Cavium has announced that its ThunderX2, second generation of Armv8-A SoC processors are available for general release.
‘ThunderX2 combines our second generation Armv8-A custom core that delivers core and socket level performance comparable to highest end incumbent server CPUs, with best-in-class memory bandwidth, memory capacity and rich IO configurations to provide a highly differentiated offering for the server market, greatly expanding our serviceable server TAM,’ said Gopal Hegde, vice president and general manager, Data Center Processor Group, Cavium. ‘We have been working with over 60 different partners including OEMs, ODMs, and independent software and hardware vendors to enable the seamless deployment of ThunderX2 based platforms, and we are pleased to announce general availability of ThunderX2.’
The ThunderX2 family includes over 40 different SKUs for both scale up and scale out applications, ranging from top bin 32 core 2.5GHz parts to 16-core 1.6GHz parts, mapping directly across Intel's Xeon Skylake server CPUs from highest end Platinum to low end SKUs.
With list prices for volume SKUs (32 core 2.2GHz and below) ranging from $1795 to $800, the ThunderX2 family offers 2-4X better performance per dollar compared to Xeon Skylake family of processors. The ThunderX2 family is fully compliant with Armv8-A architecture specifications as well as the Arm Server Base System Architecture and Arm Server Base Boot Requirements standards.
The ThunderX2 SoC family is supported by a comprehensive software ecosystem, ranging from platform level systems management and firmware to commercial Operating Systems, Development Environments and Applications. Cavium has actively engaged in server industry standards groups such as UEFI and delivered numerous reference platforms to a broad array of community and corporate partners.
‘Since the first Mont-Blanc project, our vision has been that Arm based CPUs would be a good alternative for high end servers and HPC in particular,’ said Eric Eppe, Global Head of Solution Marketing & Portfolio – HPC & Quantum at Atos. ‘The HPC community was waiting for it, BullSequana X1000 and ThunderX2's augmented memory bandwidth are offering the best-in-class TCO for customers using memory-bound applications.’
‘We've been working on ThunderX2 since its inception, it is now ready for prime time and we are really excited by this great step towards delivering Arm-based HPC platforms to our customers,’ said Agnès Boudot – VP HPC & Quantum at Atos. ‘We have a long term strategic engagement with Arm and Cavium, which will lead the way to our Exascale Program.’
‘Sandia has been actively testing a range of key codes and applications on ThunderX2 as part of our ASC Advanced Architecture Testbed project,’ said Jim Laros, Principal Member of Technical Staff at Sandia National Laboratories. ‘Sandia's experience with ThunderX2 to date has significantly accelerated our ability to expand our compute environment to support the Armv8 based system architecture along with optimized software solutions that are key to our research and development community.’
‘We are running a set of widely used HPC science codes on prototype ThunderX2 platforms, and we are seeing impressive performance out of the box, competitive with high end, latest generation server CPUs,’ said Professor Simon McIntosh-Smith, Leader of the GW4 Isambard project and Professor of High Performance Computing at the University of Bristol. ‘We are looking forward to deploying our production scientific workloads on ThunderX2 based production platforms.’
ThunderX2 SoC Family Technical Specifications:
The ThunderX2 processor family is a single chip system on a chip (SoC). Key ThunderX2 features include:
- Single chip system on a chip (SoC) server CPU
- Core and socket level performance comparable to highest end Xeon Skylake Platinum CPUs
- Second generation of full custom Cavium Arm core
- Quad Issue, Fully Out of Order
- Full SMT support – 1, 2, 4 threads per core
- Up to 2.5 GHz in normal mode, up to 3 GHz in Turbo mode
- 3X single thread performance compared to ThunderX®
- Up to 32 cores per socket delivering > 2.5-3X socket level performance compared to ThunderX
- Cache:
- 32 KB L1 instruction and data cache, 256KB L2 per core
- 32 MB distributed L3 cache
- Advanced server class RAS features covering memory, CPU, cache, CCPI2 and PCIe interfaces
- Advanced power management
- On-chip management engine for dynamic voltage and frequency scaling across the chip
- Full Turbo mode support
- Single and dual socket configuration support using 2nd generation of Cavium Coherent Interconnect with > 2.5X coherent bandwidth compared to ThunderX
- System Memory
- 8 DDR4 memory controllers per socket
- Dual DIMM per memory controller, for a total of 16 DIMMs per socket
- Up to 4 TB of memory in dual socket configuration
- 33% higher memory bandwidth and memory capacity compared to Xeon Skylake Platinum CPUs
- Flexible IO:
- Integrated 56 lanes of PCIe Gen3 interfaces, x1, x4, x8 and x16 support, 14 integrated PCIe controllers
- Integrated SATAv3, GPIOs, USB interfaces
- 16% higher IO bandwidth compared to Xeon Skylake Platinum CPUs
