New SGI Mostly Sticks to Techno Roadmap

It's been just a little over two months since Rackable Systems acquired SGI and merged the two product lines. From all appearances, it has been a fairly smooth transition. A lot of that has to do with the fact that the new company has left both sets of product offerings intact, although the future "Ultraviolet" system will change that somewhat. I'll get to that plot line in a moment.

According to Geoffrey Noer, senior director of product marketing at SGI, they "have not discontinued any of the products and have no immediate plans to do so." The product continuity is a reflection of the company's intent to combine Rackable's established customers in the Internet/cloud computing space with SGI's strength in the high performance computing market. Even before the merger deal, Rackable was looking at the HPC technical and federal government markets as a way to grow the business. Now that they have the SGI Altix products in tow, they get to those markets by default.

[...]

the next generation shared memory system will be based on the new Ultraviolet architecture. That design will use Intel Nehalem EX chips along with the next generation NUMAlink interconnect. Presumably this means the future "Tukwila" quad-core Itanium chips will never find a home at SGI. Although the 4700 line will continue to be offered for some period of time, the idea is to eventually migrate all the current users to the new architecture. "The intention is that Ultraviolet is the future of the shared memory systems line," says Noer.

According to him, more information about Ultraviolet will be publicly revealed later this year (although SGI customers have been privy to some of the details for awhile now). I would expect to see the introduction of the first new machines coincident with Intel's release of the Nehalem EX processors, which probably means early 2010.

RIP SGI Itanium.

SGI/PSC Ultraviolet Deal Dead?

There's a rumor running around that the new SGI is breaking the old SGI's deal with the Pittsburgh Supercomputing Center. This would kill the petaflop Ultraviolet system that PSC was buying. The new SGI doesn't have to honor the contract and apparently has other custoemrs willing to pay more for the system than PSC. With a limited manufacturing capability and a customer willing to pay more, SGI is taking the highest bidder for its equipment. This leaves the NSF and PSC in a lurch. It may buy SGI some money now, which it needs, but it could also leave the HPC division in trouble in the future: NSF being one of the biggest funders of HPC systems in this country other than DOE.

We'll see...

Rackle Does a Tera

So, SGI was bought for a pittance by Rackable Systems...and then changes their name to SGI. Huh. What Tera did with Cray seems to have set a precedent. Now will SGI subsume Rackable like Cray did Tera? Or should we say, "Do a China?"

Oracle Buys Sun

Oracle Corp. snapped up computer server and software maker Sun Microsystems Inc. for $7.4 billion Monday, pouncing on an opportunity that opened up after rival IBM Corp. abandoned an earlier bid to buy one of Silicon Valley's best known — and most troubled — companies.

The deal will end Sun's 27-year history as Silicon Valley's brash independent and give Oracle ownership of the Java programming language, which runs on more than 1 billion devices around the world. Oracle also will take charge of the Solaris operating system, which already has been a platform for much of Oracle's products.

It's far from Oracle's biggest acquisition during a four-year shopping spree that has cost more than $40 billion, but it may be the boldest.

Oracle, a Redwood Shores, Calif.-based business software maker, will be branching more into storage and computer hardware as it accelerates its attempts to become a one-stop technology shop for more than 300,000 corporate, government and academic customers.

"With the acquisition of Sun, Oracle is now able to make all of the pieces of the technology stack fit together and work well," Oracle Chief Executive Larry Ellison said during a Monday conference call.

Wow. Here's a contrast: Sun bought for $7.4 billion. SGI for $25 million. The mighty systems companies of the past are all falling. Ellison in charge of Sun...*shudders*

RIP SGI

Rackable Systems, Inc. (NASDAQ:RACK), a leading provider of servers and storage products for medium to large-scale data centers, today announced its agreement to acquire substantially all the assets of Silicon Graphics, Inc. (SGI) (NASDAQ: SGIC) for approximately $25 million in cash, subject to adjustment in certain circumstances, plus the assumption of certain liabilities associated with the acquired assets.

$25 million is diddly squat.

End of an era. There's nada distinctive for SGI to reemerge from Rackable (a la Cray).

This has also been a looooong time coming.

Yet More SGI Trouble

If there is one company out there that embodies "former glory", it just has to be Silicon Graphics Incorporated. Once one of the most well-respected high-performance computer and software makers of the world, it is now a mere shadow of its former self. The most recent set of financial results from the company do not bode well for the already troubled company.

SGI made most of its name back when it was the primary supplier of graphics hardware for the Hollywood visual effects industry. The company supplied this demanding industry with heavy duty graphics workstations, as well as high-performance visualisation systems. SGI's penetration in this market became clear when the brand started to appear in high-profile movies, with the most memorable being the "This is a UNIX system, I know this." case of Jurassic Park.

However, with the rise of powerful graphics hardware in cheap, ordinary computers, as well as the emergence of Linux and the advent of the Windows/Linux/Mac OS X port of Maya, led to the company slowly being pushed out of the visual effects industry, being replaced by the aforementioned cheaper options. Instead, the company focussed on government work, the scientific market and high-end servers and supercomputers.

