AND ITS NOT GOING TO CHANGE IN 2013 EITHER..............
2014 NEW WAR ERUPTS AS ARMH BASED SOLUTION ATTACK SERVER MARKET
range bound stock ..As lion roared here and than single digit in 2014 is coming up
Sentiment: Strong Sell
Mark Davis deserves credit for his armserversDOTcom article. The least that the lion should roar is the credit to the real author. Any idea who Mark Davis is and what value should be given to his May 2012 opinioins?
What is an SoC? Hint: the “S” stands for Server.
May 14, 2012 By Mark Davis
Singh only shows up hollering "TORA! TORA! TORA!" and typing in all caps when the situation is beyond repair for the short side. All longs should take comfort in his presence. You go, Singh - give full expression to the looney tunes side of your personality...
Sentiment: Strong Buy
The acronym “SoC” generally refers to “System on a Chip”. But with SoCs entering the server space, it is also taking on a new meaning: “Server on a Chip”. An SoC is a large scale integration of processor cores, memory controllers, on-chip and off-chip memories, peripheral controllers, accelerators, and custom IP (intellectual property) for specific applications and uses. As Moore’s law continues, chip process geometries shrink, allowing more transistors to reside on the same area of silicon. Traditionally, server processors have used this new real estate to add more cores. But there are better alternatives than just adding more cores for certain applications.
Increasing integration in an SoC brings a number of benefits including:
Higher performance – significantly faster and wider internal busses compared to those found in a multi-chip or multi-board solution.
Lower power – wider range of power optimization techniques can be employed in SoCs including power gating, changing bus speeds depending upon utilization, dynamic voltage and frequency scaling of processor cores and peripherals, multiple power domains, and a number of others. Additionally, having peripherals on chip avoids power hungry PHYs (analog drivers that need to drive signals between chips and boards).
Higher density – fewer components to buy, consume power, and fail.
Deeper integration of peripheral controllers and fabric interconnect technologies allow a number of advantages that cannot normally be achieved by having to go through standard bridges like PCIe.
Let’s stop and consider the components we typically will find in a standard rack-optimized volume server:
One or two processor chips, often with integrated memory controllers.
One or two chips for processor chipsets providing a range of functions like Southbridge peripherals and PCIe.
A PCIe connected Ethernet NIC, either chip or PCIe board. In today’s volume servers, this is typically one or two 1 Gb Ethernet interfaces.
A PCIe connected SATA controller, either chip or PCIe board.
Controller chip for an SD card and/or USB.
An extra cost, optional BMC (baseboard management controller) providing out of band system management control.
So, now with the availability of a purpose-built ARM® server SoC, how does this change? Everything in the laundry list above gets integrated onto a single, low power die. For example, let’s take a look at the Calxeda EnergyCore ECX-1000 series of SoCs. In each chip, we find:
A quad-core Cortex A9 CPU, configured for server workloads.
The largest L2 cache that you’ll find on an ARM server: 4 MB with ECC.
A server class memory subsystem including a wide, high-performance 72-bit DDR3/3L memory controller, also including ECC.
Integrated peripheral controllers that have direct DMA interfaces to the internal SoC busses without the PCIe overhead. Standard server peripheral controllers like multiple-lanes of SATA, multiple Ethernet controllers (both 1 Gb and 10 Gb), even an SD/eMMC controller for local boot or scratchpad storage, are all integrated on-chip.
If your server needs to connect to devices that are not integrated, there are four dual-mode PCIe controllers, supporting both root-complex and target modes, in both x4 and x8 configurations.
Instead of an optional (and expensive) BMC, management is built onto every chip, providing a sophisticated server management system that provides both in-band and out-of-band IPMI/DCMI system management interfaces along with dynamic power and fabric management.
A deeply integrated, power and performance-optimized fabric interconnect, which we’ll talk about in a future blog entry.
And all of this is designed with performance, power, and cost optimized servers in mind, delivering the industry leading performance/Watt and performance/Watt/$ servers.
