Author: Gil Russell

Intel in talks to buy Altera

Intel Q4_14_ResultsLike something out of Nassim Nicholas Taleb’s book, “The Black Swan, the Wall Street Journal reported that Intel  was in talks to buy Altera Corp. Taleb also predicts that the so called experts will then tell us why it makes perfectly good sense for Intel to acquire Altera – all after the fact of course.

To get an idea what’s involved on the money side; Intel’s market capitalisation is around $140 Billion with Altera at about $10.4 billion.

What premium Intel would have to pay is, of course, one of the finer points of the ongoing discussion. As a basis of estimate analysts are using Intel’s last acquisition of McAfee at $7.7 Billion as a benchmark indicating the acquisition could be in excess of $14 Billion making it the company’s largest acquisition to date if consummated.

Intel stock, which had risen 18% in the past year, rose 6.4% to $32 following the report of the potential acquisition. Altera stock, down 2.5% in the past 12 months, jumped 28% Friday to $44.41.

So, as a sort of red herring for acceptance of the deal, the market reacted positively – considered good feedback for the talks to continue.

Techeye Take – Why Altera?

Altera is one of the anointed companies qualified to run their programmable FPGAs on Intel’s 14 nm Fabs. The two companies have been working closely together in a number of areas and in some cases with involved third parties. Altera FPGAs, for the most part, are not involved in the consumer electronics segment but are directed almost wholly at the high end of the server and HPC markets.

We believe Intel has become deeply involved (nay dependent) on Altera’s Programmable FPGAs in their next generation data center architecture and began suffering pangs of paranoia over the company becoming too exposed to outside influences deciding that complete control over Altera was their only option (taken from Andy Grove’s guidebook; “Only the Paranoid Survive’). [Altera used to belong to AMD, Ed.]


Diablo Technologies Protects Patent ‘917

Screen Shot 2015-03-26 at 10.01.42Diablo Technologies released an announcement yesterday evening claiming a “decisive victory” in a lawsuit brought against the company by Netlist, Inc. (NLST). According to the release a “federal jury in the United States District Court for the Northern District of California ruled in favor of Diablo, with the jury Unanimously concluding that there was no breach of contract and that there was no misuse of trade secrets. Further the jury confirmed Diablo’s sole ownership and inventorship of the ‘917 patent.”

Diablo’s press release further stated:“We are extremely pleased with the jury’s verdict today,” said Riccardo Badalone, CEO and Co-Founder of Diablo Technologies. “We look forward to getting back to serving our customers and delivering on our exciting Memory Channel Storage roadmap.”

The decision further validates Diablo’s position as a thought leader and technology innovator. Diablo Technologies’ Memory Channel Storage™ architecture delivers high-performance flash storage for workloads including database, virtualization and data analytics.

Trial transcripts have not yet been posted on the US District Court, Northern District of California’s website at the time of this writing…,


Samsung Group freezes salaries

Samsung rules the roostSouth Korea’s lead chaebol, Samsung Group, announced on Thursday that it has decided to freeze salaries for its executives next year in an attempt to reduce costs, the first time since the 2009 global financial crisis. The income of approximately 2,000 executives at the business group, which includes Samsung Electronics Co. will remain fixed throughout 2015. Incentive pay will continue and will depend on each division’s profit performance.

The main culprit for the corporate belt tightening is Samsung Electronics unit, which has seen consecutive quarterly earning falls as the company continues to lose market share to Apple and aggressive low cost Chinese rivals. Samsung Electronics has seen its market share drop 7.7% to 24.4% of the global market according to Gartner Inc. Meanwhile major Chinese brands including Huawei, Xiami and Lenovo expanded their global market share to 15.5% in the third quarter – up 4.1% year-on-year.

Samsung Electronics took other cost-saving reduction earlier in the year, which include executive travel in economy class on flights under 10 hour durations and encouraging employees to take vacations instead of receiving pay for unused time off.

Techeye Take

Samsung has been warning for some time now that the profitability of its mobile phone division would be under duress – in fact they earlier declared that the chaebol’s main profit centre would be the semiconductor division which has been reporting good results. The fact that all the group’s executives will share in the discomfort will not be lost on those deemed responsible for their lack of expected Korean management skills…,

Oracle unveils X5 with Intel Inside

Oracle-Announces-X5Oracle chairman and chief technology officer Larry Ellison unveiled X5, its fifth generation of Oracle’s engineered systems, to media and analysts at company headquarters on Wednesday afternoon.

