Entries filed under “InsideTrack”

News from the HPC community before anyone else has it.

Sun visualization offering gone?

Randall at VizWorld had a post last week at his site about the disappearance of (most) of the data on the Sun vis page.

Sun logoToday, however, a fellow twitterer noticed that the Sun Visualization Webpage is almost entirely gone.  I’m completely unable to find it under their products lists, and going directly to the old page (still available under search), yields just…a Banner, but no text.  That “Get it” button goes to a form to ask for details, but nothing more.

No update since the post, but a commenter did note that Linda Fellingham left Sun in January of this year (she headed up vis at Sun).

Also posted in Visualization | Leave a comment

Inside Track: SiCortex rumored to be closing its doors [CONFIRMED]

Word on the street is that SiCortex will announce soon, perhaps today, that they are going out of business. No word on what the problem is; I know that they were working on another round of financing, and this is a tough environment to be looking for addititional financing. I’ll try to contact the company and find out for sure.

[UPDATE]: SiCortex has begun contacting people today letting them know that the third round of financing did not materialize in time for SiCortex to continue financing. Without the cash to operate they are trying now to sell the company, with a target sale by the end of June. As of today the company has suspended all sales activity.

Also posted in Business of HPC, Deadpool | 15 Comments

InsideTrack: deposits, stalking horses, Bull(s), and parachutes in the SGI Rackable deal

One of the legal documents from the proceeding today in bankruptcy court is making its way out to employees. Rackable CEO Mark Barrenchea filed an affidavit in conjunction with the ruling affirming that Rackable is a “good faith purchaser”.

Among the interesting parts of this statement, use of the term “stalking horse”

Rackable logoAs more fully described in the Sale Motion and the Agreements, pursuant to 11 U.S.C. §363, Rackable is the Stalking Horse Bidder for a sale of substantially all of the assets of the Debtors, free and clear of all Encumbrances (except for Permitted Encumbrances). …

A deposit. Like buying a house I guess.

Rackable expects to fund its proposed acquisition of substantially all the assets of the Debtors using available reserves of cash, cash equivalents and long-term and short-term investments. Rackable has deposited $1,000,000 in good funds in escrow in support of its bid, in accordance with the Agreements.

Regarding who remains, who doesn’t, and who might have a parachute

Among other things, the Agreements provide for Rackable to enter into certain post Closing arrangements with the Debtors’ CEO, Robert Ewald, and ten other members of the Debtors’ management team (the “Executives”), as part of Rackable’s bid. In addition, on or before May 1, 2009, Rackable has agreed to provide the Debtors with a list of Debtors’ Executives that Rackable proposes to retain on a long-term basis, as well as a list of Debtors’ Executives that Rackable proposes to retain on a transitional basis. Those Executives not retained on a long-term or transitional basis by Rackable will be terminated. The Agreements further provide that prior to the Closing Date, Rackable shall provide the Debtors with a list of Specified Employees to whom Rackable intends to offer employment. Rackable has no obligation under the Agreements to hire, employ, or provide transitional arrangements to any employee of Debtors or any member of Debtors’ management team other than as specified in the above-referenced lists. Rackable has not yet provided these lists to the Debtors. These matters are more fully described in the Sale Motion and the Agreements.

Evidently at least part of our speculation was right: Bull was involved at some point, and so was a group of Secured Lenders

In accordance with and after entry of the Bid Procedures Order and consent provided by Seller pursuant thereto, Rackable communicated with Bull S.A. (“Bull”) and certain of the Secured Lenders regarding the Transaction, the Auction, and the assets of Debtors and their subsidiaries. The parties explored various scenarios which might provide an increased purchase price for the Sale of assets of the Debtors. As to Bull, the communications ended prior to the Auction, were inconclusive and Rackable did not reach, and does not have at this time, any agreement, arrangement, or understanding with Bull regarding the Transaction, the Auction, or the assets of Debtors and their subsidiaries. Rackable was not constrained in any way and did not constrain in any way Bull or any other parties as to their ability to bid at the Auction, either independently or with any bidding partner or partners. As to the Secured Lenders, the communications prior to the Auction were very brief and inconclusive, and Rackable did not reach any agreement, arrangement, or understanding with the Secured Lenders regarding the Transaction, the Auction, or the assets of Debtors and their subsidiaries. There were no further communications with the Secured Lenders prior to the Auction.


