Feeds

back to article Larry Ellison's first Sparc chip and server

Oracle has announced the Sparc T3 processor and its related Sparc T3 systems at the OpenWorld extravaganza in San Francisco, giving Solaris shops who had run out of headroom on the existing Sparc T2 and T2+ machines a little breathing room – and giving Oracle a chance to chase some entry and midrange Unix server sales against …

COMMENTS

This topic is closed for new posts.
Grenade

his first own SPARC ?

I do remember rainbow falls was put on the roadmap by one jonathan schwarz, erstweil CEO of SUN microsystems.

So, no, I do not think this is proof of Oracle's commitment to SPARC.

ROCK as it should have been plus 3 years catch up vs IBM's Power, that is what we would want to see......

AC

4
1

It's basically there for high-end workloads...

If Larry Ellison wasn't committed, the chip would have died a quiet death within a week of the buyout, and nothing more would ever have been heard of it. Obviously, he thinks it's something worth sinking money into. Sun really did do some excellent engineering, it was their sales that stunk.

1
0
Anonymous Coward

Larry likes lots of cores

It's all about selling Larry's software.

16 cores * 4 chips = 64 cores *.5 core factor table = 32 licenses * $45K for WebLogic suite = $1.4M

I would say they are giving the boxes away for free to get the software but they have actually reduced the discount to only 30%.

We are going to Power7.

Exalogic is even worse. 360 cores * .5 = 180 licenses * $45K = $8.1MILLION

That America's cup was paid for with blood money.

0
0
Coat

the chip ....

was already there, and 90% of the money was already spent.....

and T3 would not get in the way of something else, since there isnt anything else.

Rock would have stolen sales from M-series and T2 but not deliver on promises, that is why it was cancelled. Or at least, that's what the sun guys told us in june last year.

AC

0
0

T3 is faster than POWER7

according to slides that Oracle showed. For instance, against a POWER750, the T3 Niagara machine was 32% faster in some benchmark. Oracle announced 8 world records, with more to come.

0
2
FAIL

Get real.

8 World records ?

*CACKLE*

Lets see... from Oracle's benchmark site.

1) SPARC T3-4 Server with Oracle Fusion Middleware 11g Achieves World Record Single-Node Result on SPECjEnterprise2010 Benchmark

Yet another obscure SPEC benchmark that nobody has heard about. Where there have been submittet a stunning 13 entries. And the only way that they could manage to get in.. second as there is no separate clustered category. Is to divide the T3-4 up into 8 domains and run like a single box cluster. *cough* *cough*

2) SPARC T3 Servers Offer an Attractive Platform for Virtualization and Consolidation

So we ran some tests in the lab, with some old x86 hardware at low utilization and then on our biggest T3 box did the same.. showing that we are better. Yeah World Record!!!

Get real.

3) SPARC T3-1 Server Delivers New Record Score on Oracle's JDEdwards EnterpriseOne Benchmark

Again one of Oracles own benchmarks. Kind of hard to not give them this one. As there are no official list or tables to really compare against anything else. Again an obscure benchmark made by Oracle themselves... What's that smell...

4) Oracle Publishes New Result on Consumer E-Commerce Site Benchmark Using SPARC T3-1 Server.

Yes again an benchmark made by Oracle themselves that haven't got an official site that you can compare results against each other.

5) SPARC T3-1 Server Posts a High Score on New Siebel CRM 8.1.1 Benchmark

Yes again an benchmark made by Oracle themselves that haven't got an official site that you can compare results against each other.

6) Oracle's SPARC T3-2 Server Shines on Industry-Standard, General-Purpose Java Benchmark

Now this is the SPECJVM2008, a benchmark that has ever been posted 6 results for.

http://www.spec.org/jvm2008/results/jvm2008.html

These results include 3 SUN x86 results and 3 iMac results....

And with 2 processors with 32 cores the T3-2's PEAK value of 323 surpassed another SUN BASE entry of 317 made with an x6270 with 2 Nehalem-EP processors.

Yeah right. Good Oracle you found yet another obscure benchmark to feed to your followers. Come one.. honestly.

7) SPARC T3-1 Server Shines on PeopleSoft Enterprise Financials 9.0 (Day in the Life) Benchmark

"This is the first publication of this benchmark by any hardware vendor world-wide."

Yes we make the benchmark ourselves and make a new version, that we can be the first to publish in and claim a world record.

8) Oracle's Newest SPARC T3 Servers Deliver Outstanding Results with Oracle Communications Order and Service Management

Yes again an benchmark made by Oracle themselves that haven't got an official site that you can compare results against each other.

9) Oracle Publishes First Result Using Online Component of Extra-Large Workload on E-Business Suite R12.1.2 Benchmark

Yes we make the benchmark ourselves, and just created a new category with nobody else in it, so we can call it a World Record. Get real. man...

