Home
Xeon E5

Intel Introduces Xeon Phi: Larrabee Unleashed

Subject: Processors | June 19, 2012 - 11:46 AM |
Tagged: Xeon Phi, xeon e5, nvidia, larrabee, knights corner, Intel, HPC, gpgpu, amd

Intel does not respond well when asked about Larabee.  Though Intel has received a lot of bad press from the gaming community about what they were trying to do, that does not necessarily mean that Intel was wrong about how they set up the architecture.  The problem with Larrabee was that it was being considered as a consumer level product with an eye for breaking into the HPC/GPGPU market.  For the consumer level, Larrabee would have been a disaster.  Intel simply would not have been able to compete with AMD and NVIDIA for gamers’ hearts.
 
The problem with Larrabee and the consumer space was a matter of focus, process decisions, and die size.  Larrabee is unique in that it is almost fully programmable and features really only one fixed function unit.  In this case, that fixed function unit was all about texturing.    Everything else relied upon the large array of x86 processors and their attached vector units.  This turns out to be very inefficient when it comes to rendering games, which is the majority of work for the consumer market in graphics cards.  While no outlet was able to get a hold of a Larrabee sample and run benchmarks on it, the general feeling was that Intel would easily be a generation behind in performance.  When considering how large the die size would have to be to even get to that point, it was simply not economical for Intel to produce these cards.
 
phi_01.jpg
 
Xeon Phi is essentially an advanced part based on the original Larrabee architecture.
 
This is not to say that Larrabee does not have a place in the industry.  The actual design lends itself very nicely towards HPC applications.  With each chip hosting many x86 processors with powerful vector units attached, these products can provide tremendous performance in HPC applications which can leverage these particular units.  Because Intel utilized x86 processors instead of the more homogenous designs that AMD and NVIDIA use (lots of stream units doing vector and scalar, but no x86 units or a more traditional networking fabric to connect them).  This does give Intel a leg up on the competition when it comes to programming.  While GPGPU applications are working with products like OpenCL, C++ AMP, and NVIDIA’s CUDA, Intel is able to rely on many current programming languages which can utilize x86.  With the addition of wide vector units on each x86 core, it is relatively simple to make adjustments to utilize these new features as compared to porting something over to OpenCL.
 
So this leads us to the Intel Xeon Phi.  This is the first commercially available product based on an updated version of the Larrabee technology.  The exact code name is Knights Corner.  This is a new MIC (many integrated cores) product based on Intel’s latest 22 nm Tri-Gate process technology.  The details are scarce on how many cores this product actually contains, but it looks to be more than 50 of a very basic “Pentium” style core;  essentially low die space, in-order, and all connected by a robust networking fabric that allows fast data transfer between the memory interface, PCI-E interface, and the cores.
 
intelphi.jpg
 
Each Xeon Phi promises more than 1 TFLOP of performance (as measured by Linpack).  When combined with the new Xeon E5 series of processors, these products can provide a huge amount of computing power.  Furthermore, with the addition of the Cray interconnect technology that Intel acquired this year, clusters of these systems could provide for some of the fastest supercomputers on the market.  While it will take until the end of this year at least to integrate these products into a massive cluster, it will happen and Intel expects these products to be at the forefront of driving performance from the Petascale to the Exascale.
 
phi_02.jpg
 
These are the building blocks that Intel hopes to utilize to corner the HPC market.  Providing powerful CPUs and dozens if not hundreds of MIC units per cluster, the potential computer power should bring us to the Exascale that much sooner.
 
Time will of course tell if Intel will be successful with Xeon Phi and Knights Corner.  The idea behind this product seems sound, and the addition of powerful vector units being attached to simple x86 cores should make the software migration to massively parallel computing just a wee bit easier than what we are seeing now with GPU based products from AMD and NVIDIA.  The areas that those other manufacturers have advantages over Intel are that of many years of work with educational institutions (research), software developers (gaming, GPGPU, and HPC), and industry standards groups (Khronos).  Xeon Phi has a ways to go before being fully embraced by these other organizations, and its future is certainly not set in stone.  We have yet to see 3rd party groups get a hold of these products and put them to the test.  While Intel CPUs are certainly class leading, we still do not know of the full potential of these MIC products as compared to what is currently available in the market.
 

The one positive thing for Intel’s competitors is that it seems their enthusiasm for massively parallel computing is justified.  Intel just entered that ring with a unique architecture that will certainly help push high performance computing more towards true heterogeneous computing. 

Source: Intel

Sandy Bridge Xeons galore! Seventeen new E5s arrive

Subject: General Tech | May 15, 2012 - 02:17 PM |
Tagged: xeon, xeon e5, Sandy Bridge E, Sandy Bridge EN, Sandy Bridge EP, lga 1356, lga 2011

Today marks the arrival of the Xeon E3-1200 single socket processor with 17 more models coming soon for two, four, or even eight socket motherboards, though according to The Inquirer Intel has no plans to scale to 16 sockets.  They come in a bewildering array of models including the Sandy Bridge E we are used to, Sandy Bridge EN which uses LGA 1356 and is intended for dual CPU motherboards as it only has one QPI and the LGA 2011 Sandy Bridge EP which scales higher thanks to dual QPI.  No triple QPI but that may still be in store to reduce the number of hops in an 8+ socket board to 2 when used in symmetric multiprocessing in the future.

