Burton Smith, a longtime supercomputer designer and chief scientist at Cray, has resigned to take a position at Microsoft. Microsoft announced two weeks ago that it planned to introduce a new version of its Windows software for scientific and engineering users, and that Smith would be involved.
interesting step. Too soon to say much about it though, but I expect more information to come along. Then we’ll see what we ought to think of it.
Windows for scientific and engineering users? They must be kidding. What else after the coward attack to linux?
For one, MS has barely ever attacked Linux. 95% of attacks are purely formed in the minds of linux zealots.
Anyways, they arn’t kidding, and people who think this is a joke are clearly clueless.
What are some of the main apps used by Science and engineering types? Apps like Matlab/Simulink, Excel. They have them on their regular desktops but are often at the limits. So having the kinds of systems MS is going for is just what they are looking for. They don’t want to mess around with some non-windows based solution. Its about getting the job done.
People can think MS taking on clusters/supercomputing is a joke all they want. But in the end MS will take over that market.
Microsoft seems to like to collect some real giants in the industry with resumes that simply don’t match the MS products we see in the market place.
If Burton and others actually got some threaded architecture action together at MS, then perhaps they may also get an OS for it in the Singularity project which would need a MTA processor to really shine.
transputer guy
Maybe he will fit in as well as Ray Ozzie *cough*
one of microsoft’s tactics has always been to buy competing companies to either steal their code or just shut them down.
hiring their programmers is a cheaper alternative i guess – they’ve done similar things with hiring google’s financial people.
lets face it, no researcher worth his/her salt would be using anything other than unix and this is just an underhand attack on unix by microsoft.
Unix sucks for supercomputing… I’m not saying windows is good. Just pointing out.
you are joking right?
what do you think is running most of the world’s supercomputers?!
aix, hpux, solaris, bsd, linux, all unix/system-v variants, not windows.
Which part about it is bad for supercomputing and how?
Personally I think Unix is very adept at supercomputing, it’s graphics/video that it shows its age in.
Even then, there are moves to fix up the problem; XCB is one example; actually replacing the crusty under pinnings of the existingn XLIB with something that allows nice ‘n clean ™ threading without need to jump through several burning hoops.
X11 is getting there, be it having to start from scratch to the ‘hold back’ nature of the old XFree86 hierachy – but thanks to the more inclusive nature of the new setup, it might take 18 months or so, but X11 will get up to speed and on a stable footing – it just takes a while unfortunately.
It is just absolutely futile endevour for MS to get into supercomputing sector, they will have absolute zero success. Their crapola Windowz garbage OS will never get any traction in 100% Unix dominated industry. I can already see roaring laughter of HPC engineers laughing in the faces of MS sales people trying to push Windows for supercomputers.
Right, And MS should never pursue the server market either.
And MS should never pursue the hand held market.
When will people just accept that in time, anything MS does, they will come out king at. They have the home desktop, the work desktop, the handheld market, tablet market, strong foothold if not lead in the media center/dvr market, and working their way towards taking the server market.
People who don’t want to go writing ground up apps and doing a lot of work just to be able to due there work will suck these things up like mad. Most people will take just pulling some computers out of boxes, hooking them up, and a few clicks later they are their over having to deal with unix, or creating custom applications and so forth.
I’m sure Developers for programs that need this kind of thing, like Mathworks are eagerly awaiting this. Probably some CAD developers too.
It is just absolutely futile endevour for MS to get into supercomputing sector, they will have absolute zero success. Their crapola Windowz garbage OS will never get any traction in 100% Unix dominated industry. I can already see roaring laughter of HPC engineers laughing in the faces of MS sales people trying to push Windows for supercomputers.
BWAHAHAHAHAHAHAHA! Can I quote you on that? Seriously, dudes … a lot of people tend to write MS off but, then, over time we see headlines like “MS Tops In Embedded Systems Market”.
MS doesn’t care if it loses in the short term. It makes investment and it keeps hammering away at competitors until it takes away significant market share.
Look at the server market. People were shocked recently when MS beat Linux in server sales. Linux was once thought to have this category sewn up. Nope. MS didn’t care. It kept improving its server products, making them bone-simple to operate, and far more secure than the past. It seems to be working.
Look at the emerging smartphone market. Symbian and MS are already competing fiercely for the next generation of phones. Even MS-hater Nokia has been leaning toward Redmond lately. Palm is using Microsoft’s OS on its smartphones.
Here’s a good piece of advice: If you’re gonna write MS off, make sure that you’re watching your back. Because odds are, MS will eventually catch up. It has the money and the time.