Also posted in Business of HPC, Featured Stories | 1 Comment

InsideTrack: SGI notifying customers that the Rackable acquisition has been confirmed

SGI has begun notifying customers that the judge in its bankruptcy case approved the sale of SGI to Rackable during the hearing today in district court. The deal and the remaining paperwork are set to complete on May 11.

Some terms have been disclosed in an affidavit provided to employees earlier today, which I excerpt here. Neither company has issued a release. Update: 05012009 The official releases are linked here.

Also posted in Business of HPC, Featured Stories | Leave a comment

InsideTrack: Rumors that two bidders survived the SGI process

Industry sources say that only two bidders made it through the diligence process willing to submit a bid: Rackable, and a group of creditors, and that Rackable has the winning bid at this point. We posted at the beginning of the week about rumors of who was digging around. Evidently they didn’t come out happy with what they found.

No word of who was in that creditor group, and that is likely to not change. The debt holders filed a motion to seal the bids, making them invisible to the general public, and the judge partially concurred, making the names of the bidders available to Rackable but not to the public. This last bit is sole sourced, and I’m looking for corroboration. If you can corroborate, or if you can point me to a corroborating document on the interwebs, send me a note or leave a comment.

Also posted in Business of HPC | 1 Comment

InsideTrack: rumors of other bidders for SGI [Updated]

I’ll show you mine if you show me yours…here’s what I’m hearing about companies sniffing around SGI and looking at the potential of bidding against Rackable.

Cisco: given their Unified Computing Platform (codenamed California, covered here), this move could make sense as a lever into HPC, but its not clear that Cisco wants to get into HPC. UCS is more strongly aligned with the enterprise (Cisco/Sun might have worked). Of the two pure play HPC companies SGI has the strongest commercial high end computing customer base, and this could help Cisco cross-market its UCS and other solutions. Cisco has a market cap over $100B, and at the end of last fiscal year had just over $5B in cash and equivalents, and they are profitable. If they want in they could easily outbid Rackable.

HP: Although HP and IBM dominate the Top500, HP hasn’t had a strong presence in the top 10 for a while. SGI’s R&D in HPC could reinvigorate their offering. HP has a market cap of just about $85B right now, with $10B of cash and equivalents on hand. [Update: Readers have also pointed out that HP is the only other large-scale Itanium vendor in the US.  Buying SGI would be a consolidation of technology and talent surrounding the
platform.]

NEC: NEC has a strong presence overseas, but they are weak in the US. An SGI buy could give them leverage into the Americas, but would cause problems with SGI’s extensive federal business (which I assume would remain separate as they did when Bob Bishop — a native Australian — was head of the company). The company had cash on hand of about $2.5B US a year ago (the most recent info I could find). I think this one is unlikely, but I’ve heard it a couple times now.

Bull: Bull is a significant HPC vendor in Europe, but does very little elsewhere. In fact, France (52%) and the rest of Western Europe (27%) account for 79% of the company’s business, so you can see that this is a company in strong need of diversifying its customer base. Also, SGI’s R&D at the high end would be a nice complement to their cluster-based business. As of their last annual report they show only about $400M cash on hand, but they could clearly make a decision to invest in future growth and make a competitive bid for the company. I also think this one is unlikely, but I like it better than NEC.

I guess that HP is probably the most practical bidder, but it’s also the most boring. Bull is interesting and European ownership could work; Cisco seems possible but it would require knowledge about their plans in HPC that we don’t currently have. NEC seems unlikely given cultural differences, although the company’s research heritage could serve SGI’s future roadmap well.

Guess we’ll know on the 30th.

