aggregation
- Heterogeneous network baseband processing
- CPRI / eCPRI fronthaul interfaces
Defense and Aerospace
- Radar signal processing and target tracking
- Electronic warfare and signals intelligence
- Software-Defined Radio (SDR) platforms
- Secure communications and encryption acceleration
Medical Imaging
- High-resolution ultrasound signal chain processing
- CT / MRI image reconstruction pipelines
- 8K×4K video acquisition and real-time processing
- DICOM image streaming and transfer
Industrial and Test Equipment
- High-speed data acquisition systems
- Machine vision processing pipelines
- Industrial automation real-time control
- Protocol analysis and test instrumentation
Software Development: Vivado Design Suite
The XCKU035-1FBVA900I is fully supported by AMD’s Vivado™ Design Suite, which provides:
| Tool |
Function |
| Vivado Synthesis |
RTL-to-gate-level synthesis |
| Vivado Implementation |
Placement and routing |
| Vivado Simulator |
Functional and timing simulation |
| Vivado IP Integrator |
Block design and IP integration |
| Xilinx Power Estimator (XPE) |
Pre- and post-implementation power analysis |
Vivado is available in free WebPACK and full-featured paid editions. The XCKU035 is supported from Vivado 2014.1 onwards.
Additionally, AMD’s Vitis™ Unified Software Platform enables high-level synthesis (HLS) for accelerating algorithm development targeting the XCKU035 device fabric.
Part Number Decoder: XCKU035 Speed Grades and Variants
The XCKU035 device is available in multiple speed grades and package options. The table below shows common variants:
| Part Number |
Speed Grade |
Package |
Temp Range |
I/O Count |
| XCKU035-1FBVA900I |
-1 |
900 FCBGA |
Industrial |
468 |
| XCKU035-1FBVA900C |
-1 |
900 FCBGA |
Commercial |
468 |
| XCKU035-2FBVA900I |
-2 |
900 FCBGA |
Industrial |
468 |
| XCKU035-2FBVA900E |
-2 |
900 FCBGA |
Extended |
468 |
| XCKU035-3FBVA900E |
-3 |
900 FCBGA |
Extended |
468 |
| XCKU035-1FFVA1156I |
-1 |
1156 FCBGA |
Industrial |
520 |
| XCKU035-1FFVA1156C |
-1 |
1156 FCBGA |
Commercial |
520 |
Temperature Range Key:
| Suffix |
Range |
Use Case |
| C (Commercial) |
0°C to +85°C |
Consumer / Commercial |
| I (Industrial) |
–40°C to +100°C |
Industrial / Harsh environments |
| E (Extended) |
–40°C to +100°C |
Extended screening |
The XCKU035-1FBVA900I specifically targets industrial-grade applications where reliable operation across wide temperature excursions is mandatory.
Compliance and Certifications
| Standard |
Status |
| RoHS (Restriction of Hazardous Substances) |
Compliant |
| ECCN (Export Control Classification Number) |
3A991.d |
| USHTS Code |
8542390001 |
| TARIC Code |
8542399000 |
| MSL (Moisture Sensitivity Level) |
MSL 3 |
Comparison: XCKU035 vs. Adjacent Kintex UltraScale Devices
| Feature |
XCKU025 |
XCKU035 |
XCKU040 |
XCKU060 |
| Logic Cells |
247,680 |
355,474 |
488,960 |
722,880 |
| CLB LUTs |
141,120 |
203,128 |
242,400 |
331,680 |
| DSP Slices |
840 |
1,440 |
1,920 |
2,760 |
| Block RAM (Mb) |
13.1 |
19.1 |
21.1 |
32.1 |
| GTH Transceivers |
16 |
20 |
24 |
32 |
| Max User I/O |
396 |
468 |
520 |
520 |
The XCKU035 occupies the sweet spot in the Kintex UltraScale lineup, offering significantly more DSP capacity and logic than the XCKU025 while maintaining a cost advantage over the larger XCKU040 and XCKU060.
Frequently Asked Questions (FAQ)
What is the XCKU035-1FBVA900I used for?
The XCKU035-1FBVA900I is used in high-performance applications requiring DSP-intensive processing, including 100G networking, medical imaging, wireless infrastructure (4G/5G), and defense signal processing systems.
What is the difference between -1, -2, and -3 speed grades?
The speed grade indicates the maximum switching performance of the device. The -3 speed grade is the fastest, while -1 is the slowest (but lower power) within the Kintex UltraScale family. All speed grades of the XCKU035 operate at VCCINT = 0.95V in standard operation.
Is the XCKU035-1FBVA900I RoHS compliant?
Yes, the XCKU035-1FBVA900I is fully RoHS compliant and lead-free.
What development tools are supported?
The XCKU035-1FBVA900I is supported by AMD Vivado Design Suite (v2014.1 and later), Vitis HLS for C/C++ design entry, and the Xilinx Power Estimator (XPE) for power analysis.
What temperature range does the XCKU035-1FBVA900I support?
The “I” suffix designates an Industrial temperature range, meaning the device is specified for operation from –40°C to +100°C junction temperature.
Can this device be used for PCIe applications?
Yes. The XCKU035-1FBVA900I integrates a PCIe Gen3 x8 hard block, enabling high-bandwidth PCI Express connectivity without consuming additional logic resources.
What package is used and what is the pin pitch?
The XCKU035-1FBVA900I uses a 900-ball Flip-Chip BGA (FCBGA) package with 0.8mm ball pitch and a 31mm × 31mm body size.
Summary
The XCKU035-1FBVA900I from AMD Xilinx is a robust, industrial-grade FPGA that delivers:
- 355,474 logic cells for complex design implementation
- 1,440 DSP slices for high-throughput signal processing
- 20 GTH transceivers at up to 16.3 Gb/s for multi-protocol connectivity
- 468 user I/Os in a compact 900-pin FCBGA package
- Industrial temperature range (–40°C to +100°C) for demanding environments
- 20nm low-power process for optimal performance-per-watt
With its industry-leading price-performance-per-watt ratio and broad ecosystem support through Vivado Design Suite, the XCKU035-1FBVA900I remains one of the most versatile mid-range FPGAs available for design engineers today.