I agree absolutely. But you said it “odds are, MS will eventually catch up. It has the money and the time.” but no quality. They will be haphazardly put together layers slapped on top of another. That is MS’s way of writing OSes. Check out Vista…I know that it is not feasible to write from scratch…but you can only recycle crap for so long!
When will people just accept that in time, anything MS does, they will come out king at.
Uh, no. In fact, apart from the OS and Office product lines (and, recently, Servers), MS has actually had a lot of trouble getting dominance in other markets.
Take the DVR market: they are actually very far behind other products (such as TiVo). Sure, they own the Tablet PC market, but it’s a very, very small market (and not increasing much).
In Web services they are hurting, in Game Consoles they face stiff competition from Sony (the leader) and Nintendo (the true innovator).
I think you are overestimating Microsoft’s ability to conquer new markets…and definitely, from your post, you don’t understand what supercomputing is about at all.
Uh, no. In fact, apart from the OS and Office product lines (and, recently, Servers), MS has actually had a lot of trouble getting dominance in other markets.
You do realize that those markets generate BILLIONS of dollars in profits and allow MS to outlast and outpurchase competitors in other markets, right?
Take the DVR market: they are actually very far behind other products (such as TiVo).
Media Center is more than just a DVR. It’s a brand new category.
Sure, they own the Tablet PC market, but it’s a very, very small market (and not increasing much).
Several generations from now, people will be buying nothing but tablets. It just takes continued investment. MS has the time and money.
In Web services they are hurting…
Not true. .NET is actually extremely popular and becoming more popular every day with devs.
in Game Consoles they face stiff competit
ion from Sony (the leader)
Assuming that Sony ships something in a year.
and Nintendo (the true innovator).
Get real. Nintendo is finished, whether it knows it or not. The console fight is going to be between Sony and MS. Nintendo is kind of like Sega, at this point.
I think you are overestimating Microsoft’s ability to conquer new markets…
Why? He made some good points about MS winning markets that it was supposed to never even have a chance in.
and definitely, from your post, you don’t understand what supercomputing is about at all.
But that’s the whole point. MS (like other server vendors) has reduced the technical overhead for building networks. Why should supercomputing remain solely within the realm of the ivory tower? Why not allow companies to build supercomputers out of retail components rather than proprietary boxes to tackle tough problems? The *nix vendors want you to believe that you either need to go completely-freeware or completely-expensive-commercial in order to build a supercomputer. MS can promote the plan that there’s another, simpler path. I don’t see anything wrong with that. It will undoubtedly have a lot of appeal to companies.
“Why not allow companies to build supercomputers out of retail components rather than proprietary boxes to tackle tough problems?”
Yes, and we’ll all have MS-Crays in our bedrooms, and government organizations, banks, multinationals, military and others will just stare in envy with their jaws on the floor. Please Billy come and liberate me from my socalist heterosexual delusions.
You do realize that those markets generate BILLIONS of dollars in profits and allow MS to outlast and outpurchase competitors in other markets, right?
In theory it would, but that’s not what we’re seeing. What we’re seeing is that MS pours money in order to try to conquer new markets without suceeding.
Look, MS is very, very rich, but it’s still ONE player out of a very large number of players. It can’t just buy its way into all of these other markets, and looking at the facts tends to prove this.
Media Center is more than just a DVR. It’s a brand new category
One that isn’t selling at all compared to DVRs, which are really taking off.
Several generations from now, people will be buying nothing but tablets. It just takes continued investment. MS has the time and money.
That’s pure speculation. You can’t know what people will be using several generations from now, and you certainly don’t know what Microsoft’s position will be in that market at that time. What if Apple released an iTablet? With their brand power, they could dominate that market like they did with the iPod.
In any case this isn’t a tenable argument as it is highly speculative. For now, the tablet market is marginal, so Microsoft’s “dominance” is irrelevant.
Not true. .NET is actually extremely popular and becoming more popular every day with devs.
By Web services I meant things like MSN, which has lost money for MS since day one. Here they’re facing Google, which is a very real threat to MS.
Assuming that Sony ships something in a year.
Why wouldn’t they? They’re the market leader.
I have to say, I’m starting to find your absolute confidence in MS rather suspicious. I really hope you’re a dedicated fan and not a MS employee…
Get real. Nintendo is finished, whether it knows it or not.
Uh, no. Not by a long shot. Take it from someone working in the industry: Nintendo ain’t done – their revenue from GBA and DS are sufficient to help them overcome the average results of the GameCube (which only sold marginally less than the Xbox, by the way).
