... they couldn't ask him about the $34 million loss and the 2500 jobs to be axed.

like (the now defunct) Star Office, or (the now defunt) Forte/Sun One studio...

For webservers, databases and file servers the T1000, and T2000 are ahead of anything else especially in terms of performance per Watt

... they couldn't ask him about the $34 million loss and the 2500 jobs to be axed.

Even worse, all of the benchmarks that Sun throws around for these things themselves require some fairly specific configuring of certain software on Solaris, and recompiling in Forte or Sun Compiler Studio (whatever it's called now) as Sun reps have been telling you for years whenever a gcc query has popped up. Quite frankly, a lot of people decided that it all wasn't worth the hassle years ago.

Such metrics are pretty meaningless overall. All CPU makers now, including AMD and Intel, are talking about putting more cores on a chip and doing more in terms of parallel applications and threading (corner cases for performance improvements really) when people really want to do a task twice the size in half the time and get more through - and the x86-64 processors will still always cream SPARC there. In the seven years since I seriously looked at Linux/x86 and Solaris/SPARC head-to-head, 4370 pystones/sec on an UltraSPARC versus 17,543 pystones/sec on a 1.4GHz Athlon was a pretty big no brainer, and that's why lots of academic institutions in particular jumped off. I don't see that the situation has improved.

As a casual bystander, I can't help but notice that while you were careful to point out a metric of raw performance and performance per watt, you carefully avoided talking about performance per dollar.

I'll leave it as an exercise to the reader, and Jonathan, to work this one out, considering that by various metrics a good three quarters of the servers they manage to sell are x86, and of those, probably a good 70% - 80% have Linux specified as their OS. Unless they can get some returns on their spending on SPARC and Solaris then it's all just dead weight.

Oh look! a single 8 core 1.4 GHz SPARC system is better than a 16 core 2.9 GHz Xeon System in raw performance.

In performance per watt the Xeon box will look like a joke. Each of the Xeons in that box take 130 Ws so 4x130 is 520 Watts for the cpus alone. The UltraSPARC on the other hand consumes 95 watts normal max 123 Watts.

Why does an Intel based system need 5x the power and 4x the cpus to produce worse results than a single UltraSPARC chip?

Doesn't look like your ignorant statement " the x86-64 processors will still always cream SPARC there."

BTW Academic institutions are back. http://hpcvl.org/hardware/victoria-falls.html

"Once fully installed, the cluster, called "Victoria Falls" will be able to process almost 10,000 threads.""

Contrary to the current hype, all that extra (and expensive) software you gratuitously added to an otherwise economical machine is generally unnecessary. I configured it with just the RHEL 5 and it came to $13,000.

I really hope all Sun customers have your kind of faith. Really. I do.

Repeating and regurgitating Sun's own benchmarks, verbatim, counts for very little.

On raw performance? Four Xeons versus one SPARC, all at twice the clock speed (not that that counts for much as a comparison)? Errr, no.

But you keep telling yourself that. One SPARC would never keep up with one Xeon in a month of Sundays, and the sorts of target workloads that Sun seems to be using would have to be so parallel and so concurrent as to be totally unrealistic. Not everything is or can be, and even in the article I linked to the UltraSPARC couldn't even outperform the Opteron on that.

I'm sure Sun can find lots of other completely arbitrary, moving target units of measure such as 'performance per watt' to make SPARC look better. Whether that is really enough, we'll have to see. Something tells me that Sun hasn't learnt from what happened after the dot com boom.

Right on cue. Speaking of arbitrary units of measure...... You can't just tot things up on Sun's power calculator on their web site and expect that to answer a real world question.

It doesn't. It would help if you actually looked at what the results tell you, and it would also help if you actually knew what the power consumption cost of a Xeon was. You balance that versus the raw performance, and you could halve the CPUs to two and halve the performance of a Xeon to cut power right back (a more realistic test) and its raw performance would still be better.

What people are looking at is whether it is worth spending the money to get any future power savings, versus having the raw performance per cycle. Nobody cares about Sun's theoretical performance per watt. Fact is, compared to an Opteron box, the guy worked out that his UltraSPARC would have to be at least twice as power efficient to feel the effect of cost savings over a period of several years for Coolthreads to be worth it. Sorry, but that scenario doesn't add up.

Hate to break it to you, but the raw performance of SPARC has lagged behind x86 for a very, very long time.

