The IT Business

There is an article at tech crunch about the dislike of Intel in regard of a 12 inch notebook with Atom proc. I think it´s not only the loss of sales for bigger mobile procs. The problem is more wide-ranging.

The cheap x86 procs for servers are largely subsidised by the desktop sales. It´s almost the same die, thus the development costs are shared, the manufacturing costs are shared. It´s classic economy of scale ... it´s the reason, why a Intel Xeon is cheaper to produce than a RISC CPU. Without the desktop sales you have to increase the price of your server CPU or reduce the margins (server CPUs are an high-margin business)

Now let´s assume you have a different architecture with different design targets, thus it isn´t possible to share the development costs. This can severely disrupt your business modell. You can´t share the costs of your server procs with all the desktop sales. As i wrote before: The current dominance of x86 servers are an result of the decision of a finnish computer science student and many gamers, because the fact that the dominant gaming os runs on it x86 only. Thus Intel has a vast interest in not selling Atom in all possible markets. I perfectly understand Intel that they don´t like 12-inch netbooks. 10 inch displays are to small for many people, but 12 inch is a different story.

This the upcoming netbooks and nettops have may have an interesting effect to unrelated markets by changes in the cost models due to decreasing sales.

Recent layoffs (besides the Sun RIF):
Lenovo - 2500
IBM - 16000 (rumored)
Unisys - 1300
Microsoft - 15000 (rumored)
Logitech - 500
SGI - 225 (15%)
EMC - 2,400 (7%)
Dell - 1,900
HP - 24,600 (announced in September 2008)

Tough time at the moment ...

From the release notes of FreeBSD 7.1-RELEASE:

>The DTrace, a comprehensive dynamic tracing framework and dtrace(1) userland utility have been imported from OpenSolaris. DTrace provides a powerful infrastructure to permit administrators, developers, and service personnel to concisely answer arbitrary questions about the behavior of the operating system and user programs.

(via Tim Bray)

An tweet on twitter lead me to some thoughts. The question in this tweet was "What will be after Linux?" At first this seems as a question of a Solaris fanboy and obviously you would expect that i would opt for Solaris in this article. I have my opinion but i won´t discuss it here. I won´t even give a hint because honestly spoken, i have no idea. It´s relatively easy to look 6 months in the future, it´s much harder to do this for 12 month and it´s outright impossible to do this for a point in time in a 5 years interval (okay, when you make your predictions in a generic way, it´s easy).

When we look into the history of computing, we see many technologies with a vast amount of market share having problems later on. Mainframes are such a example. Yeah ... Sun is such an example ... nobody expected in 1998 that Solaris would have the role of something else than the market leader in web services 10 years later. Many thought of Longhorn as the next big thing killing all other OSes. We´ve got Vista. In ten years we will think about Vista in the same way as Windows 98SE. Otherway round we see many systems which were almost dead with a bright future today. Think about MacOS 9 and MacOS X. Apple was pretty much a few years ago and today the market share on notebooks looks as near to 50%, at least when you walk through the train between Hamburg and Berlin. Other way round: 5 or 6 years ago the Palm Pilot was the unconquered leader of it´s market ... today ... the complete market was sucked up by cellular phone vendors and the vendor of a mp3-player on steroids. Apache vs. lighttpd or nginx. Sendmail vs. Postfix. Ingres vs. Oracle. CP/M vs. DOS. dBase vs. Access et al.

This example should show us one thing: There is no thing in IT that keeps it´s lead forever. Comebacks are possible. Complete removal from the market place is possible. It´s foolish to assume that any piece technology is excempted from this rule.

So Linux will encounter the same lifecycle. Think this is impossible? What would happen, if Linus Torvalds decides to take his midlife crisis and starts to do research on real penguins in a polar station. What would happen, if one of the large proponents (Red Hat, SuSE, Canonical) would collapse under the weight of this or the next recession? That´s not a linux-only thing ... just think about the "Steven Jobs is ill/dead/the new iGod" rumours.

