IBM Eye

Layoffs continue at IBM, this time it seems concentrated on GTS (Global Technology Services) employees, with 1600 more anounced May 30th. IBM didn’t disclose where they were, but according to local news, about 100 of those are in Triangle, 50 in Colorado, a dozen or so in Endicott, 72 in MA and 25 in South Florida.

An anonymous IBMer metioned to me that there were 1000+ in Atlanta earlier this month, but I didn’t see that confirmed anywhere. It seems like that size of number would have been all over the AJC.

IBM today announced what it claims to be the “fastest microprocessor ever built.”

Per the www.theregister.co.uk/2007/05/21/ibm_power6_p570/

The updated p 570 system that can hold one to eight of the new 4.7GHz dual-core Power6 chips. When running the TPC benchmark, an eight-way version of the p 570 reached a score of 1.6m, which IBM reckons is three times the performance per core of HP’s top Itanium-based Superdome server. The bad news, however, is that customers won’t be able to buy the exact IBM box and software used in the TPC test until late November. HP is expected to have updated its Itanium servers by then.

It is built in a 65 nm process, houses 790 million transistors and runs at a clock speed of 4.7 GHz. (Coincidentally, the original IBM PC had a 4.77MHz CPU). Compared to the Power5 chip, the new version offers more than double the clock speed, four times the cache (8 MB) and surprisingly, has a similar power consumption profile.

Designed to work in virtualized server environments, a mid-sized Power6 powered server is capable of running 50 or more AIX partitions and to copy one partition to another with a mere few seconds of interruption. It has as many cycles / second as there are bytes on a DVD. In terms of bandwidth, the Power6 offers users an aggregate bandwidth of 300 GB/s – which in theory is enough to download Apple’s entire iTunes catalog in about 60 seconds.

Read more:
The Register (they broke the story - they’ve changed the text throughout the day, but this morning they had test results from Oracle’s website that were promptly removed. IBM may have pushed up their PR due to the news, like they did with the High-K release. If you don’t follow the register, you ought to.)
Slashdot discussion
IBM has confirmed that POWER6 has the same pipeline depth and roughly the same execution unit configuration as the POWER5

Revenue in the Lotus division grew 30 percent during the latest quarter compared with the final quarter of 2005, IBM reported last week. The company will demonstrate the service at its annual Lotusphere customer conference this week in Orlando, Fla.”

That’s what Ed Brill was referring to when he said he was very proud of the results his division posted this quarter. That really is wow kind of growth. It’s easy to dismiss IBM’s software because services is the major division at the company, but Lotus customers are almost apple-like in their devotion to it. And with Lotus Connections, IBM has signalled they will drive and inovate in the space. And the slick new look and feel of Lutus 8 due mid-2007 lays to rest much of the critcism that Lotus hasn’t kept up with the times. 5 years ago, customers were fearful that IBM was letting Lotus die a slow death, but IBM is in this space to win. Go big blue.

Also, only somewhat related, I was thinking that this should be added to my last post but I decided to add it here instead. Standard & Poor’s Corp. analyst Richard Stice said some people saw problems with the company’s hardware sales results. (IBM’s hardware-focused Systems and Technology group’s revenue only gained 3% year over year to $7.1 billion.) “There’s also some general profit-taking or negative sentiment associated with the technology sector at the moment,” Stice says. Noting things like IBM’s leading share position in a number of key sectors, Stice kept a strong buy opinion on the shares anyway. Stice wasn’t the only one to see more positives than negatives for Big Blue.

“While we acknowledge that there is no near-term catalyst for IBM shares, we are encouraged by multiple signs in the 4Q06 financial results of improvement in the struggling services business and strong momentum in software,” Citigroup analyst Richard Gardner said in a research note Jan. 19. (Citigroup has done business in IBM within the past five years, but Gardner certifies that the views in the report are not affected by compensation.)

IBM has been in the news this week for announcing a breakthrough in High-k Metal Gate Technology. Though IBM was the lead, they partnered with AMD and others to share risk and expense. Intel singularly announced a similar breakthrough the same week. I’ve provided what I believe is a sufficiently detailed summary of the breakthrough (I’ve been summarily disappointed by every article I’ve read). If you’re interested in the physics or math of it all, Wikipedia is, as always, fantastic.

Background:

Moore’s law for processor speeds is in jeopardy because of limitations imposed by the use of Silicone Dioxide to insulate circuitry in microchips. One problem with SiO2 is that 5 atoms thick is as thin as they can get it, imposing a hard limit on how small the chips can get. The second problem is that at that width, noticeable current begins to seep out (technically “leakage currents due to tunneling” results), lots of heat builds up and lots of power is consumed. It has proven very difficult to find a suitable replacement material.