Sorry it's pathetic.

// Jesper

1
2

@Jesper Frimann

Good ole Jesper. Never denies himself. When Sun wins a benchmark, that benchmark is "crafted by Sun" (including benchmarks crafted by IBM) or "cherrypicked by Sun", etc etc. He has numerous explanations. All non-IBM winning benchmarks are immediately rejected for different reasons, one reason weirder than the other. For instance, "even though you need four POWER6 to match two Intel Nehalem on TPC-C the POWER6 is faster, because the cores are faster". Who can win against such a fantastic logic? No one.

IBM is free to respond to these world records, Jesper. Let us see what POWER7 can do. If people claims it is the fastest cpu in the world, then it should have no problems?

.

Jesper, FYI. Next year, the Niagara T3 will be upgraded. For the first time, the cores will be improved. Earlier, Sun just added more and more cores and threads. But the cores were always very simple cores. Beginning from next year, the cores will be beefier, have out-of-order execution, etc etc etc. This will result in 3-5x better thread performance. :o)

And 2013 there will be 8-socket T3+ servers. Jesper, those new upgraded T3+ servers will kill everything that IBM can throw at them. I know even now, you will reject all those benchmarks and results, but I just wanted to prepare you mentally for T3+ arrival. The performance of an 8-socket T3+ server with 1024 beefy threads will be truly shocking.

Oracle has admitted that the new T3+ cpus will start to eat into the normal SPARC server market share (because T3+ have beefier cores) but that is ok.

But wait until you see 64-socket T4(?) servers with 16384 beefy threads and 64 TB RAM. Eat your heart out, IBM. :o)

1
1
WTF?

Kebafart

Where are the specint, and specjbb results that Sun touted the last time around? Oh, yeah they can't compete there anymore so they won't publish those.

0
0
Thumb Down

Welcome to the Real World

"IBM is free to respond to these world records, Jesper. Let us see what POWER7 can do. If people claims it is the fastest cpu in the world, then it should have no problems?"

Listen Keb. Why would anyone use time and money to try to beat benchmarks that cannot be compared public, cause it's Oracle almost internal benchmarks.

It would be just as pointless as Oracle posting rPerf number for T3, as for IBM or HP to try to figure out what obscure benchmark Oracle/SUN now have chosen to publish.

The only ones that listen are people like you. "Decision makers" listen to Industry standard benchmarks, and sorry Oracle haven't published one yet for the T3 systems.

// jesper

1
0

@Jesper Frimann

When IBM did their own benchmarks to show the superiority of some obscure thing, and when Sun did the same benchmark with over 10x better performance without any optimizations at all, you also objected as "cherrypicked by Sun" etc etc etc.

Anyway, if POWER7 beats the T3 easily, it should be no problems for IBM to run the same benchmarks over an afternoon with no optimizations at still beat T3, yes? So, go ahead. Or, maybe the POWER7 does not beat T3? I can not wait until the T3+ with 3-5x better thread performance arrives next year.

Before T3, when IBM compared POWER7 to T2, I had no objections. I accepted all benchmarks by IBM and I admitted that POWER7 is faster. I did not reject any IBM results. But, you always reject every benchmark that shows IBMs cpus to be slower. And the rationaly you provide is somewhat interesting. I would never have come up with such a weird logic no matter how much I thought. I do not think ill logically, those thoughts never come to me. Weird.

.

Regarding the rest of the benchmarks, we have to wait and see when Oracle publishes those benchmarks. But as I said, if POWER7 is that fast, then there should no problems to best the T3.

0
1
Coat

wake up.

"with over 10x better performance without any optimizations at all, you also objected as "cherrypicked by Sun" etc etc etc."

What benchmarks are those ?

"Or, maybe the POWER7 does not beat T3? I can not wait until the T3+ with 3-5x better thread performance arrives next year."

Oracle claims up to x3 better thread performance... But Yosemite falls, as the processor is called have half the number of cores compared to the T3, and surprise it runs at 50% greater frequency. So .. the chip is not 3-5 times faster. It's more likely 1.5% faster.

You have been constantly peddling T3 against all other processors, and now it's here and then you start peddling T3+. Man you sound like a Itanium sales guy.

// Jesper

1
0

@Jesper frimann

You know what IBM benchmarks I talk about. You discussed it with me. T2 Niagara was 10x faster, and you talked about cherry picked benchmarks by Sun. But, it was IBM's own home brew benchmark! And still you are never satisfied. Even when Sun runs IBM's own benchmarks you dismiss the Sun results. How can Sun win against such a fantastic logic? :o)

.

Interesting, I didnt knew that about T3+. Can you show links on that?