In my opinion Linux is already on the downward path. There wasn´t an innovation in Linux that gave me the thought "Wow, that´s cool" for a long, long time (And to be honest: For many features i thought as cool features in Linux in my early UNIX years i´ve learned later on that they were implemented somewhere else (IRIX, Solaris, some old BSD et al) before) ... but your perspective may vary as your milage. This is nature of opinions. Dbase was moved out of market by other competitors making a faster transition to Windows. lighttpd is used by admins, who think that Apache httpd has grown to far to result into a stable and efficient webserver.

Furthermore: The user communities in the open source world are more fluid. The Microsoft ecosphere is a little bit different. Microsoft is only able to survive it´s constant underdelivery in regard of their operating system because of their applications. You have to use Windows, if you want to use MS Word or MS Excel. But as i wrote a while ago: In Open Source the binary of an application is just a Makefile away. Thus there is no application-enforced vendor lock-in. When people don´t like their old OS (out of whatever reason) anymore or just want features of another OS, it´s just an rsync to the new away. Commercial IT would take a little bit longer because of already existing runbooks and processes. But Unixes aren´t that far away from each other to make this impossible.

Despite what many want to think: Linux isn´t immune to this. So the interesting question is ... what will be after Linux? LinuxNG, OpenSolaris, BSD, Windunix? That´s an interesting question and i have no answer to it. Just an opinion. It the next Linux a already known operating system. Or a completly new operating environment?

What would be the path of Linux after such an downward path ... it´s a community development. The next interesting question is: Are highly dispersed development communities capable to restart a franchise? Like other restarts: MacOS 9 to X (non-opensouce), like Solaris 9 to 10 (open-sourcE). Both restarts were triggered from large companies with a large interest in the restart and deep pockets to pay developers. Or would the community just move to another prefered development platform moving Linux into a niche, like the BSDs today?

After writing this article i had an additional answer to the question of a customer: "Why does Sun still develop Solaris instead of supporting Linux the development" aside from all the technological and commercial reasons: Because Linux needs Solaris. Without a strong, innovative competitor the downward spiral (as the innovation would be just limited to supporting new hardware) would just go faster and this would open an opportunity to other systems. The otherway round Solaris needs Linux ... Solaris 10 would look different without the large impact of Linux in the market and the balance of power between Windows and. Unix would look differently. It´s the same with MacOS. The Apple developers were in need of an improving Windows to restart the MacOS franchise with X. Without it Apple would be a large part of computer history, but not of contemporary IT.

So, we should think about the time after Linux as well we should think how Linux gain strength in such a phase. But as i wrote at start, it´s hard to think about it, when predictions are such a hard business.

Cisco seems to look for new markets - Cisco planning significant data center assault:

Perhaps the most important example of that will be a new Cisco blade server system expected next year. This will take the company into the data center compute space, right up against longtime stalwarts — and up to now, Cisco partners — IBM and HP.

This matches to rumours and hints i´ve got in in the last few weeks. But i don´t believe that Cisco will play a large role in this business. There are already several players in this game ... IBM,HP,Dell,FSC and Sun. And more important: People think of Cisco as the supplier of their networking equipment ... not of their servers ... and Cisco hasn´t a long track record as a supplier of servers. Cisco needs a real differentiator ... i´m curious what this differentiator might be ... but at the moment i´m not convinced about the success.

And there is another problem: Network admin departments, storage admin departments and server admin departments have seperated supplier spheres in many companies. And Cisco is a supplier for the network admin departments and they rarely connection to the server departments. And: Did you know Sun sold an hardware load balancer for a while? No? Cisco will have a similar problem ...

There is an interesting report at xtremesystems.org with some photos about overclocking an AMD Phenom II to almost 6.3 GHz. Okay ... you need liquid nitrogen to do so, so it isn´t really relevant for most people, but 4,13 GHz with air cooling is interesting as well. An observer writes:

- We were able to run Crysis @ 6,1 GHz (2 loops CPU-test)
- Max CPU-Z screenshot and validation we got with 4 cores was 6280 MHz (no dirty tricks)
- With air cooling (Cooler Master Z600) highest we got Crysis stable was @ 4,13 GHz (1,58 V)

Looks like the AMD Phenom II has some potential in regard of it´s frequency.

