AMD Naples High-Performance Server Chips Pack 32 Cores, 64 Threads – Based on Zen With 8-Channel Memory, 128 PCIe Lanes, Launch in Q2 2017
AMD Naples has been previewed and will soon be heading out to high-performance servers around the globe. Based on AMD’s Zen architecture, Naples chips will feature a very dense architecture, packed with lots of compute performance.
AMD Naples High-Performance Zen Based Server SOCs Previewed – 32 Core, 64 Threaded Chips That Pack Over 19 Billion Transistors, Aimed at The Compute Hungry 1P and 2P Platforms
AMD Naples has been crafted to challenge Intel in a space they have long dominated with their Xeon lineup. This hot battle will take place in the server market with AMD Naples at one end and Intel’s Xeon at the other. AMD has several Zen Naples SKUs ranging from 16 cores and up to 32 core variants. Each of them is multi-threaded so performance is going to be top-notch on this part since Zen has a behemoth 52% IPC increase over its Opteron based predecessors. Along with performance, 14nm FinFET would also prove a great process for efficiency optimizations.
AMD Naples Features:
- A highly scalable, 32-core System on Chip (SoC) design, with support for two high-performance threads per core
- Industry-leading memory bandwidth, with 8-channels of memory per “Naples” device. In a 2-socket server, support for up to 32 DIMMS of DDR4 on 16 memory channels, delivering up to 4 terabytes of total memory capacity.
- The processor is a complete SoC with fully integrated, high-speed I/O supporting 128 lanes of PCIe 3, negating the need for a separate chipset
- A highly-optimized cache structure for high-performance, energy efficient compute
- AMD Infinity Fabric coherent interconnect for two “Naples” CPUs in a 2-socket system
- Dedicated security hardware
The Zen core comprises of a CPU complex. Each CPU complex features four cores that are connected to 8 MB of L3 cache. Zen based server chips with 32 cores can feature up to 64 MB of L3 cache which is quite disruptive. This along with 64 threads available for hyper-threaded workloads would make for a very competitive product in server platforms. A single Ryzen 8 core chip packs 4.8 Billion transistors so we can take a guess that Naples would be packing over 19 Billion transistors which is a hefty amount for a server chip. For full details on the Zen microarchitecture, you can check out a detailed article here.
AMD Naples Performance Metrics:
The new AMD server processor exceeds today’s top competitive offering on critical parameters, with 45% more cores, 60% more input / output capacity (I/O), and 122% more memory bandwidth. AMD compared performance metrics of their flagship chip with Intel’s top Xeon offering, the Xeon E5-2699 V4. The Intel chip is based on the Broadwell architecture and uses a similar 14nm process node.
Some key comparisons indicate that a dual socket platform utilizing AMD Naples chips features more cores (128 vs 88), more memory channels (16 vs 8), more memory capacity using 16 GB DRAM sticks (512 GB vs 384 GB), higher memory frequency (2400 MHz vs 1866 MHz) and higher PCIe lanes (128 vs 80).
“It is exciting to see AMD back in the server conversation with a new CPU and a sound strategy for why it is the right processor for the modern datacenter and the cloud computing era,” said Matt Eastwood, senior vice president, Enterprise Infrastructure and Datacenter, IDC. “Looking at the product details announced today, it sounds like a compelling combination that will give IT buyers a unique new option to consider when making their next upgrade.” via AMD
Upcoming Intel and AMD Server Platform Comparison:
|Intel Xeon E5-2600/4600 V5 (2S/4S)||Intel Xeon E7-8800/4800 V5 (4S/8S)||AMD Naples Platform (2P)|
|PCH||Lewisburg PCH||Lewisburg PCH||SOC|
|Socket||Socket P (LGA 3647)||Socket P (LGA 3647)||SP3 LGA socket|
|Max Core Count||Up To 26||Up To 32||Up To 32|
|Max Thread Count||Up To 52||Up To 64||Up To 64|
|Max L3 Cache||35.75 MB L3||38.5 MB L3||64 MB L3|
|DDR4 Memory Support||6-Channel DDR4||6-Channel DDR4||8-Channel DDR4|
AMD Naples Packs a Mammoth 128 PCIe Lane I/O Architecture – 128 PCIe Lanes on 2P Servers
The latest slides in the preview show that AMD will be offering an optimized GPU server platform with Naples CPU. The AMD Naples platform will feature a maximum of 128 PCI Express Gen 3.0 lanes. The huge number of lanes will allow support for a vast amount of devices. The detailed block diagram shows that a 1U rack will be able to support up to 32 NVMe devices and four discrete GPUs. The 1U rack will feature two InfiniBand EDR interconnects for data communication between storage and server systems.