Ellison introduced the company’s X5 as “the future of the datacentre” based on Intel Xeon® E5-2600 v3 processor family (Haswell-EP with up to 32 cores) and support for high bandwidth NVM Express (NVMe) flash drives.

The X5-2, a 1U two socket server, is designed and optimised for running Oracle Database in a clustered configuration. Optional four NVMe drives can be used to accelerate Database performace via Smart Flash Cache. This server is targeted at high-density vitualization environments.

The X5-2L, a 2U platform, is targeted for single-node databases and enterprise storage applications. The supports up to 758GB of memory, and configured for a maximum of 50.4TB of direct attached storage.

Also announced was Oracle’s NVM Express (NVMe) design providing up to 6.4TB of hot-swappable flash providing 2.5X the data rate of older SAS3 SSD interface drives using PCIe Gen3 Small Form Factor NVM SSD drives (12Gb/s vs. 32Gb/s). NVM Express flash technology is optimized to accelerate Oracle Database using a feature called Database Smart Flash Cache. This feature keeps recently accessed data warm in flash storage, reducing the chance that the database needs to fetch the data from slower magnetic media that may be direct attached or resident on a NAS/SAN fabric. In addition to the high-bandwidth interface to the NVM Express SSDs, the flash technology itself has been engineered to be high-endurance and write-optimized for Oracle Database.

Ellison’s new “vision” entails connecting datacentres efficiently and at lowest cost to the cloud – “There has to be some degree of compatibility between the public cloud and your private datacentre”, Ellison said.

Ellison emphasised Oracle’s “new strategy” using Intel processors to compete for the two-socket core business. The new “Virtual Compute Appliance X5” converged infrastructure system, consists of compute servers and software defined networking.

That integration comes in the form of th Virtual Compute Appliance X5 converged infrastructure system, consisting of compute servers, software-defined networking and Oracle designed hardware. Ellison went on to highlight the company’s abilities in software defined configuration of server and storage networks on VCA, supporting infiniband internal networking with external connectivity provided by Ethernet and Fibre Channel to link with existing networks.

Included within the X5 product portfolio are Oracle’s Big Data Appliance for Hadoop and NoSQL big data jobs and Exalogic X5-2 for private clouds.

Ellison described Oracles Zero Data Loss Recovery Appliance capable of full data recovery with real-time “redo” transport and fully automated recovery functions, log re-examination with extraction of malicious transactions followed by re-entry of those processes again allows the appliance to be restored to any point in time.

Further the appliance, which can handle thousands of databases with backup connections to on-site datacentre, remote datacenters and cloud. “The big deal is it’s fully automated, so it’s easy to operate, and you never lose data. It’s a no brainer appliance as we have, “Ellison stated.

Ellison reminded the audience that “Oracle manufactures tests and supports all of these products in-house”, naming rivals Cisco, EMC, VMware, Microsoft and Red Hat hinting at more expensive and fragmented support by rivals. Further “One appliance alone can handle thousands of databases with potential backup connections to on-site datacentres, remote datacentres, and the cloud.” he said.

“The big deal is it’s fully automated, so it’s easy to operate, and you never lose data. It’s as a no-brainer appliance as we have,” Ellison remarked.

He further stressed Oracle has manufactured, tested, and support all these pieces in-house, calling out rivals Cisco, EMC, VMware, Microsoft, and Red Hat and hinting at more fragmented (not to mention expensive) deployment options. All X5 machines are available now.

TechEye Take

The rumor of the Intel invasion of Oracle has been circulating since OracleWorld 2012. This is a major shift for Oracle. The company’s management, currently in the midst of a “reinvention period”, includes the fact that Larry Ellison is executing a gradual accession plan as he moves toward retirement.

The X5 release is seen as one aspect of the company’s new strategy – one in which the company protects their private datacentre market base while adjusting to a world increasingly enveloped by the evolution of open hardware, software and the cloud. Ellison is a sharp toothed shark and Oracle is having a problem finding a way to replace his natural instincts – how this evolves is another one of those “only in the valley” stories.

