Product Advantages

Virtex® UltraScale+™ HBM FPGAs provide the highest on-chip memory density with up to 500Mb of total on-chip integrated memory, plus up to 16GB of high-bandwidth memory (HBM) Gen2 integrated in-package for 460GB/s of memory bandwidth.

Innovative embedded HBM controller and breakthrough integration enables maximum bandwidth, efficient routing complexity, and optimized power efficiency for workloads that process large datasets from AI inference, video transcoding, next-generation firewalls, to data warehouse.

Key Features and Benefits


Enormous Memory Bandwidth

460GB/s of memory bandwidth delivers 20X more bandwidth than a DDR4 DIMM

Maximized HBM Bandwidth

Extended AXI interfaces to enable 3.7Tb/s operation and flexible addressing over pseudo-channel boundaries help users get efficient routing and the most usable bandwidth in and out of the HBM stacks

PCIe® Gen4 x8 with CCIX

Cache coherent compute using CCIX ports are suitable for complete end-to-end solution for multi-100G ports

32.75Gb/s Transceivers

Up to 96 transceivers on a device provides 3.1Tb/s SerDes bandwidth

Footprint Reduction

By eliminating I/O to external packaged memories and providing massive on-chip memory, these FPGAs enable small footprint system design without comprising performance

Find Out More

View product specifications or talk to someone now.


Explore New Possibilities with Virtex UltraScale+ HBM FPGAs


Compute Acceleration

Compute Acceleration

Up to 28 TOPs (INT8) and 460GB/s of memory bandwidth provide performance and capacity for a large number of database and data analytics workloads.

Network Acceleration

Virtex UltraScale+ HBM FPGAs store all flow tables and allow concurrent access for Network Functions Virtualization (NFV) and Open vSwitch (OVS) deployment in Software-Defined Networking (SDN). Doubled memory density offers scalability and bigger query size to accelerate random rule databases in the firewalls and routers.

Network Acceleration

Storage Acceleration

Storage Acceleration

Enterprise-class data center IPs, high data processing capability, and massive memory density accelerate sophisticated storage algorithms for new storage architecture.

Artificial Intelligence

The interconnect between FPGA and HBM shortens the data movement. High memory bandwidth with low latency unlocks compute capacities for AI inference applications such as autonomous driving, neural networks, and multi-layer perceptron (MLP).

Artificial Intelligence
Product Table
System Logic Cells (K) 962 962 1,907 2,852 1,907 2,852
DSP Slices 2,880 2,880 5,952 9,024 5,952 9,024
HBM DRAM (GB) 4 8 8 8 16 16
HBM AXI Interfaces 32 32 32 32 32 32
PCIe Gen3 x16/ Gen4x8/CCIX 4 4 4 4 4 4
GTY/GTM Transceivers (32.75/58Gb/s) 32/0 32/0 64/0 96/0 64/0 96/0
I/O 208 208 416 624 416 624
Download Product Selection Guide


Filter Results
Default Default Title Document Type Date
Get Started

Training Courses

Xilinx hands-on FPGA and Embedded SoC design training provides you the knowledge to begin designing right away

Core Technologies

Xilinx delivers the most dynamic processing technology in the industry, enabling rapid innovation with its adaptable, intelligent computing

Tools & Utilities

Xilinx provides comprehensive tools and utilities to maximize productivity of your design

Development Boards & Kits

Jump start your development with Xilinx evaluation boards & kits

Popular Resources

Learn from design advisories, get answers for the most common issues. We’re here to help you


Featured Videos

All Videos

Adaptable Advantage Blog