PS: By the way ... somebody out there with a spare UltraSPARC T2 system? I would purchase some liquid nitrogen

There is some discussion about many-cores hitting a "memory wall". Robin Harris writes about it in his article Many-cores hit the memory wall. But i think most of this article think not for enough. Who says, that you just can have many cores. You can have many-memorycontrollers as well. The future solution for memory is perhaps not the fastest bus. It´s the bus that transport as much data as possible with the fewest possible amount of socket pins. The real wall may be the pin budget of modern processors as the number of memory controllers is limited by the possible connections to the outside world.

I don´t think there is a wall at 16 cores in processor design because of memory bandwith mandated by some basic principles, there is just a economic limit of manufacturing costs that prevents you from simply increasing the pin count and thus the numbers memory channels.

And so the problem of such articles like Jon Stokes "Analysis: more than 16 cores may well be pointless" is just the point, that they think in x86 technology and x86 economics. And even there he might be only correct when you just multiply the cores. There is a reason for connecting three channels of memory to an Intel Nehalem.

At the end the memory-wall discussion is somehow a validation of a Sun technology. You have to do something different. The idea of putting 4 memory controller on a T1 for example. The threading concept in the UltraSPARC T1 and T2 was introduced for a reason, too . The basic idea of this concept was "Live with the limited memory bandwidth and with the latency of memory and make the best out of it". The T1/T2 just switches to another thread in case of a memory stalled thread.

The TLB bug almost ruined the AMD Phenom franchise. But it looks like if Intel has a similar problem with the Nehalem. According to Fudzilla the Nehalem is haunted by a TLB bug too and refers to the errata document of this processor. It looks like history repeating.

At page 16 of the the Intel Core i7 Processor Extreme Edition Series and Intel® Core™ i7 Processor they write:

AAJ1. MCi_Status Overflow Bit May Be Incorrectly Set on a Single Instance
of a DTLB Error
Problem:A single Data Translation Look Aside Buffer (DTLB) error can incorrectly set the Overflow (bit [62]) in the MCi_Status register. A DTLB error is indicated by MCA error code (bits [15:0]) appearing as binary value, 000x 0000 0001 0100, in the MCi_Status register.
Implication: Due to this erratum, the Overflow bit in the MCi_Status register may not be an accurate
indication of multiple occurrences of DTLB errors. There is no other impact to normal
processor functionality.

But in the clarification for errata AAJ1 on page 37 of the same document they state:

In rare instances, improper TLB invalidation may result in unpredictable system behavior, such as system hangs or incorrect data. Developers of operating systems should take this documentation into account when designing TLB invalidation algorithms. For the processors affected, Intel has provided a recommended update to system and BIOS vendors to incorporate into their BIOS to resolve this issue.

Doesn´t sound like There is no other impact to normal processor functionality. I don´t think, that this bug has the scale of the AMB TLB bug, but i think customers are a little bit sensitive today when they hear TLB and bug in the same sentence.

Update: AAJ42 on Page 27 is interesting, too:

Under certain conditions when C6 and two logical processors on the same core are enabled on a processor, an instruction fetch occurring after a logical processor exits from C6 may incorrectly use the translation lookaside buffer (TLB) address mapping belonging to the other logical processor in the processor core.

Funny discussion about boot-up times of pSeries systems:

Hey, I manage 7 p595 squadrons and they are all different. For the most part, the startup time depends on the amount of I/O you have installed. Two of mine are 3 frame p595's with 64 hba, 64 nic, about 20 internal HDD, dual power systems and I think 10 d20 I/O drawers. These servers take about 90 minutes just to initialise the hardware, not including starting any LPAR's. So tell your friend that this is normal and take a good book and a really large coffee with him each time he needs a restart

Hmm, it will be interesting to see where the acquisition of Transitive by IBM will leave Apple (Rosetta is a Transitive technology) or SGI (their ability to execute Irix/MIPS code on their Linux based systems). At least the last system can be viewed as a competition to IBM.