Qualcomm joins ARM-based server race
Chip giant advertising for engineers to work on system-on-chip for low power server space, challenging Intel
The ARM processor architecture is looking well beyond its mobile device heartland and eyeing a whole range of products from embedded gadgets to servers. Several chip providers are chasing the emerging market for low power blade servers for the cloud, an area where ARM thinks its licensees can displace the incumbent Intel. Qualcomm has joined Marvell, pioneering start-up Calxeda and others in planning a system-on-chip for servers and is advertising posts for specialists for this project.
According to EETimes, Qualcomm's job advertisements, for at least three software engineers, reveal its server ambitions. They say the engineers will work on "architecture/design and system prototypes of Qualcomm's new ARMv8-based server SoC ASICs for the power optimized server market.
The ads also say: "Primary responsibilities include specification, development, porting, integration and shake-out of server platform management software and firmware on prototypes utilizing Qualcomm's new SoC."
The nascent ARM-based server movement is expected to get a major boost from the ARMv8 architecture, the first generation to support 64-bit instruction sets and so level with Intel's x86 in the high end segment. A rush of commercial products is expected in 2014 once v8 has been fully commercialized. However, some early movers like Calxeda and Marvell are already shipping 32-bit server SoCs, though these are mainly to steal a march on rivals and attract customer interest, rather than mass sales. Cloud giants like Microsoft are already testing prototype ARM-based platforms (Microsoft has carried out projects with Calxeda) and Dell, HP and Mitac have all worked on 32-bit prototype platforms.
Among the other chip firms known to be working on ARM server SoCs are AMD, Applied Micro, Cavium, Nvidia and Samsung, and Huawei advertised for ARM server engineers last year too, though these may be for boxes rather than silicon. Texas Instruments has also announced server SoCs, but is targeting them mainly at specialized companion servers or embedded platforms rather than taking on Intel in the data center.
The companies known to be preparing ARM server chips generally aim to start shipping 64-bit SoCs in 2014. Calxeda and Marvell are shipping 32-bit SoCs for servers today with limited traction, mainly in prototyping systems for Dell, Hewlett-Packard and Taiwan's Mitac.
A server move by Qualcomm would be another step away from the firm's previous strategy of sticking to mobile devices - just a few years ago, it said it wanted to concentrate on what it did well, but with its core smartphone sector facing price competition and eventual stagnation, it must seek new bases just as the ARM platform is - in tablets, 'post-PC' products, the internet of things, home networking, and in future servers and wireless infrastructure, especially small cells.
ARM to get server revenues next year, says East
ARM signed two architecture licenses in Q4 which is a rare thing for ARM to do. It has only given out about a dozen in its 22 year history.
The architecture licenses were for ARM's new 64bit v8 processors which are principally aimed at servers.
ARM said the architecture licenses were intended for "use in multiple end markets including supercomputing, mobile and enterprise."
EW asked ARM CEO Warren East who the licensees were but he wasn't going to say.
Asked if they were for chip design companies or for people who own a lot of servers, he declined to be drawn.
A recent event for ARM was being included in Facebook's Open Compute Project which specifies interchangeable x86 and ARM processors for its server farms.
Is it a new thing for ARM to work with its customers' customers?
"It's always been the case that we work with the thought leaders in a particular space," replies East, "we have had a direct relationship with Nokia since the 1990s. It's part of the intelligence gathering process. We try to work with the thought leaders."
ARM's penetration into servers has been a long time coming. "Expect to see some server revenues in 2014. That has been the same for the last five years," says East, "five years ago I said we'd have our first server revenues in 2014."
Asked if there was any significant metric he could point to which showed ARM did a superior job to x86 in the server space, East replies: "Some of the results from the HP Moonshot project compared ARM with existing solutions which ran on x86, and looked at the cost of ownership, and they were fairly stunning results which were better than the theory."
ARM's pervasiveness rolls on. Sales of ARM processor into consumer products including digital TVs and STBs doubled in 2012 and sales of ARM-based microcontrollers went up 25%.
Progress into laptops was not much helped by Microsoft's Surface which runs Windows on ARM, but East points out that Surface was only the beginning of Windows running on ARM - and it was a very restricted beginning with very few companies engaged in the process.
Whether finfet or FD-SOI makes the running at 20nm is not a critical factor for ARM whose designs can be made on either process.
"FD-SOI looks very good to us," says East, "the libraries are the same as for bulk. That's one of the good things about it. It’s a much easier process than making fins."