Technology and architecture of HP ProLiant AMD-based 100 - series G6 (Generation 6) servers

Abstract
This technology brief describes the key technologies implemented in HP ProLiant 100-series G6
servers based on AMD™ processors. As of this writing, the AMD-based 100-series G6 server
platforms are limited to the ProLiant DL165. A link to the QuickSpecs for this server is listed in the “For
more information” section at the end of this technology brief.
Introduction
HP ProLiant AMD-based 100-series G6 server platforms are high performance, low cost, ultra-dense
rack server nodes for compute and I/O intensive environments. ProLiant AMD 100-series G6 servers
include these new technologies:
AMD Opteron™ six-core 2400 Series processors, including high efficiency (HE) processor options
The ProLiant Onboard Administrator Powered by Lights-Out 100i (LO 100i)
Input/output (I/O) technologies such as PCI Express (PCIe) and faster Smart Array controllers that
incorporate common form factor components
High efficiency power supplies to provide the required amount of power and improve power
efficiency
The technologies discussed in this paper are implemented in the ProLiant DL165 G6 server.
Processor technology
The HP G6 AMD-based 100-series servers use the AMD Opteron 2400 series Six-Core processor.
These processors are based on AMD's 45 nanometer (nm) quad-core architecture. In addition, these
processors use Direct Connect™ architecture 2.0, HyperTransport™ 3.0, HT Assist™, and 8x ECC
error correction. The Six-Core processors fit into the same socket 1207 architecture as previous four-
core AMD Opteron processors. Each processor operates at speeds of up to 2.9 GHz, has 512 KB of
L2 cache memory, and shares a total of 6 MB of L3 cache.
Direct Connect architecture 2.0
Instead of front-side bus architecture, Direct Connect integrates the memory controller into the
processor and directly connects CPUs to the I/O subsystem and other processors (Figure 1). Direct
Connect architecture uses direct communication links (HyperTransport links) between each CPU,
between CPU and I/O, and between CPU and memory. Direct Connect architecture currently scales
up to 12 cores which provide superior memory and I/O capability, near native virtualization
performance, and a range of power bands
1
that place a priority on low power consumption.
1
Power bands refer to a new metric developed by AMD to reflect power consumed by the processor and its integrated
memory controller during peak workloads. This metric is based on AMD’s measurement of Average CPU Power (ACP). For
more information on ACP, see the whitepaper at www.amd.com/us-
en/assets/content_type/white_papers_and_tech_docs/43761C_ACP_WP.pdf
2