It is looking like a very good year for Intel’s E5000 series though…,


Intel earnings Jump: it’s memory bundling

Intel Q4_14_Results

Intel released its fourth quarter 2014 results yesterday afternoon with income jumping 39 percent on improved demand for personal computer and server system chips. The company allowed that it is  expecting a somewhat flat first quarter for 2015 which led shares 1.9% lower in after-hours trading. The PC Client Group’s earnings improved by three percent while the Data Centre Group’s earnings improved by 25 percent. Overall revenue increased by six percent  year-on-year and gross profit margin exceeded 65 percent.

Analysts polled by Thomson Reuters expected per-share earnings of 66 cents and revenue of $14.71 billion.

For the current first quarter, Intel projected revenue between $13.2 billion and $14.2 billion and gross margin of 60 percent, plus or minus a couple percentage points. Analysts, on average, were expecting revenue of $13.76 billion and gross margin of 61.2 percent, according to Reuters.

For the year, Intel projected revenue to rise by a mid-single digits percentage rate and achieve a gross margin of 62% of revenue, plus or minus a couple percentage points. Analysts, on average, were expecting revenue to rise 4% and gross margin of 63.4%, according to Reuters.

PC Client Group Improves

Intel’s integral attachment to the PC market greatly affected earnings as the PC market growth slowed and consumer market demand was satisfied with less costly tablets and high capacity smartphones. The uptick in PC demand last spring has had a positive effect on earnings and aided in the company’s turnaround effort to become “the” dominant supplier in the mobile market. With Intel’s 14 nm manufacturing muscle Brian Krzanich is now “loaded for ARM” vowing to place 40 million Intel chips into tablets now dominated by ARM Holdings PLC.

3D 256 Gb NAND-Flash Bundling?

No mention was made by Intel of its  recently announced 3D 256 Gb NAND-Flash devices. Intel has what can only be called an obsession with its ability to control the memory side of the sales equation without owning any of the fixed assets to produce it.

Analysts have been wondering why Micron was not more upbeat on the announcement; it is,  after all the controlling partner in Intel-Micron Flash Technologies, Inc. (IMFT). Sources indicate that Intel will most likely begin bundling Processors and Memory kits with Intel claiming the lion’s share of margin leaving Micron to its own pursuits with its share of output.

Last but Not Least

The Data Centre Group is rumoured to be the earnings darling of the coming quarters with sources indicating market moving announcements over the first half of 2015. Those announcements concern Intel’s SDN for Cloud Computing efforts…




Samsung browses Blackberry

Samsung Browses BlackberryA breaking story  today entails Samsung’s approach to Blackberry Ltd. in talks to buy the embattled smartphone maker for a reported $7.5 billion – Samsung is apparently buffing up their intellectual property portfolio to stave off a continuing onslaught by Apple Inc. The news sent Blackberry’s stock price up 30 percent.

Samsung offered a trading range of $13.25 to $15.49 per share, a premium of 38 percent to 60 percent over Blackberry’s current trading price according to Reuters. What’s not clear is the depth of the deal and what it might entail – there’s speculation that the deal has several versions that are currently under discussion. That the story was leaked and by who leaves one wondering whether this is a negotiation tactic or just business as usual.

BlackBerry announced in November a high-profile security partnership with Samsung. The partnership aims to wed BlackBerry’s security platform with the South Korean company’s own security software for Galaxy devices.

Blackberry has been struggling to regain lost momentum in a competitively crowded market. Samsung’s smartphone business is experiencing losses from lower price Chinese competitors which now are affecting their semiconductor division – reports of bargain prices for NAND-Flash smartphone memory devices are widely circulated.

Techeye Take

Analyzing Samsung’s needs, wants and desires indicate the company is obsessing over security – key to the company’s entry into the electronic wallet market space. Secure communications is and will continue to be a premier element of the continued evolution of the smartphone market – even governments demand it.

The question remains, will this set off a bidding war for Blackberry?..,


Both Samsung and Blackberry have denied that they are in talks for Samsung to takeover the Canadian company.