The E5-2400 (SB-EP) has eight cores and is targeted straight at AMD's lower price, lower power consumption chips as well as offering a noticeable improvement over the already launched E3s.  The E5-2600 family with its dual QPI is more suited for high powered applications that need several powerful processors working in tandem but not to the levels that the E7 series provides.  By offering such a wide variety of choices, especially a family of what for Intel are very low cost processors they are really putting a lot of pressure on AMD and the soon to be released Piledriver family.

Inq_intel_xeon_e5_4600_diagram.jpg

"If you were planning on buying new servers in the coming weeks and months, Intel just gave you a whole lot of homework. And if you work at Advanced Micro Devices, you're getting some homework, too."

Here is some more Tech News from around the web:

Tech Talk

 

Source: The Inquirer

Intel Announces New Xeon Processor to Power the Cloud

Subject: Processors | March 6, 2012 - 02:26 PM |
Tagged: xeon E5-2600, xeon e5, xeon, Sandy Bridge E, lga2011, Intel

XeonE5.png

SANTA CLARA, Calif., March 6, 2012 – Addressing the incredible growth of data traffic in the cloud, Intel Corporation announced the record-breaking Intel Xeon processor E5-2600 product family. These new processors deliver leadership performance, best data center performance per watt, breakthrough I/O innovation and trusted hardware security features to enable IT to scale. These processors are not only at the heart of servers and workstations, but will also power the next generation of storage and communication systems from leading vendors around the world.

Forecasts call for 15 billion connected devices and over 3 billion connected users by 2015. The amount of global data center IP traffic is forecasted to grow by 33 percent annually through 2015, surpassing 4.8 zetabytes per year, more than 3 times the amount in 2011. At these levels, each connected user will generate more than 4GB of data traffic every day – the equivalent of a 4-hour HD movie. This will increase the amount of data that needs to be stored by almost 50 percent per year. ­ In order to scale to meet this growth, the worldwide number of cloud servers is expected to more than triple by 2015.

“The growth in cloud computing and connected devices is transforming the way businesses benefit from IT products and services,” said Diane Bryant, Intel vice president and general manager of the Datacenter and Connected Systems Group. “For businesses to capitalize on these innovations, the industry must address unprecedented demand for efficient, secure and high-performing datacenter infrastructure. The Intel Xeon processor E5-2600 product family is designed to address these challenges by offering unparalleled, balanced performance across compute, storage and network, while reducing operating costs.” The key requirements to enable IT to scale are performance, energy efficiency, I/O bandwidth and security. With the best combination of performance, built-in capabilities and cost-effectiveness, the new Intel Xeon processor E5-2600 product families are designed to address these requirements, and become the heart of the next-generation data center powering servers, storage and communication systems.

Leadership Performance with Best Data Center Performance per Watt
Supporting up to eight cores per processor and up to 768GB of system memory, the Intel Xeon processor E5-2600 product family increases performance by up to 80 percent, compared to the previous-generation Intel Xeon processor 5600 series. The family also supports Intel Advanced Vector Extension (Intel AVX) that increases the performance on compute-intensive applications such as financial analysis, media content creation and high performance computing up to 2 times.

Additional built-in technologies such as Intel Turbo Boost Technology 2.0, Intel Hyper-Threading Technology and Intel Virtualization Technology provide IT with flexible capabilities to increase the performance of their infrastructure dynamically. These performance advances have led the Intel Xeon processor E5-2600 product family to capture 1510 new dual socket x86 world records.

Modern data centers must improve the raw performance they deliver, but also do so efficiently by reducing power consumption and operating costs. The Intel Xeon processor E5-2600 product family continue Intel’s focus on reducing total cost of ownership by improving energy efficient performance more than 50 percent as measured by SPECpower_ssj 2008 compared to the previous generation Intel Xeon processor 5600 series. These processors offer support for tools to monitor and control power usage such as Intel Node Manager and Intel Data Center Manager, which provide accurate, real-time power and thermal data to system management consoles. In addition, Intel’s leadership performance allows IT managers to meet their growing demands while optimizing software license and capital costs.

Sandy-Bridge_EP_7_p.jpg

I/O Innovation and Network Capabilities
With the unprecedented growth in data traffic it is essential that systems not only improve computational abilities, but also enable data to flow faster to support data-hungry applications and increase the bandwidth within the data center. The Intel Xeon processor E5-2600 product family meets these needs with Intel Integrated I/O (Intel IIO) and Intel Data Direct I/O (Intel DDIO). Intel DDIO allows Intel Ethernet controllers and adapters to route I/O traffic directly to processor cache, reducing trips to system memory reducing power consumption and I/O latency. The Intel Xeon processor E5-2600 product family is also the first server processors to integrate the I/O controller supporting PCI Express 3.0 directly into the microprocessor. This integration reduces latency up to 30 percent11 compared to prior generations and with PCI Express 3.0 can up to triple the movement of data into and out of the processor.

