Our shared goal as a society to reduce the negative environmental impact on our planet- is ENQUANTUMs inspiration and mission.
Reducing carbon footprint with green and sustainable technology, dramatically increasing the power efficiency on network data processing rather than conventional and traditional data center facilities.
Following recent global events, silicon chip manufacturing and production supply chains have been strained and adversely affected, resulting in increased lead times for silicon production.
Application-Specific-Integrated-Circuit (ASIC) is a coordinated circuit chip customized for definitive purpose, alternately intended for general use. The biggest disadvantage of ASIC is the restriction to one functional purpose without any option to reprogram the chip, this makes ASIC be listed as a high carbon footprint technology.
Field-Programmable-Gate-Array (FPGA) is the solution to achieve a longer life cycle product and dramatically decrease time to market. FPGA has an inherited feature of field programmability resulting in reduced CapEx.
A FPGA acceleration card is a system including FPGA as a processing unit, with peripheral components, memory- and various interfaces such as PCIe and Ethernet.
FPGA as a Data Center is an innovative concept ENQUANTUM pioneers to assimilate in a network entity virtualization. The advantages of data center on an FPGA are clear, along with lower carbon footprint - turning it into a Green Data Center, comes the consolidation of network infrastructure providing an absolute game changing benefit compared to universal solutions.
The upgradable long-lasting flexible platforms, ensure easy system maintenance and upgrade to any customer’s need or topological network change without purchasing additional expensive equipment to stay in line with constantly evolving and changing network standards.
Furthermore, no vendor lock using Commercial-Off-The-Shelf (COTS) product gives a convenience in operating and maintenance of a system. Along with those mentioned above, reduced operating costs is an additional advantage of our product.
The growing demand for increased bandwidth, subscriber capacity and reduced latency drives the urgency for efficient large scale dedicated data processing systems. This task is traditionally addressed by software accelerated platforms such as DPDK and VPP at the cost of server resources and floorspace.
Dedicated hardware implemented in ASIC meets those demands while retaining the disadvantageous time to market and development costs.
The ultimate solution solving this concern is FPGA technology, utilizing high-speed FPGA platforms. Performance and TTM needs are easily met maintaining the ability to upgrade without changing hardware.
ENQUANTUM has developed a data processing architecture that introduces all those innovative features, provided by Xilinx on their exceptional UltraScale+ family of FPGAs.
The operational advantage comes from a highly parallelized processing data pipeline inside the FPGA, using high performance embedded system memory that allows sub-microsecond latency for a full packet process. This technical superiority creates an ideal system for ultra-low latency, highly reliable communications channels required by IIoT, 5G networks and beyond.
Utilizing Xilinx UltraScale+ family of FPGAs, ENQUANTUM DPU manages to meet and exceed the performance metrics set for B5G. The remarkable efficiency of those platforms enables unprecedented flexibility in implementation of the final product.
Those high-performance FPGA platforms are often paired with HBM2 memory directly attached to the FPGA’s fabric. This robust memory infrastructure allows highly scalable network operations.
A major market disruptor that was introduced by Xilinx is a functionality called DFX or Dynamic Reconfiguration, a system feature that authorizes a netlist-level change in a running system. The primary goal is to provide a technology that will allow consolidation of networking infrastructure and enable smarter more efficient datacenters.
Relying on ENQUANTUMs unique parallelized data processing architecture, the DPU can operate at sub-microsecond latency and at terabit-scale while consuming substantially less energy. This groundbreakingly optimized high performance data processing architecture lays the foundation for the majority of the necessary network processing requirements.
So, what are the major KPIs for a DPU?
-
Performance: Processing latency and bandwidth capacity
-
Scalability: expanded subscriber count without purchasing additional equipment and licenses
-
TCO (Total Cost of Ownership): The purchase price of an asset, plus operating costs, over the asset’s lifespan
ENQUANTUMs DPU effortlessly meets all these KPIs at a competitive level.
The outcome is a robust DPU that enables extraordinary savings on OpEx while simultaneously addressing the growth demand retaining a competitive value to our customers.