“BlackBerry has not engaged in discussions with Samsung with respect to any possible offer to purchase BlackBerry,” the company said in a statement Wednesday.

Post Script

Where there’s smoke there’s generally fire – stay tuned…,

Intel announces 3D NAND-Flash

IMFT Sign - Lehi

Rob Crooke, VP & GM of Intel’s Non-Volatile Memory (NVM) Solutions Group was last up in the company’s day long Investor Meeting today in Santa Clara.

Though last, he had the most newsworthy announcement about the company’s future memory intentions.

Intel announced it is back in the memory business – 3D NAND-Flash that is (mass production in-house is conditional though).

Crookes’ revelation ends any rumination on Intel-Micron Flash Technologies 3D Flash development – it also includes SK Hynix when the device goes into production 2Q 2015. Evidently those who have been nice have early sample devices according to sources.

The specifics:

  • 4G hole array 32 layers deep | (216 x 216)(Array) x 25(Layers) x 2(MLC) = 256 Gbits
  • 1TB in 2 mm package
  • SSDs: 10TB and up planned
  • Production 2H 2015 – IMFT (Lehi, Utah facility mentioned) & SK Hynix
  • Intel can also produce internally
  • Replacement of HDD with SSD in all PC and Mobile devices

Crooke allowed that the devices will not use Intel’s cutting edge 14nm technology but a slightly relaxed geometry  – Micron is on record at 16nm geometries for 3D NAND. The openly known fact that prevaricating about Flash Geometries may hold sway – a hefty dose of caveat emptor is recommended.

The announcement coincides with reports that Intel and Micron are involved in a project with EMC2-DSSD – an effort to produce the first NAND-Flash In-Memory Database appliance.  The proffered memory type may be a custom type expressly tailored for the application and may be produced in-house by Intel – more on this as roll-out time nears.

IBM pays Globalfoundries $1.5 billion to cut losses

IBM Not Servers Not SemisUpdate: IBM advised late Sunday that it will be making “a major business announcement” Monday morning along with its’ third-quarter results according to Bloomberg News.

This comes as no surprise to industry watchers. Sanjay Jah, CEO of Globalfoundries Inc. is well known as a shrewd negotiator – fees for taking over IBM’s semi ops were reported as high as $2 Billion.

IBM is shedding the company’s brick and mortar structure piece by piece to facilitate what the company sees as its new destiny. Commoditisation of semiconductor and hardware server content is seen as the motivation.

Margins in both businesses have decreased to the point where economies of scale must come into play – requiring ever-larger investments with ever decreasing margins draining capital away from the company’s core business strategy.

IBM announced that it is investing $3 Billion over five years on semiconductor research in a move to reassure its customer base that the company is continuing basic research to advance hardware and software technology indicating that the company will still be supporting high end research.

What’s not clear at this time:

  • Number of people affected
  • Timeframe
  • Whether all semiconductor operations are included
  • Power PC
  • Intellectual Property
  • Operating charges

We’ll be following up on details as they become available.


In a statement made this morning IBM will pay Globalfoundries Inc. $1.5 billion over the next three years to take over the company’s loss ridden semiconductor operation.

Globalfoundries will become IBM’s exclusive server processor foundry for 22nm, 14nm and 10 nm server and Power processors for the next decade.

IBM will take a third-quarter pretax charge of $4.7 billion.

 Job Retention

Globalfoundries will offer jobs to those affected in East Fishkill, New York and Essex Junction, Vermont. Workers at IBM’s commercial microelectronics business are also included in the offer.

Globalfoundries will gain access to IBM intellectual property and technologies related to IBM Microelectronics under the 10 year partnership agreement – making the foundry one of the largest semiconductor IP portfolios holders in the world.

 Trusted Foundry

The U.S. Government has used IBM as a supplier as a “trusted foundry” supplier for decades. Globalfoundries is privately owned by United Arab Emirates company called Advanced Technology Investment Company, or ATIC, a subsidiary of the Emirate of Abu Dhabi.

There’s speculative agreement that Intel is the most likely candidate to take over the “trusted foundry” business from IBM.

AMD Cuts Workforce – Sea Change 101 for Sailors?…,

AMD LayoffsJust one week and a day after assuming her new roll as AMD’s CEO Dr. Lisa Su announced a reduction in force amounting to seven percent of the AMD’s current workforce of 10,149 employees.