There were several other confounding factors that have helped in SGI's decline. The company prematurely announced a switch from MIPS to Itanium in 1998, after which it quickly became clear that Intel could not deliver the Itanium on time; to make matters worse, Itanium would not reach the speeds as promised by Intel. This left SGI with a set of outdated MIPS-based products, and they quickly had to release new models with faster processors to remain competitive. The switch to Itanium was continued later on, and at one point, the SGI Altix formed the base of the Columbia supercomputer, which was built out of 20 Altix 3000 systems with 512 Itanium processors each. At the time, this was by far the faster supercomputer in the world.

In 2006 the company announced the end of its legendary MIPS product line, as well as the end of the IRIX operating system, preferring Linux instead, all as part of a plan to bring the company back to positive figures. Not long after, the company emerged from bankruptcy, and things started to look up again. The company even re-entered the famed visualisation market with the SGI Virtu running Linux or Windows Compute Cluster.

However, due to the economic downturn, the company has made a few steps back during the last quarter. I won't bore you with the figures, but suffice to say it's not going well. The company is hoping to get government support, rooting for a slice of the stimulus pie president Obama and the US Congress are working on right now.

If they need some of the stimulus package to survive, they're probably toast. Then again, we've been saying that for a loooooooooooooooong time. From our POV, they're the undead HPC maker.

They. Just. Won't. Die.

OTOH, some of the things they have on their HPC roadmap - if they can survive that long - are damned fascinating if they can pull them off.

SGI's Cell Phone Supercomputer Node

With 10,000 Processor Cores, Silicon Graphics Molecule 'Concept Computer' Uses Half the Power and Just 1.4 Percent of the Space of Comparable PC Cluster

Silicon Graphics, Inc. today offered a glimpse of the potential future of dense, power-efficient computing with the Silicon Graphics® Molecule™ concept computer.

Like a futuristic concept car points to potential innovations in transportation, Silicon Graphics Molecule is a concept computer that illustrates how the latest low-watt, multi-core consumer electronics technology, such as the Intel® Atom™ processor, can be combined with breakthrough Silicon Graphics® Kelvin™ cooling technology to pack more than 10,000 cores into a single rack.

Engineers at Silicon Graphics research labs developed the system to show how consumer electronics technologies and emerging marketplace trends might someday be applied to overcome the limits of today's high-throughput clusters. The Silicon Graphics Molecule concept computer balances processor speed, sustained memory bandwidth, and power consumption. The system has been shown to deliver sustained results on scientific and business problems from seismic processing to rendering and distributed searching.

Features of the Silicon Graphics Molecule concept computer include:

* High concurrency with 20,000 threads of execution — 40 times more than a single rack x86 cluster system
* High throughput with 15TB/sec of memory bandwidth per rack — over 20 times faster than a single rack x86 cluster system
* Greater balance with up to three times the memory bandwidth/OPS compared to current x86 CPUs
* High performance with approximately 3.5 times the computational performance per rack
* Greener with low-watt consumer CPUs and low-power memory that deliver 7 times better memory bandwidth/watt
* Innovative Silicon Graphics Kelvin cooling technology, which enables denser packaging by stabilizing thermal operations in densely configured solutions
* Operating environment flexibility, capable of running industry-standard Linux® implementations, with Microsoft® Windows® variants on some configurations

If someday brought to market, a single-rack system based on the Silicon Graphics Molecule concept computer would offer the computing power and memory bandwidth of more than 750 high-end PCs, yet it would consume less than half the power and less than 1.4 percent of the physical space.

Now that's closer to a HPC desktop. It would have to have a benchmark of 1.2 GFlops/core to make it onto the list. However, first off, it's only a concept computer. Secondly, this will make John at work very happy. Third: reliability. How often do I replace a CPU? Is it a 1% failure rate? What's the MBTF? Better be more than a billion hours: one million hours means there's a CPU failing every 4 days per node. I'd rather not be on the other side of the phone when a researcher calls irate that his 10 million core job had to get restarted again. Fourth, this still eats 40 kilowatts per node...minimum. hmmm. Big power plug. Better consumption per flop, but...

Okay. Mental exercise. I want 1024 nodes here (sorta a magic HPC number). I am going to assume a usefulflop rate of 500 Mflops/core and a gigabyte of memory per core (we really like 2, but...) That means this system would have 10,240,000 cores and more than 10 Petabytes of memory. (o.O) The useful sustained computational rate of 5 petaflops. However, we have a power requirement in excess of 41 megawatts just for the CPUs[1]. Not counting memory, chipset, spinning disk or interconnect. Ouch. MBTF of 1 billion hours? Means I am yanking a failed CPU every 4 days. Wait is the MBTF even close to that?

Looks interesting, but, again, just a concept uber'puter node. We'll see how this meshes with SGIs HPC plans.


1. *mutters* Heat Death of the Singularity.