Also posted in Business of HPC | 5 Comments

InsideTrack: Penguin CEO on the SGI/Rackable deal

After the announcement yesterday we started rifling through our Rolodexes here at insideHPC HQ looking for industry comment. Predictably, most of the folks we got in touch with were at public companies and loathe to talk on the record (though in some cases I swear I was able to hear hands rubbing together in glee). But we did get in touch with Charlie Wuischpard, CEO of Penguin Computing, and in an insideHPC exclusive he was willing to share his take on the goings on:

Honestly, I’m not surprised at the deal since I knew Rackable was shopping but I think it’s going to be very problematic…not the least of which is because you now have two very different companies, with different focus and culture, who have both lost loads of money in the last 2 years. If you follow Rackable, they have still been touting the web business and the data center in a box (shipping container) concept. Very data center focused and no real IP or skill set in HPC (by their own admission and remember the failure of Rapidscale). I suppose you could plop Rackable’s management team and balance sheet into SGI and keep the focus on HPC but even under ideal circumstances, that will be difficult.

At the end of the day, as I’m sure you would expect me to say, we see this as a positive event for Penguin! It puts the SGI install base in play (and some of their remaining best people)…and I think Penguin should enter more short lists as a result.


Leave a comment

InsideTrack: stimulus money headed into HPC on the back of other programs

I was on the phone with Charlie Wuischpard, the CEO of Penguin Computing, today and got some interesting insight into possible HPC impact from President Obama’s stimulus spending.

From the various posts I’ve made and pointed to on the subject, you’ll remember that while there is nominal added funded for computing in the stimulus act, it doesn’t come across as anything like a priority.

I was heartened to hear from Charlie that he is seeing what he believes is the front wave of a new round of spending about to start in HPC as grant writers (in this case hunting for NIH money) approach his company about helping them figure out the computational piece of their scientific research proposals. Looking over the President’s priority list (things like manufacturing, healthcare, construction, and new energy) I certainly see many opportunities for large-scale computing to provide benefit as an enabler of these research priorities.

And in the end, this could actually be better for the HPC community: products built for the requirements of today are usually just better products than those built for the requirements of tomorrow, or those built to satisfy a research agenda around the product itself.

1 Comment

InsideTrack: SGI makes official layoff announcement, eliminates 15% of workforce and some senior management [UPDATED]

SGI confirmed layoff rumors tonight with the release of an official announcement

SGI logoThe company has refined its focus and is restructuring to adjust to changing customer and financial market realities. As part of its restructuring, the company will reduce its workforce by approximately 225 positions, or roughly 15 percent. The reductions will include several executive and senior-level positions.

Although the release doesn’t specify where the cuts happened, a source close to the company says that the cuts were primarily on the commercial side of the business, while the federal business was left mostly intact. Those that were let go left the company yesterday.

The executives who were dismissed are not named, and I assume we’ll have to wait for subsequent filings with the SEC or updates to the web site to figure out if anyone on the senior team is leaving.

[UPDATE: comparing today's exec bios page with earlier versions of that page insideHPC notes that Shahin Khan, VP of marketing, Barbara Stinnett, VP of Services, and Bill Trestrail, VP of APAC, are all missing. Presumably victims of the cuts.]

Also insideHPC has news that Dave Parry, Senior Vice President, Server Products and Operations, will leave the company in January. No word on whether this is part of the restructuring or whether Dave has just finally had enough turmoil for one career. This is bad news for SGI — Dave is a strong technical force in the company and an all around swell guy.

According to the announcement, the company will retain and focus its efforts on servers, visualization, and innovations like the Industrial Strength Linux Environment.

There isn’t a good time of the year to be out of work, but it seems especially poignant to lose a job at Christmas. Our thoughts and prayers are with those affected by SGI’s actions, and with those who now have to set to work moving the company forward.

Also posted in HPC | 1 Comment

InsideTrack: SGI to lay off workers [UPDATED]

While we don’t yet have details, a source sources inside SGI has confirmed for insideHPC that layoffs are expected at the company today. No word on size (see update below) and scope, or on which businesses will be affected.