The fallowing of ~700 people follows two rounds of layoffs under Rory Read’s three year tenure.

AMD did not provide any information about where the cuts would be made – the company recently split into two divisions “Computing and Graphics”, and “Enterprise, Embedded and Semi-Custom”. Contacts indicate that “Computing and Graphics” will receive a majority of the “hits”.

Last Thursday’s warning by Microchip Chief Executive Steve Sanghi that a correction will spread more broadly across the industry in the near future sent shares of chipmakers lower last Friday.

Microchip is a broad spectrum supplier into the Chinese and Asian marketplace, booking revenue only after it is shipped by distributors – a closely coupled supply chain that quickly indicates impending sea changes. AMD shares a similar situation in China, the company’s biggest market, substantiating Microchip’s warning. Whether this is the beginning of a prolonged downturn or is merely another “noise blip” on the radar is entwined in controversy.

Strangely enough, AMD’s arch nemesis Intel, reported rather glowing results on Tuesday indicating that AMD might be suffering from Intel’s competitive resurgence in Asia. The fact that Intel is devoting resources to system level integration at the SoC level may now be having an effect on both competitors.

AMD’s experienced a 65% drop in quarterly profits and is expecting the current quarter to be 13% lower than the period ended in September. The company’s share price fell 6% to $2.49 in after-hours trading. The share price has dropped 43% in three months as of close Thursday.

Su went on to assure analysts that the company was moving toward customized chips for applications beyond videogames hinting at two customers that had the potential of bringing in $3 Billion in additional revenue over the next three years.

TechEye Take

The first time I saw Rory Read perform in front of analysts was somewhat of an embarrassment. He became so animated on stage that an additional two flaps of his arms per minute would have gotten him airborne (I heard that he fired his stage coach soon thereafter). His resignation came as no surprise, only late by three years. Lisa Su was the only stand-up with credibility and has remained so since.

Can we expect Dr. Su to right the AMD ship? She’s very smart and well experienced in the land of semiconductors and if anyone can accomplish the miracle required to make AMD a player she’d be my pick…,

Micron a $50 stock – maybe more…,

Micron_Lehi_UtahBarron’s claims that DRAM demand and a lack of producers will drive Micron’s share price to over $50 in their October 6th issue. They cite business PC replacement and Big Data as the market drivers behind the price climb and the fact that there are only three major producers remaining.

The simple deduction is that the DRAM market will be capacity limited for the foreseeable future.  Of course this doesn’t factor in splits between Flash and DRAM demand confusing the production mix – end result is a higher price for both.

An interesting nuance to Barron’s forecast for Micron is the introduction of a next generation non-volatile memory that reduces the price of storing very large database images.

Glimpses of HP’s version in “The Machine” using Memristor based memory is scheduled for launch in 2018 – implying that the first production devices will need to be extant by early next year. HP’s record on the Memristor Project has missed each and every promised milestone so the success expectation probability is low.

Tell Tales Out of School

An intriguing story making the underground rounds in the Valley concerns the existence of an extremely secretive program involving a new, high speed, non-volatile memory coupled with DRAM. No it’s not the Diablo Technologies, Inc. Memory Channel Storage (MCS)  – though somewhat similar it couples extremely dense non-volatile storage with low-latency parallel caching loads of high-speed low-power DRAM main storage.

The membership is limited to an exclusive set of players on both the supplier and user sides.

This is in step with a major effort to move from SATA serial interface non-volatile memory (SSD) to a high performance parallel interface. The discussion centers on whether the transition will include NAND-Flash or will begin a fresh start with the next generation replacement.

The idea has begun to percolate through the JEDEC Standards Committee. Sources predict that this will be accelerated through the standards process by an influential member group at JEDEC.

Killer Elite Application

What is the application – the one that motivates the factory to produce massive amounts of these devices. My contact looked me straight in the eye with that “you idiot look” and exclaimed, “Everything”. That’s when I got it…,







HP joggles PC/printer divisions – again

Whitman's-SamplerOnce Again, HP has decided to evolve the PC and Printer operations as a distinct and separate corporate entity.