I’m sorry, but that’s either pure fanboyism or well-rehearsed astroturfing. Nintendo isn’t going anywhere.
Why? He made some good points about MS winning markets that it was supposed to never even have a chance in.
No he didn’t. He claimed that MS won those markets, but the fact is that apart from a few core markets, MS isn’t the leader in those markets it’s trying to enter.
But that’s the whole point. MS (like other server vendors) has reduced the technical overhead for building networks. Why should supercomputing remain solely within the realm of the ivory tower? Why not allow companies to build supercomputers out of retail components rather than proprietary boxes to tackle tough problems?
I see you don’t know much about supercomputing either. Supercomputers are already built out of retail components to minimize cost, and the vast majority use Linux for the exact same reason.
Linux is highly configurable, free and open, all things that give in a very strong edge in the field of supercomputing. Basically, you don’t need to bother with a vendor at all, you can just build your supercomputer yourself, and that is what is very appealing to research centers, universities and companies. This isn’t the right market for MS to propose a “simple” path, because there is already one there that’s simple, free and open.
As with consoles, Microsoft is a late entry to the market, and trying to move into the space IBM and Sony will be carving out with Cell. The future I see emerging is that Cell workstations and rack mounted boxes, whether of the IBM or Sony variety, will be fully deployed before Microsoft and Intel have generated an equivalent dollar value of paper.
The console and supercomputer areas are the two biggest zones Microsoft needs to succeed in if they’re to move beyond being a one product shop. Latin America, Africa, the Far-East and the European Union are embracing their own solutions. Within the decade, Microsoft will be a, primarily, American based vendor with marginal influence outside of its home market.
Of course MS is dominating the Tablet PC market! They’re the only ones in that market!
In theory it would, but that’s not what we’re seeing. What we’re seeing is that MS pours money in order to try to conquer new markets without suceeding.
Name some of these new markets that MS is pouring billions into “without succeeding”.
One that isn’t selling at all compared to DVRs, which are really taking off.
No wonder. They’re not competing against DVRs.
By Web services I meant things like MSN, which has lost money for MS since day one.
Wrong, yet more BS. MSN has been profitable since 2003.
Here they’re facing Google, which is a very real threat to MS.
Google is a one-trick pony. It derives nearly all of its revenue from advertising. You claim that MS isn’t winning new markets? Neither is Google.
That’s pure speculation. You can’t know what people will be using several generations from now, and you certainly don’t know what Microsoft’s position will be in that market at that time. What if Apple released an iTablet? With their brand power, they could dominate that market like they did with the iPod.
Of course it’s speculation. I never claimed otherwise. But I’m still right — and you’ll see.
Why wouldn’t they? They’re the market leader.
Because market leadership with a legacy product is no predictor for future success.
I have to say, I’m starting to find your absolute confidence in MS rather suspicious. I really hope you’re a dedicated fan and not a MS employee…
Of course you do. Because nobody could possibly support MS other than a MS employee, right? Just like somebody who supports PS3 has to be a Sony employee… /sarcasm
Uh, no. Not by a long shot. Take it from someone working in the industry: Nintendo ain’t done – their revenue from GBA and DS are sufficient to help them overcome the average results of the GameCube (which only sold marginally less than the Xbox, by the way).
Uh, hellooooooooo. We’re talking about the console market. Not handheld game devices. As I pointed out, Nintendo is finished in the console market. From now on, they’re going to place 3rd. So, it’s really a question of whether they can tolerate losing to MS and Sony, and whether their efforts are better focused on their strengths in the handheld market. Personally, I’m amazed that Nintendo hasn’t found a way to leverage its way into the cellphone platform credibly.
No he didn’t. He claimed that MS won those markets, but the fact is that apart from a few core markets, MS isn’t the leader in those markets it’s trying to enter.
Be specific. What markets are you referring to?
I see you don’t know much about supercomputing either. Supercomputers are already built out of retail components to minimize cost, and the vast majority use Linux for the exact same reason.
You’re not paying attention. Using Linux or Solaris or whatever to build a supercomputing grid/cluster isn’t about taking a bunch of off-the-shelf components and plugging them into one another. That requires considerable sweat and configuration. Microsoft’s focus will undoubtedly be to make the entire configuration process easy and seamless. If you pay attention, that’s how MS operates. That’s how MS wins.
Linux is highly configurable, free and open, all things that give in a very strong edge in the field of supercomputing.
Yeah, yeah, yeah, yadda, yadda, yadda. You guys said the same thing about embedded systems, and MS owns that market solid.
say NO to monopolies
say NO to proprietary standards
say YES to open source
say YES to healthy competition
say YES to open standards