Note that this follows layoffs of roughly 6,000 at Sun last month, and rumored (though unconfirmed) layoffs coming to IBM’s Deep Computing division.

We’ll keep these in our thoughts and prayers.

[UPDATE: Additional word from those in a position to know is that Bo sent out an email last night putting the size of the layoffs at between 10% and 20% of the workforce, and it looks like it might be Friday before we know which 20%. Details on the timing are still sketchy.]

Also posted in HPC | 4 Comments

InsideTrack: Rocky Mountain Supercomputing Center

Our faithful reader, Rich Hickey, delivered this great report on Butte’s Rocky Mountain Supercomputing Center, courtesy of his wife Roberta.  She was kind enough to venture out and attend the city council meeting whereat the center plans were discussed.

Here is the info I got from tonight…the IBM is being delivered December 27th and it is a water cooled power 7 or whatever the heck you said. One step down from a Blue Gene. The RMSC is a non profit organization that will precipitate the clients and schedule time on the system as admistration for the system. The state is funding the program long term and will own the computer. MERDI is leasing the rack space for the system in the Thornton building.
Funding is being provided by state and federal grants for the next 2 years, with the possibility of longer term funding.
The US Air Force already has expressed interest in building a paramilitary medical rescue simulation center here in Butte, using the supercomputer…and Rocky Mountain Laboratory has already signed contingency contracts to use the system. Both state schools will, of course, have access to it.
Several high profile people came out to support the project, including Dr Alex Philp, professor at MSU, Senator Jon Tester, Ray Rogers Vice President of the National Center for Health Care Informatics, Robert Rice of Neogents and of course, me.
Several of the county commissioners still have reservations about supporting the project financially in such economic times, however have pledged to look into all funding sources the county has access to in the next 10 days. A vote is scheduled at the next meeting next Wednesday.
Dr. Alex Philp said the funding will bridge the gap between the installation of the machine and the beginning of state funding in July 09. The funding will be used to provide operating capital and hire a center director. [Roberta Hickey]

Great scoop from the Hickey family!

Also posted in HPC | Leave a comment

InsideTrack: Bill Kramer Leaves NERSC for NCSA

This just in.  We just received word that Bill Kramer has been named the new Deputy Project Leader for the NSF-sponsored Blue Waters supercomputing project at the National Center for Supercomputing Applications [NCSA].  He is currently the General Manager at the National Energy Research Scientific Computing Center [NERSC].

Bill has been a long-time staple of the high performance computing industry on the west coast.  Prior to working at NERSC, he held positions down the street at NASA.  From 1986 to 1994, he was the Branch Chief of Computational Services at the Numerical Aerodynamic Simulation [NAS] Systems Divison at NASA Ames Research Center.

Good luck to Bill in his latest endeavors.

Also posted in HPC | Leave a comment

InsideTrack: Sun may be winning some business in a shakeup down under

A source close to insideHPC has given us a glimpse into some of what is going on in HPC in Australia with the Australian Bureau of Meteorology HPC acquisition. The acquisition covers several existing installations, with incumbents NEC and SGI. The acquisition calls for a machine each at APAC and CSIRO, with a hot spare at the BOM. APAC researchers seem to prefer a cluster and NOT a vector based system as their primary machine, whereas BOM/CSIRO are happy with a vector system.

Word from the InsideTrack initially was that IBM has a good shot since they often follow the UK Met organization, which recently went IBM. However the picture was unclear, as APAC evidently aren’t keen on IBM. It now appears that IBM has been ruled out, as has Cray. HP didn’t bid.

Which leaves NEC, SGI, and…. In fact the InsideTrack has been told that the announcement of the winner is delayed due to extended negotiations with Sun. We’ll definitely keep our eye on this one.

Leave a comment

InsideTrack: SC08 Cluster Challenge Technical Brief

As those of you attending SC08 this year begin to fill your calendars, we wanted to bring your attention back around to the second annual Cluster Challenge.  Last time, we introduced you to the teams and their respective vendor sponsors.  This time, we introduce you to the individual applications required during the performance qualification runs.  Brent Gorda graciously gave us access to his crack team of committee members working on this year’s application requirements.