HP came close to selling both divisions during the short reign of Leo Apotheker. After the discovery of a massive over payment for Autonomy Corp. HP’s Board decided Leo had to go and PC & Printers had to stay.

Slipping to the number two position behind Lenovo, HP has decided to spin the combined organization into a separate entity under the aegis of Dion Weisler as CEO (Weisler is an exec in the PC and printer operation currently). Patricia Russo will be installed as the Enterprise company’s new Chairman (former lead independent director).  Meg Whitman will remain CEO of the Enterprise company and oversee corporate guidance of the PC/Printer entity as Chairman.

What difference does this make? Reporting structures loaded with changes in culpability mostly, freeing Whitman up for minding the Enterprise store and:

  • Aligns Weisler for the fall when and if the PC/Printers Division comes in under plan.
  • Allows time to position the PC/Printer Group for a potential sale.

HP has been struggling in their efforts at penetrating the Cloud with their Moonshot technology – Whitman may find the ice a little thin for skating this Winter and into next Spring.

HP’s merger discussions with EMC recently ended. We’re left wondering if what we are now seeing is part of a “Plan B” by HP’s Board of Directors…,


Intel shows off in-memory-database Biz

Intel-IDF-'14-Copy-SizeIntel’s Developer Forum 2014 annual meeting at San Francisco’s Moscone Center wound down yesterday. My assignment is to continue research on a technology that’s now ramping.

The computer industry is at the beginning of a major architectural shift – “In-Memory Database” (IMD) systems, originally aimed at solving near real-time solutions for analytic problems have successfully been applied to cognitive computing problems as well. The nascent application of “cognitive computing intelligence and predictive analytics” toolsets to IMD equipped servers is thought to be the first step in a new era in computing – quite possibly the next big thing.

The Google Effect
At the 2000 Intel Developer Forum in San Francisco a relatively unknown entrepreneur, while having a Keynote fireside chat with Andy Grove, said he’d like to take the entire Internet and put it in memory to speed it up – “The Web, a good part of the Web, is a few terabits. So it’s not unreasonable,” he said. “We’d like to have the whole Web in memory, in random access memory.”

The comment received a rather derisive reception from the audience and was quickly forgotten. The speaker, Larry Page, an unknown at the time, as was his startup company, Google – the company’s backbone consisted of 2,400 computers at the time.

Fast forward to the present – system vendors found their future in Big Data has a lot of the look and feel of Google’s “free to the public” offering. Google was the first to successfully deploy a massively parallel processing (MPP) network commercially using commodity servers – one that was delivering real-time data access on a worldwide basis. Their competitors realized that they could no longer remain competitive with systems that relied on high latency rotating magnetic media as the main store – in fact, solid state disks (SSD) are considered somewhat slow for the new realities of Big Data analytic computing.

The development – called “In-Memory Database” mounts the entire database (single system image – even enormous ones) into large scale memory arrays of Registered DIMMs – closely coupled with Multi Core Processors. The resulting increase in throughput accelerates not only transaction processing but also analytic application performance into real time. The frosting on the cake is that this architecture change applies to good advantage in the emerging cognitive computing space.

SAP – HANA, In-Memory Database Computing
In 2006 Hasso Plattner, Co-founder of SAP AG, took a bottle of red wine, a wine glass, some writing implements and paper to the garden behind his house. By the time he reached the bottom of the bottle there wasn’t much written on the paper. But he had reached the conclusion that in-memory systems were the future. Mr. Plattner had realized that for SAP to remain competitive it needed to innovate – Plattner believed that by changing the server design to accommodate massively parallel processing with enough memory to load an entire database when combined with columnar based storage software would have a revolutionizing effect on processing speeds for OLTP and OLAP applications.

Gathering a small group of PhDs and undergrads at the Hasso Plattner Institute, Plattner expressed the in-memory idea he wanted them to explore. The first prototype was shown in 2007 before an internal audience at the company’s headquarters in Waldorf, Germany. SAP management was skeptical that the idea would work – the team needed to prove that the concept of in-memory database would work under real world conditions.