Regarding the x3 better thread performance, I talked to the chief engineer behind T3 and he said 3-5x times better thread performance. Maybe Oracle do not want to exaggerate publicly and therefore only claims "up to 3x better thread performance" instead of "3-5x"? To me, he said 3-5x. It seems the official numbers are "up to x3".

0
0
Thumb Up

Glad to see it

It's nice to see someone who can finally sell the extraordinary technology that Sun is able to build.

4
0

Re: Glad to see it

I don't think selling servers was Sun's problem, it was the cost of developing them for an increasingly niche market. Still it is nice to see another decent architecture (basically, anything non-x86) being developed, although the chassis of the new servers is rather drab compared to Sun's wacky plastic covered ones of yesteryear.

1
0
Linux

Good to see that SPARQ's still alive.

I'm guessing this wont be GPL'd hardware though. I really didn't think Oracle would do such an about face on Sun's FOSS as soon as they got EU approval for the purchase. It's sad to see what they're doing to such an innovative albeit mismanaged company. Hopefully this can at least serve as a lesson for FOSS developers on what to avoid spending their time on in the future. Bloody corporate leechers!

0
0
Linux

BMSEER = Kebbabert ?

Interesting, I didnt knew that about T3+. Can you show links on that?

Ok what are you drinking.. it's just a few posts above:

Kebbert wrote this:

"And 2013 there will be 8-socket T3+ servers. Jesper, those new upgraded T3+ servers will kill everything that IBM can throw at them. I know even now, you will reject all those benchmarks and results, "

Funny.. Is Kebbert a new Oracle substitution for BMSEER, a alias used by slightly technical Marketing people to spread the Gospel.

The key word here is "up to". For example there is code where POWER6 is perhaps "up to" 20 times faster than POWER5, on a core to core basis. This can be code that Decimal floating point numbers, where POWER6 has a Decimal floating point unit. Or it can be code where POWER6 can use it's VMX (vector) execution unit.

But again you fail to mention that T3+ as you call it only have half the cores, of the T3. So.. well.. yes up to x3-5 times better single threaded performance doesn't translate into 5 times better throughput.. You are dreaming.

// jesper

0
0

@Jesper Frimann

No, I am not BMSEER. I told you, I work in a large finance company. One of the most famous in the world.

.

"....But again you fail to mention that T3+ as you call it only have half the cores, of the T3..." I fail to mention that, because I didnt knew it. Again, do you have links about that? I didnt knew that.

.

"....yes up to x3-5 times better single threaded performance doesn't translate into 5 times better throughput.. You are dreaming...." I have never implied it would be 5x better performance. That is your own interpretation. I only said it will have 3-5x better thread performance, that is all I said. I never said T3+ would have 3-5x better performance.

I know that 3-5x better thread performance does not mean the whole chip is 3-5x times faster. Of course I know this because I have told you many times that just because one POWER6 core is faster than Niagara T2 core, it does not mean the whole POWER6 cpu is faster. Just because the thread performance of T3+ has been increased, does not mean that the whole cpu is 3-5x faster. I know that.

IBM does not know that. If IBM had better 3-5x better thread performance, IBM would have claimed the whole cpu is 3-5x faster. Just as IBM claims "one core is faster, then the entire CPU is faster". "One thread is 3-5x faster, then the entire cpu is 3-5x faster". But this is FAIL LOGIC.

You are used to how IBM thinks and reasons and therefore it is you that are dreaming. Not me.

0
0
Big Brother

Hmm.. Well

Well you sure sound like BMSeer.

"I fail to mention that, because I didnt knew it. Again, do you have links about that? I didnt knew that."

http://regmedia.co.uk/2009/09/11/sun_sparc_roadmap.jpg

T3+ aka Yosimite falls. 8 threads and 8 cores.

"Just because the thread performance of T3+ has been increased, does not mean that the whole cpu is 3-5x faster. I know that."

Besides from your usual rant about POWER6 and what it means being a fast processor core versus throughput of a whole chip. Which basically still just showcases that you didn't really get what that discussion was all about.

I think that good increase in single threaded speed is just what the doctor ordered for the Niagara style processors. But one has to understand that this increase in single threaded speed does not mean that the processor is having flour in your mouth and whistling at the same time. The whole trick is to have both single threaded throughput AND good chip throughput at the same time.

Personally my guess is that Oracle will go from Fine grained multi-threading to SMT and then implement out of order execution, with perhaps an extra execution unit. Basically addressing the main problem with the Niagara style processors. That should pretty much account for the proposed speedup's.

This would also make the Niagara family a more useful processor, for the types of workloads that are used today, basically slowly turning a niche product around and pointing it in the same direction as POWER7 and Nehalem-EX.

But in doing that they'll basically sacrifice the whole idea behind the processor IMHO. And it is to be seen if the T3+ will do more throughput on a per chip level than the T3.

// Jesper

0
0
This topic is closed for new posts.