.: Ricky Kendall from ORNL is managing the computational chemistry portion of the competition by introducing GAMESS into the mix. Ricky had this to say:

GAMESS is a great application for the participants because it demonstrates the needs of cluster resources for one rather demanding field computational chemistry.  Clusters configured to deal with the
GAMESS suite are able to run many applications.  GAMESS stresses all of the subsystems of the computational resource, in particular memory, interconnect, processor performance.  It also demonstrates some of the legacy aspects that computational scientists must deal with in the real world.

.: Tom Spelce of LLNL is managing two applications for this year’s competition.  The first, WPP, simulates ground motion during earthquakes.  According to Tom, it was recently in the spotlight during a simulation of the 1906 SDF earthquake.

WPP was chosen because one of the problems it simulates is the ground motion resulting from earthquakes. It is a relatively simple code which has some potential for source code modification to improve performance. WPP is the same code used by LLNL researches to reenact the 1906 San Francisco Earthquake on its 100 year anniversary.

Tom also had the task of bringing a CFD app [really a toolkit] into the competition.  He chose the widely used OpenFOAM [Open Field Operation and Manipulation] Toolkit.  According to Tom:

OpenFOAM is an OpenSource, parallel Computational Toolkit that can simulate a wide range of physical phenomena, from chemical kinetics, fluid flow and heat transger, to solid dynamics, electromagnetics and
even financial engineering.

.: Doug Fuller of ASU is the bio guru of the group.  Bioinformatics codes is still an emerging sector of HPC.  Doug chose two bio-centric applications for the competition, POY and RAxML.

POY and RAxML are two separate, related bioinformatics applications that will expose students to newly developed parallel software that is on the frontier of this young and exciting field of computationally intensive research.  In particular, the datasets and results produced by these applications will be readily understood by the students and the audience: from genetic sequences, the students will use these applications to analyze the genetic differences among organisms and their relationships to each other.

We were also able to get an inside look at this year’s Cluster Challenge theme.  If you recall, last year’s theme featured air racing.  Keeping in line with the competitive atmosphere, the Cluster Challenge theme for 2008 will be cycling.  Austin has a large contingency of cyclists, not to mention its the home of Lance Armstrong.  Trek bicycles graciously donated plenty of eye candy for the event.  This, being the perfect forum for such an organization.  Where better to recruit the latest crop of CFD specialists right out of school?

Make sure to peddle your beach cruiser over to the Cluster Challenge showcase and check out the exciting competition.

Also posted in HPC | Leave a comment

InsideTrack: Rocks on Solaris

Earlier today, we posted a preview of the Sun booth at SC08.  In it, we speculated on the appearance of the Rocks cluster distribution in their booth, running Sun’s Solaris operating system no less.  For once, our speculation was correct!  Thanks go out to Mason Katz, Greg Bruno and Anoop Rajendra [Solaris port project lead for the Rocks team] for confirming the news.

Anoop’s response:

Our display at the Sun booth will consist of -
1. A small cluster running an alpha version of Rocks on Solaris.
2. Fully automated provisioning of Solaris compute nodes, and
Thumper/Thor (NAS) appliances as part of the cluster from a Linux
frontend.
3. Rolls support
4. MPI support using the Sun HPC Cluster Tools product.
5. Sun Grid engine batch system support, and demonstration of running
MPI jobs using SGE

He goes on to comment….

The most important take-away from this, (we hope) will be that – Rocks
brings the same ease of installation and use to deploying Solaris on a
cluster, as it has brought to Linux for the past few years.

Very cool stuff!  We’ve heard rumors on this for quite some time, but no concrete news to report.  It has now been confirmed and you heard it first on InsideHPC.

Also posted in HPC | 2 Comments


Video Archive

insideHPC.com is a production of insideHPC, LLC. © 2006-2013 Sitemap