I'd laugh if that wasn't so sad. Why do you think they left in the first place? I'm sure Sun gave them a nice deal and some new toys to play with ;-).

The issue here is how much of each thread can be completed per second (is it better to get more threads and work through each second, or is it better to have more of them?) versus the initial cost of the machine versus the time it takes for the power cost savings to outweigh the initial cost. That's what matters.

Coolthreads (and multiple cores in general) is simply a tough sell for people wanting to complete more of the same tasks in less time following Moore's Law, and that accounts for the majority.

The bottom line at roughly the same cost of acquisition the Sun boxes perform better and at a lower power envelope.

No. Adding the extra memory and drive, which I was not including before, comes to $25,000. RHEL5 comes with Xen. That was the whole point of it being so late. You keep wanting to stack on useless and expensive software.

$39,000 is 56% more than $25,000. That is not "roughly the same cost". Why don't you send me $39,000 and then I'll send you back $25,000 and we'll call it even. OK?

For that matter, even using your figure, I'd agree to the exchange if I had to send back $32,000. I could use a quick $7,000.

"In a May 2006 report, IDC reported that 66.3% of the AMD-based Galaxy servers in Sun's portfolio ship with Linux pre-installed. By September, that number had increased to 71.5%."

"A Sun representative contacted for this article declined to update those numbers, and said the company does not break out the number of servers shipped for analyst reports."

BTW, you seem to have some emotional investment in this. People don't usually throw around terms like "vile ignorance" regarding issues which they are viewing objectively. I don't really care whether Sun or HP "wins" this contest. But the evidence is simply not on your side, even using your own figures.

Well, I know you want to nitpick without addressing the crux of the arguments, but I'll bite.

http://tweakers.net/reviews/649/last/database-test-sun-ultrasparc-t...

There should be some reasonable data in there for you to digest. The crux is that Sun's x86 Opteron servers are outselling SPARC by a wide margin. It's a question of volume, and that's why Sun had to move to it. Having their SPARC be outperformed and still be more expensive in the same target market isn't helping.

Given that Linux has traditionally been installed on x86 and that's where Sun's SPARC lunch was eaten, that's fairly logical. It's difficult to justify Solaris under those circumstances.

In all honesty, that doesn't surprise me. If the opposite was true, the Solaris folks would be trumpeting it from the hills. They're not.

That's the long and the short of it, and alas, there isn't much cheer to be had. Sun have some good stuff, and with good leadership they could be doing very well, but due to inertia and poor management decisions that isn't going to happen.

Your ignorance is even more telling. Cpus consume more energy at higher clock rates and dissipate more heat.

Oh you are still stuck like a broken record on that article you posted

Read and comprehend first. Then do some research and come up with some real data to the topic at hand. Some up to date data would be very useful.

I wonder why the coolthreads systems are doubling in revenue YoY.

I forgot that you don't live in the real world.

You're tripping sweetheart. What matters to people is what work gets done in that clock cycle for the amount of power they use and what they paid. It really is that simple.

Yer, because that's a potential Sun customer (you know, money?), and that's what matters to him. As it is, he's looking at performance per watt per dollar and the raw figures just aren't there. Specweb justifies nothing to him.

You're making a huge assumption that Specweb is actually saying something because you have no data yourself, and you've proven yourself totally incapable of explaining it or discussing what's happening.

When you've come from nothing doubling is easy. Once again, no exact figures from Sun ;-).

Also RHEL 5 only supports 2 processors. You need RHEL Advanced Platform (more $$$). RedHat's product matrix is clear about this.

What? Configuring a box similar to the Sun boxes makes the HP proliant 32K. End of story. Your nitpicking isn't going to change that.

I wouldn't because I am not as short sighted as you are. Power and cooling costs alone would handily eat that margin over the life time of the system (typically 5 years)

All versions of RHEL support 16 cores just fine. HP gives you a choice of a RH support contract or not.

No, it comes to $25,000 and a 56% difference in price is hardly nitpicking.

Surely you can handle simple arithmetic?

I ask again. Why are you so desperate to prove that that the Sun is not more expensive, even when your own skewed figures show it to be?

I mean, if it is, great; I don't really care. I'm mainly curious what has you in such a tizzy.

It appears you can't. Where is the extra 425-407 watts of power the thing consumes gone??