The Solution:

IBM et al found that certain hafnium http://en.wikipedia.org/wiki/Hafnium alloys (probably Hafnium silicate - HfSiON) can be used as a more effective insulator than silicon dioxide and is planning to use the element to produce faster and more energy efficient chips, by allowing circuitry scaling to be reduced to less than 45 nanometers. Integrating high-k / metal gates will address the power consumption issue; a major barrier to scaling chips and continuing with Moore’s Law. It is thought that with this weeks announcements, Moore’s law has been extended through the next decade.

Another huge benefit to the Hafnium Solution is that it can be implemented without requiring major tooling or process changes in manufacturing. It seems that most of the alternative gate dielectric materials were impractical for existing manufacturing facilities resulting in potentially Billions of Dollars to move to the next generation of chip circuitry.

On-Topic Tangent:
Dr. Rajarao “Raj” Jammy was the project lead on this. He has 50+ patents and is one of those guys that you only ever see working in a University somewhere or at IBM. Anyway, there’s a great interview with here at Reed Electronics following a conference on High-k dielectrics. A quote:

…(In) reference to SiO2 or silicon oxide or silicon oxynitride gates, those gates with polysilicon electrodes have stopped scaling. So once these dielectric stacks or gate stacks have stopped scaling, we really had no option left. But for performance enhancement, people continue to use new ideas, like mobility enhancement. And therefore, they continue the scaling that the industry is so used to. But in some point in time we have to get back to the dielectric and try to see how we can continue to scale the dielectric also – part of it for the improvement in the coupling that we achieve at the channel, and therefore drive more current; but also the key part of it is reducing the leakage that comes from the gate dielectric.

IBM announced on Saturday an agreement to acquire Softek Storage Solutions Corporation, a privately held company based in Vienna, Virginia. They have been an IBM partner for as long as they’ve been independent from Fujitsu (1996) and have good penetration in the Fortune 1000, claiming 60% of them as customers. Interestingly, Softek just bought Enigma Data Solutions, a mainframe software developer, in November.

Softek helps companies manage data, assisting with data uptime management, transferring data from one environment to another, as well as assisting with backup recovery plans for businesses. IBM has installed Softek’s flagship TDMF product “on thousands of services engagements worldwide” making it a decent idea for IBM to dip at both ends of that pond. Per Hoovers:

The company’s products are used to monitor, provision, and manage enterprise storage assets (including servers, mainframes, and networked storage devices). Softek also offers software for implementing storage area networks (SANs), monitoring service levels, and backing up and recovering data, as well as professional services such as consulting and data migration.

Financial terms were not disclosed and I haven’t caught and online whispers, but I would guess the acquisition to be in the . The acquisition is subject to regulatory approvals and other customary closing conditions. Upon completion of the agreement, Softek will become part of the Storage and Data Services business unit in IBM Global Technology Services.

IBM released earnings after the bell today and the news was largely good. Revenue rose 7.5% to $26.4 billion and profit was up $3.54 billion, or $2.31 a share, up from $3.19 billion last year. That was slightly above “expectations”, but press has been glowing of late (another great IBM article was in this month’s FastCompany (subscription required) , talking about how math can be sexy). It looks like IBM’s stock will dip for at least a few day (like I know what I’m talking about) but I really believe this is going to be a big year for the company.

The encouraging part of the release was that services division is growing much faster than it had been (maybe IBM wasn’t lying when they said there were some contracts in Q3 that were held up). The blemish on earnings was the hardware division, which, other than Bob Hoey’s fine advertisements, have been pretty neglected. That said, IBM will fare as their services division does. Great quarter IBM.

There’s a nice discussion over at slashdot spurred by the defection of Don Ferguson, the man behind WebSphere, to Microsoft. The discussion can be summed up by the following:

1. what kind of person leaves IBM to go to the evil empire?
2. IBM management is out of touch and IBM is a sinking ship (I know that you only hear the squeaky wheel, but the level of ire still concerns me - though I stand by the prediction that 2007 will be a great year for IBM)
3. Websphere is terrible and clunky
4. The last few versions of Websphere have been nice (though it’s a framework.. not really an app)

I thought these comments were noteworthy:

Aa pro-WebSphere Poster

Let me tell you two things:

1) Big companies needs big support. Who will guarantee their servers will be up’n'running 24×7? Who will pay the fines if a failure stops the big company from operating for, say, 3 hours? That’s the IBM’s market. IBM is big enough (and have people enough) to support this kind of
company.

2) In my experience as a Java developer, I can say WebSphere is one of the fastest application servers in the market. Even faster when running in real servers (not that cheap toys [serverpronto.com]). JBoss (opensource) is really good, but isn’t enough for some companies. The difference between JBoss and WebSphere is that JBoss is made for developers (it’s easy to install/configure) and WebSphere is made for performance. It’s not a trivial task to install/configure, but once configured, it is fast as hell

Do you need middleware

Well.. ..I am not going to shill for IBM, because really, I’ve worked with the hairy mess that is WebSphere, and it’s like everything from IBM - a lifestyle choice. You don’t just recommend it like you would Zope or FoR.