Using contacts to advance the project, Mr. Plattner persuaded Colgate-Palmolive Co. to provide transaction data for the project. He also persuaded Intel’s Craig Barrett to secure the latest microprocessors for the labs ongoing effort. The company also set up an R&D facility in Palo Alto to be in close proximity to their innovation and research partner Stanford University.

SAP HANA was officially announced in May 2010 with shipments commencing with the release of SAP HANA 1.0 in November. The market was slow in adopting the technology convinced that it was still in an early stage of development. Analytics and the need to score a real reason for their customers to mount their IT to the cloud provided the market conditions SAP’s HANA needed to press its adaptation. SAP over time adopted HANA to the Cloud through successful partnering with a wide array of vendors making it the company’s fastest growing segment.

During the development of HANA, SAP discovered the amount of physical memory required to store an entire database could be reduced substantially (compressed) – in some cases by 100X. This had the effect of reducing power (less memory required) and made database searches more efficient (reduction of the empty set). The market implication was that the price of memory per gigabyte had finally reached a price/performance breakeven point in an application that could not be accomplished at that price any other way. DRAM producers have found their next “Killer Application”.

IBM’s Watson – Cognitive Computing Public Debut
IBM’s Watson is a Big Data analytics system running on 2,880 PowerPC cores with 16TBytes of DRAM. Estimated cost is reportedly just over $3 Million and it requires 200kW of power to operate. Watson’s inner workings have not been publicly released – what is known is that it runs under a tool IBM calls DeepQA, implemented in conjunction with Hadoop (a Java implementation of MapReduce) that runs under the SUSE Linux Enterprise Server Operating System.

IBM introduced Watson to the public by competing it against human opponents on the game show “Jeopardy” in February 2011 – establishing IBM and the Watson Brand in the public mind when it won the $1 Million Dollar prize for charity.

Watson’s ability to semantically interpret language implies a native ability to understand the context of questions – including puns and word plays that it handled amazingly well – questions of this nature typically remain a significant challenge for machine-based systems.

Watson’s creators have stated that the algorithms are “embarrassingly” parallel – the implication that the core engine is highly MapReduce in nature rather than the more traditional graph analytics approach. Conventional network control is adequate for such an engine reducing costs and falls within a Software Defined Networking (SDN) framework.

IBM previously missed the industry shift to data management from ISAM files to relational databases in the 1970’s even though they were the inventor of RDMS systems. Oracle took full advantage of this colossal gaff much to IBM’s dismay.

IBM announced the establishment of the Watson Business Unit in early March investing upwards of $1 Billion in the new entity. What is surprising is that the company had a fully established cloud based offering replete with a supporting ecosystem around Watson (now physically occupying three rack cabinets instead of the original nine). There is no lack of customer interest in Watson with over 1,000 third party developers signed on to date.

IBM emphasizes Watsons’ natural language capabilities and analytics to process and synthesize information in a manner similar to the way humans think – enabling quick comprehension and evaluation of large of amounts of human style communication data to generate and evaluate evidence based hypotheses – to adapt and learn from training, interaction and outcomes.

Server Commoditisation – IBM Going Fabless?
“Watson” is at the beginning of a bundling “strategy” by IBM that’s in line with its continued separation from its hardware origins. IBM’s internal politics sometimes show in decisions made by disparate groups within the company in efforts to preserve their own “silage”.

The persistent and widely spread rumor that IBM was selling their low-end server division began circulating in April 2013 with Lenovo the most likely buyer – it passed into obscurity before becoming a reality in January 2014. The trend toward server hardware commoditization is the driving force behind the sale. Margins in the low-end server space have decreased to the point where economies of scale must come into play – requiring ever-larger investments with ever decreasing margins draining capital away from the company’s core business strategy. Watson, on the other hand, is viewed as a “maximum best-fit scaling technology” for capitalizing on IBM’s capabilities as a company.

Recent rumors that IBM is accepting bids for its semiconductor operations are being taken seriously and lean toward Global Foundries as the favored bidder. IBM announced that it is investing $3 Billion over five years on semiconductor research in a move to reassure their customer base that the company is continuing basic research to advance hardware and software technology. The company has entered talks of selling the East Fishkill, N.Y. Fab to Globalfoundries Inc. though a definitive agreement has yet to be announced.