At $0.10 per kw, that works out to about $350 per year. Add another 33% for cooling and that makes about $466/year savings. That represents a couple thousand over the 5 years you set as the useful life of the server. A fraction of what you need to "handily" make up for the exorbitant price of the Sun box. Of course, if that money had been invested elsewhere, accruing, say, 10% per year instead of being tied up in an overpriced server it would have made $8500 over that 5 year period instead of a couple of thousand.

If you would like to waste more time in a futile attempt to prove... whatever it is you are trying to prove, be my guest. It's your time to waste.

It seems that some people are just stating things about SPARC and x86 as facts, without proving it with the hard numbers? They have no substance in their words, no hard figures, no proof?

Opinions are like assholes; everyone things everyone else's stinks. The reality is that one shouldn't take benchmarks all that seriously; look at the information and make the decision yourself - don't rely on so-called experts or benchmarks because most of their time they're wildly distorted to what reality is.

Look at the charts. The chart on the left is Total Server shipments. On the right x86 shipments.

"Sun indicated that it had sold slightly more servers, but a much sharper rise of Opteron sales reveals that Sparc sales are still on the way down."

Sun sells more SPARC systems running Solaris than all the x86 systems combined. See above.

But the opposite is true but your feeble mind can't comprehend it. Sorry try again.

BZZZT! Smugmug is a Sun customer and has since posted very positive feedback about Sun.

It doesn't show that at all. What it shows is a yearly growth comparison. Sorry, I really don't know what it is you work on, or what interest you have in this, but you simply cannot read. Did you not read the article, or did you just spin in your chair?

"Sun indicated that it had sold slightly more servers, but a much sharper rise of Opteron sales reveals that Sparc sales are still on the way down.

Sun sells more SPARC systems running Solaris than all the x86 systems combined. See above.
You can't read a simple article. See above.

You've been presented with two articles that say otherwise. Failing to read the first article in order to see what you want to see, and then completely ignore the second because you saw what you wanted to see in the first is..........a pretty damn sad thing to do in all honesty.

"BZZZT! Smugmug is a Sun customer and has since posted very positive feedback about Sun.

"Sun indicated that it had sold slightly more servers, but a much sharper rise of Opteron sales reveals that Sparc sales are still on the way down. At the moment, Niagara generates returns of 100 million dollars per quarter, on a total of 1.3 billion in the area of servers."

Those two graphs showed server unit growth, versus how much their x86 sales are growing. What they did not show was total units sold as you think that they do. From TFA:

""Sun indicated that it had sold slightly more servers, but a much sharper rise of Opteron sales reveals that Sparc sales are still on the way down. At the moment, Niagara generates returns of 100 million dollars per quarter, on a total of 1.3 billion in the area of servers."

if in Q405 Sun sold 90,000 units and in q406 they sold 100,000. In Q405 they sold 21000 x86 boxes and q406 they sold 29,000.

then they sold 2000 more Sparc boxes than before.

You've just described the point of why I posted that article (growth of shipments). It illustrates the point within the context of this thread:

Sun are funding an entire hardware platform that was totally outgrown and outperformed both in growth and overall revenue by their x86 business, and what they ship each quarter.

SPARC sales are declining. x86 is just taking up more of the whole, and they're accounting for close to all of the growth, if any, that Sun gets each quarter.

Sun simply can't afford to hang on to that and expect to compete with their own x86 servers that they sell either in terms of growth, revenue, margins or the costs that they incur.

Their SPARC business is going to continue to decline unless they can put SPARC where IBM's Power is or pull their fingers out and give SPARC the raw power of x86 to attract people. Neither looks likely.

It occurs to me that I have focused on one tree and momentarily forgotten about the forest. I would like to correct that because I feel that this is an important point that should not be allowed to fall through the cracks.

We have been arguing about the performance per kw in purely monetary terms: this month's electric bill. While it is true that this is what most businesses would be interested in, we should not forget that there are many other important reasons that a more energy efficient server is desirable.

Dependence upon foreign sources of fuel is something about which many of us are concerned. Conservation of remaining fuel reserves is also something which should be of concern. And whether one believes in global warming as a threat or not, few would take the position that burning fossil fuels is not damaging to the environment for reasons which go beyond the effects of greenhouse gasses.

For all of those reasons, the electric bill is only a small part of the actual cost of a less efficient server. In the long term, these are reasons which should be significant to the businesses in question. But even if they are not, we are all in this together. The next door neighbor who drives a needlessly inefficient SUV affects us all. Likewise, the business that uses less efficient servers affects us all.

I seem to have developed a bit of tunnel-vision earlier in this thread, which the above should rectify.