The 1U rack is termed “Maximize Compute Density” so we may probably be looking at the denser Naples configuration. The one with 32 cores and 64 threads can be expected here. The 2U rack is termed “Maximize Performance/Node”. This could mean that we are looking at a different or same configuration of Naples. But, the 2U rack allows for more GPU support. The 2U rack can support up to 8 discrete graphics cards, 26 NVMe devices and a single IB EDR link. When running two chips, 64 PCIe lanes will be dedicated for communication between the two Naples chips which leaves the rest of 64 PCIe lanes for I/O use (per core).
AMD has listed down some advantages of both 1U and 2U racks. These include 4-6 directly attached discrete GPUs. RTG is actively working on delivering Vega based products around the same time as Naples so we can see some of them coupled to Naples processors. The interconnect provides P2P communication between the GPUs. There’s also no PCIe 3.0 switch required which effectively reduces cost, latency while delivering full bandwidth between GPU and CPU.
All AMD Naples chips will feature the latest Infinity Fabric coherent interconnect solution that allows for high bandwidth and high-performance server scalability. The new fabric delivers near-performance CPU scalability with various SKUs that feature different core count.
|WCCFtech||AMD Naples||AMD Summit Ridge|
|L1 Instruction Cache||32 KB x 32||32 KB x 8|
|L1 Data Cache||64 KB x 32||64 KB x 8|
|L2 Cache||512 KB x 32||512 KB x 8|
|L3 Cache||64 MB||16 MB|
|Base Clock||1.4Ghz||Up To 3.6 GHz|
|Turbo Clock||2.8Ghz||Up To 4.0 GHz|
AMD Naples Server Platform Uses 8 Channel Memory System
The Zen based Naples server platform will be featuring 8 channels for main system memory. This would allow for denser memory capacities on the racks as each channel supports 2 DDR4 DIMMs making up a total of 16 DIMMs across the 8 channels. A 2P server can house a total of 32 DDR4 DIMMs that can feature up to 4 Terabytes of memory capacity. Each chip can access a maximum of 170 GB/s of DRAM bandwidth.
Such racks will be fully optimized within several workloads for HPC environments such as molecular dynamics, rendering, graphics and data analytics, DNN and O&G.
In another slide, AMD has listed several validated platforms which they are working with OEMs to bring to the market when Naples launches.These range from 1U (2 CPU / 4 GPU), 2U (2 CPU / 8 GPU), Sled: 2P (Multi CPU / 1 or 2 GPU) and Blade: 2P (Multi CPU / 1 to 3 GPU) systems. AMD will also be offering server cards in various form factors. The add-in-cards will be available in the standard PCI Express form factor while OEM-only will include MCM and MXM modules.
A 2 socket solution should mean a total of 64 cores and 128 threads along with denser memory capacities that make Opteron sound like a kid in the park. We would also see next generation HPC server chips which combine 32 high-performance cores alongside massive Radeon Instinct cards based on Vega GPUs that will be used to crunch FP64 calculations.
AMD Naples Launches in Q2 2017 – Deep Dive On Architecture on 8th March 2017
That has been in the plans for quite some time and with Naples rolling out in Q2 2017. AMD has stated that they will be holding two presentations of Naples server chips on 8th March at 4:55 PM (PST). These will be part of the Open Compute Summit and AMD will deliver a further deep dive into the Naples Enterprise and Server platform. AMD Naples is a very powerful server and enterprise level processor platform and we can’t wait to see how it performs against Intel’s Skylake based Xeon offerings that are also coming this year and pack up to 32 cores.