Well, as Transitive delivers one of the foundations of IBMs PowerVM technology, i expected, that they would acquire Transitive at some point in time. Today IBM announced such a move according to this press release: IBM Announces Plans to Acquire Transitive.

The Transitive stuff allows IBM to run Linux x86 code in a Power LPAR. Okay, i don´t really know if it´s sensible idea to use an expensive Power system to virtualize a cheap x86 system. But they need something like that as Linux on Power doesn´t really have a large support by ISVs.

Just did an short look on the new Top500 list:

• Wow ...A Constellation is an own architecture type in the Top 500 list
• The are just 9 Itanium systems left in the list.
• The total sum of Gigaflop per second increased from 11705491,49 on 2415203 to 16958600,19 on 3121579 procs. Thus the yield per CPU increased from 4,85 to 5,43 Gigaflops per proc.
• By the way: Ten year ago, the list sumed up to 29367,61 gigaflops. That´s an performance increas of over 577 times in 10 years.
• When compare the actual increases with Moores law in it´s first derivation (Moore talked about transistor count, not performance), we see some interesting developments.Moores law suggests in it´s first derivation an increase in perfomance very 18 months. Over the 10,5 years period Moore´s law suggested an increase by the factor of 128. So the performance increase of the top 500 clusters was higher than Moores law suggest. The Top500 lists suggests a doubled performance in a little over 12 month instead on the long run.
• But this doubling is just correct when you look at the whole cluster. Let´s look at the CPU level and the list of 06/2007. In this list every CPU yielded 4.05 Gigaflop. The number of 06/2008 ist just increased by the factor of 1.3 in 18 month.
• Back to cluster level: In 06/2007 the TOP500 yielded 6974873 Gigaflop/second. In the following 18 months the the list showed an increase of 2.43 (as already mentioned, the total computing power of the Nov 2008 list was 16958600 GF/s)

Reuters writes in Fujitsu to buy Siemens stake in PC joint venture:

Japanese electronics conglomerate Fujitsu Ltd will buy Siemens AG's 50 percent stake in their computer joint venture for 450 million euros ($567 million), aiming to boost its presence

They just talk about the PC business in this article, so there is no new information about the FSC server business.

There is something really strange in the numbers for the EMC Symmetrix DMX4. A colleague gave me this hint. Just look on the specification sheet. They state on page 2 that this system can house up to 2400 disk (albeit they restrict that in the footnote by writing that more than 1920 drives are just available on RPQ). But that´s not the point.

Just look at the numbers at system capacity on page 3: They state on this whitepaper, that you can plug 1920 1 TB hard disks into the system. Okay .... let´s take the example for RAID 5 with 7+1. You could cut 240 of this RAIDs out of 1920 disks (1920/8). 240 multiplied with 7 Terabyte is 1680 Terabyte. But the maximum system capacity is stated by EMC with 585.91 Terabyte. WTF ?!?!? Where is the missing Petabyte? The first approx. 670 1 TB disks in your DMX4 gives you capacity, the other approx 1250 disks gives just heat dissipation and some nice blinking lights.

I don´t really think, that there is an error in my calculations . Why? When you calculate this for 73 GB you get 122 Terabyte ... almost the system capacity number stated on the specification sheet(you have to deduct some of the storage for internal usage, so it´s slightly less).

I smell a architectural weakness here. Something like "640 Kilobyte is enough for everyone" decision ... just on a larger scale

The third quarter of HP is the one we call the fourth quarter at Sun. So it covers the time from April to June. There is one interesting chart in the earnings presentation of HP - the revenue for Enterprise Servers ans Storage :

The next earnings announcement will be an interesting one. In the next earning announcement they can´t say that they grew year over year in this area. Nevertheless, there is an clear downward trend in this chart. Perhaps this is the reason why they bought EDS. They need that business to cover the decreasing revenue from their server branch.