IBM is slowly being transformed into a mostly software and services company using commodity, software defined hardware. That it’s going fabless is no surprise – the question of who will fill the void of developing next generation semiconductor processes and the attendant processor architecture development.
In 2013 the odds were firmly on Intel – the lack of furthered commitment in IDF 2014 shakes this conclusion but remember that the E7 version will not be ready for prime time till next year or at best very late this calendar year.

IBM, deciding to take Watson to market, set out to solve cost, power and footprint issues through industry collaboration. The effects of this collaboration will have far ranging effects on the company, its hardware product line and industry partners.

IBM’s larger than usual presence at the Intel Developer Forum in 2013 with a keynote delivered by Diane Bryant, Intel Senior Vice President and General Manager of the Data Center Group further signaled IBM’s continued segue with Intel toward high end servers.
Intel’s Rack Scale Architecture

Intel has been developing its version of the “Disaggregated Server” named “Rack Scale Architecture” or RSA.

At the core of the Rack Scale Architecture is a technology Intel calls “Silicon Photonics” – developed under the premise that a system wide integrated silicon photonic based data highway woven into a hierarchical communication fabric will support massively parallel computing systems into the foreseeable future and remain a baseline architectural model for future growth. Copper interconnects do not scale reliably in server systems at data rates much above 10 Gbs per channel (multiple fiber channels (10) are combined to establish interconnects like 100 Gbit Ethernet).

The idea of a “silicon photonic” highway offers system architects freedom to allocate computational resources “at will”. This blends well with Software Defined Networking down to the computational element – essentially making an entire data center a virtual machine.

Key to this idea is the use of fiber optic cable capable of carrying 100 Gbps and up data channels (cluster of 4 fibers at 25 Gbps each) called “Silicon Photonics” by Intel.

Diane Bryant brought Andy Bechtolsheim – Founder, Chief Development Officer and Chairman of Arista Networks on stage to announce the company’s first shipments of the “Top of Rack Switch”. Bechtolsheim stated that Intel’s Silicon Photonic’s solved the cost issue allowing Arista’s TOR Switch to enter the market. Andy added that switches extending transmission distance from 200 meters to 2 kilometers required for Cloud data centers would be shipping in volume in Q1 CY 2015.

Intel’s Big Data Analytics Market Outlook
Diane Bryant saved the best for last in her keynote segment. She stated that McKinsey reported big data analytics can improve margins up to 60% through increased sales per visit through improved management of inventory and through optimized product pricing. Cost of compute has declined 40% and the cost of storage has declined 100% making it truly cost feasible to deploy these big data analytic solutions. She added that the E5V3 analytic server units were announced in a separate announcement on Monday. Unfortunately nothing was said about the massive E7s now in development.

Bryant went on stating “within a couple of years Hadoop will be the number one application. It will be running on more servers than any other single application. It will be more common for Enterprise IT than Enterprise ERP system. The big data market is growing at 35% CAGR it’s projected to be a $150 Billion business in silicon systems, software and professional services by 2020.”

TechEye Take Away
We’re not sure what happened between IBM and Intel. Comparing IBM’s presence last year compared to this year’s IDF was completely different. Relationships between companies can take wild swings over internal problems that are kept far from the public eye and we suspect that this may well be operative here. IBM is most interested in the E7 version which remains unannounced though sources report this is scheduled for some time in Q1 2015. We think the apparent lack of mutual devotion is temporary and helps to quiet internal silo wars at IBM for the time being.

Do not be surprised if Intel’s Data Centre Group breaks out into a separate, standalone forum next year.

Intel is working on multiple technology fronts to develop next generation data center architectures capable of real time transaction processing and analytical processing. Keep also in mind that these machines are completely capable of running Cognitive Intelligent Computing currently the domain of IBM but will first ramp in 2015 in an application span called Cognitive Analytics.

Remembering that analytics also includes voice and real-time voice translation leaves wide implications into a number of consumer space applications – think of a gate keeper service melded into cellular phone contracts.

In any regards Mark Bohr is still holding court over Intel’s process development – one of the company’s solid IDF anchors that’s still left at the company. The news is that Intel can build 14 nm FinFet 300 mm wafers in volume and is well on its way to 7 nm with a stop at 10 nm.