But in the end you buy software in this class for a few key reasons:

1. Ability to interface directly with many platforms. (see #2)

2. The ability to write software that runs on many platforms. And I don’t mean Linux or Windows when I say platforms, I mean like mainframe, mini, datacenter, server, etc.

3. The ability to write really big systems.

When I mean really big, I am saying, you know like supporting an e-commerce website with 80,000 http request per second. They are rare, but they are out there. Although the core of the product is IBM HTTP Server, which is a fork of Apache, the key is in the tuning.

Here is the test I recommend when people ask me about it: can you run a query against your live database to determine orders/transactions placed today?

If you can, than don’t worry about Websphere or middleware at all. You are fine. Your site or app is still “small” (not a pejorative).
If you can’t, than it means you probably have a big system. And maybe you need middleware.

Money Magazine has a v
ery positive article about IBM at CNN on Friday. This is the kind of
positive publicity I was wanting when I wrote my embittered (and
retrospectively ironical)”Why I hate HP”
post in August. IBM, in spite of sluggish services growth in Q3 seems
to be firing on all cylinders and people (journalists and analysts)
really seem to like their decision to acquire software firms like a
squirrel preparing for a long winter (I’m still on the fence). But it’s
fun seeing positive things about good old IBM. Enjoy it while you can
because press attitudes can change overnight and it’s way easier for a
lazy journalist (or blogger as the case may be) to criticize than it is
to thoughtfully praise. I really like where IBM is right now
fundamentally. They are doing a very nice job of delivering short term
while showing great vision with investments, research and staffing for
the long haul.

A few quotes from the article:

IBM went on a software buying binge last year and analysts
think this was a good move.

“They’ve made some key acquisitions in the last year, and
they’re not overspending on them,” said Christopher. “They’re buying
companies with established businesses that aren’t too big, but that can
grow.”

“Internationally, they’re extremely strong - they’re viewed
as the gold standard”.

And with IBM trading at a discount to all of its top competitors, you
could say the same about the stock.

Information Week has a nice article about the IBM OmniFind Yahoo Edition. It gives paints a favorable picture of the alliance from the perspective of both firms.

IBM is hoping that users of its free software will like what they see and want to add to more sophisticated tools that deliver more substantial revenue. “We actually believe that search is really only the start of the value proposition for enterprises,” says Andrews.

Yahoo, meanwhile, hopes to increase awareness of its services among businesspeople. “We haven’t been in the enterprise business space per se,” says Eckart Walther, vice president of product management for Yahoo! Search, “but our products are used in the enterprise.” He points to Yahoo Messenger, Yahoo Mail, Yahoo Small Business, and Yahoo HotJobs as examples.

IBM has continued fleshing out Tivoli with its fat wallet, purchasing Vallent Corporation, a privately held company that makes software allowing Cell Phone providers to granularly monitor its network quality for a purported $200 million. Consolidation has been happening fast in the OSS space and it looks like convergence is playing out such that there will be half a dozen or so major players in a year or two:

• Amdocs
• HP
• IBM
• Oracle
• Telcordia

Smaller, independent players such as Vallent and Intec ripe for being brought into the fold of a larger company. So far this year, Oracle has bought MetaSolv, Amdocs - Cramer, CA - Wily and Syndesis picked up CoManage.) And Telcordia, who I described above as a survivor, is rumored to be for sale and in talks with SAP.

Anyhow, with the merger, IBM becomes the leader in quality of service tools for the wireless side of the telecom industry. Vallent Corporation specializes in wireless network and service performance management to help mobile operators forestall network outages and gauge the real-time quality of service being experienced by customers connected to their networks. Network failures are costly from a lost revenue and customer dissatisfaction standpoint so the QoS tools Vallent sells are indespensible. Vallent was formed by the mergers of Watchmark, Comnitel and Metica and, as such, nearly every carrier in the world uses their products. The private company is thought to have annual revenues of aroung $70 million with a sales price of $200 million.

Most notably, the Vallent acquisition will enhance IBM’s partnership network by adding many deep relationships with big names around the world the world. Per telecom industry pub Light Reading:

The move will bolster IBM’s position as one of the OSS industry’s biggest players, as Vallent, which has more than 200 mobile operator customers, will give the IT giant a wealth of wireless carrier contracts and wireless service assurance capabilities.

This seems like a great acquisition (Don’t they all? What’s a couple hundred million to us?). As they have demonstrated with contracts with Telstra Corporation and Bharti Airtel, IBM seems to be one of the few players with whom companies feel comfortable inking major dollar, many year deals. IBM gets a major arrow in its quiver with the QoS tools and the added (primary?) benefit of a huge rolodex ripe for the upsell.