The high-performance processing power along with Intel Integrated I/O and advanced storage features such as PCIe non-transparent bridging and asynchronous DRAM refresh, makes the Intel Xeon processor E5-2600 product family also an ideal choice for storage and communications solutions.

Increasing bandwidth demands driven by server virtualization and data and storage network consolidation have led to strong growth in 10 Gigabit Ethernet deployments, with adapter port shipments exceeding 1 million units in each quarter of 2011. Today’s announcement of the Intel Ethernet Controller X540 demonstrates Intel’s commitment to driving 10 Gigabit Ethernet to the mainstream by reducing implementation costs. This industry-first single-chip 10GBASE-T solution is designed for low-cost, low-power LAN on motherboard (LOM) and includes flexible I/O Virtualization and Unified networking support at no additional cost.

Trusted Security
The Intel Xeon processor E5-2600 product family reaffirms Intel’s commitment to providing a more secure hardware foundation for today’s data centers. Intel Advanced Encryption Standard New Instruction (Intel AES-NI14) helps systems to quickly encrypt and decrypt data running over a range of applications and transactions. Intel Trusted Execution Technology (Intel TXT15) creates a trusted foundation to reduce the infrastructure exposure to malicious attacks. These features in partnership with leading software applications will help IT protect their data centers against attack and scale to meet the demands of their customers.

Extensive Industry Support
Starting today, system manufacturers from around the world are expected to announce hundreds of Intel Xeon processor E5 family-based platforms. These manufacturers include Acer, Appro, Asus, Bull, Cisco, Dell, Fujitsu, HP, Hitachi, Huawei, IBM, Inspur, Lenovo, NEC, Oracle, Quanta, SGI, Sugon, Supermicro and Unisys.

price.png

Source: Intel

Intel moves on the server room with 40 processors due for release

Subject: General Tech | January 3, 2012 - 01:58 PM |
Tagged: Intel, server, cpu, Romley, Ivy Bridge-H2, xeon e3, xeon e5

The server room will be getting an update over the next year thanks to Intel releasing numerous CPU models based on different architectures.  First up comes Romley with a total of seven 8-core Xeons, a half dozen 6-core Xeons including both the E5-1660 and 1650 as well as the E5-2640 and relatives, five 4-core Xeons and a single dual core CPU.  That will take us until close to summer.  By then Intel will be working on eleven different Ivy Bridge-H2 series CPUs including the Xeon E3-1290v2 as well as seven more higher end processors including Xeon E5-2470, which will take us towards the end of 2012.

In addition to the regular lineup, DigiTimes also lists four low power Xeons which will arrive in 2012 including the 8-core Xeon E5-2650L.

serverroomimages.jpg

"Intel is set to launch 40 new processors including those for its upcoming Romley platform, in the first half of 2012 with the company to release 20 models each quarter, according to sources from server players."

Here is some more Tech News from around the web:

Tech Talk

Source: DigiTimes

Bulldozers at Knights Corner; duelling server chips

Subject: General Tech | November 16, 2011 - 12:36 PM |
Tagged: xeon e5, xeon, servers, opteron, knights ferry, knights corner, interlagos, hp, dell, bulldozer, acer

As you would expect, no sooner does AMD release news on its new line of Bulldozer era Opterons, Intel follows suit with news on their next generation of server chips.  AMD hit the news and the server room first thanks to interest shown by Dell, HP and Acer.   These vendors have based a series of 2U servers on AMD's new chip as well as a family of blade servers.  Dell's Poweredge C6145 was probably the most ambitious, with 4 sockets you can have 128 cores and 1TB of DDR3 in a 2U rack mount server and FusionIO was suggesting the inclusion of their 1.2TB Iodrive Duo card to ensure your storage media can keep up.

Intel also spoke with The Inquirer and other news sites about their new Xeon E5 processor family as well as providing more information about Knights Bridge. Intel has reached out to a different set of clients for the new Xeon, focusing on NVIDIA's latest target market of High Performance Computing (that HPC acronym you see hanging around Fermi).  They tout over 10,000 chips sold, some of which are sitting pretty in the TOP500.  Also on display was their Knights Ferry accelerator board, again targeted for the HPC crowd that NVIDIA has been courting.

So this processor generation we have Intel and NVIDIA fighting it out for HPC customers, while AMD seems to be without major competition in high density computing, although ARM has certainly been making inroads into that market.  

dell-2u-quad-socket-opteron-185x185.jpg

"AMD's partners have shown a small but impressive array of Bulldozer Opteron kit. Dell's 2U eight socket beast was arguably the most impressive of the lot on show in Munich, but AMD will know it needs more than just one vendor in its fight against Intel. Thankfully it has the might of HP also showing that its traditional rackmount and blade servers can make use of AMD's Bulldozer silicon."

Here is some more Tech News from around the web:

Tech Talk

 

Source: The Register

PC Perspective
Upcoming Events
  • 5/23 @ 6:00pm PDT / 9:00pm EDT
    This Week in Computer Hardware
  • 5/29 @ 7:00pm PDT / 10:00pm EDT
    PC Perspective Podcast