!!!The Ultimate K8L Thread 2007 & Beyond!!!

Quote:

---

Originally Posted by Alxea

http://www.tgdaily.com/content/view/31977/135/

Demoed Agena FX! Confirmed SSE4A & a 2.9Ghz speed bump.

---

Pardon me, but did you even read the article? Here is an exact quote:

Quote:

---

Details about Agena FX, which is rumored to launch with clock speeds up to 2.9 GHz, were not disclosed and AMD avoided showing any benchmark performance numbers.

---

The article clearly states the 2.9GHz is "rumored" and that AMD confirmed nothing about operating frequencies. Not trying to flame, well actually maybe a little bit, but part of want you wrote goes completely against what was stated in the text of the article.

So who's gonna skip Conroe and "Get All Three"?

Phenom + RD790 + R650... a tasty combination!

Suman

Quote:

---

Originally Posted by sdsdv10

Pardon me, but did you even read the article? Here is an exact quote:



The article clearly states the 2.9GHz is "rumored" and that AMD confirmed nothing about operating frequencies. Not trying to flame, well actually maybe a little bit, but part of want you wrote goes completely against what was stated in the text of the article.

---

http://www.theinquirer.net/default.aspx?article=39440

To correct that, B0 stepping gives speed increase. So you must not know about B0 stepping from my guess... no?

Quote:

---

Originally Posted by Sumanji

So who's gonna skip Conroe and "Get All Three"?

Phenom + RD790 + R650... a tasty combination!

Suman

---

I'd have sx with dat! ;)

Quote:

---

Originally Posted by Alxea

http://www.theinquirer.net/default.aspx?article=39440

To correct that, B0 stepping gives speed increase. So you must not know about B0 stepping from my guess... no?

---

Your original link to Tom's hardware had no reference in the article to stepping B0 and in fact stated the AMD confirmed nothing about clock frequencies. You then admit you were wrong, followed by trying to pass it off as my ignorance buy linking to an article by "The Inquirer", a publication known for its accurate reporting. Nice....

I linked to toms hardware when? lol And I'm sure you thought "inaccurate"? So your saying the IQ is accurate reporting? When thats not the only source of info you can get it from. Guess you don't look too hard for world wide facts... ;D

http://www.google.com/search?hl=en&s...celona&spell=1

If the ATI take over from AMD wasn't a true enough rumor once thought of to be impossible nothing would be true with your close mindedness... "when it happened". Next thing your going to tell me is K10 never happened, and benchmarks thought AMD has done and are saying are 50%+ faster then C2Q are really just hype?

When we have 400,000+ on ALU and 300,000+ on FPU performance at only 2.5ghz agenst a 3ghz C2Q that really says how fast a cpu is with truthfully said benchmarks from the ES sources, when intel only got 20% less in ALU and 50% less in FPU I'm sure C2Q is just killing K10 with a 70% less performance and power advantage all at if put on equal speeds. If being 70% ahead means K10 is slower and C2Q some how is faster by magic (A 4.7ghz C2Q equals a 2.5ghz K10 "OCing K10 anybody? a 6ghz C2Q is needed to equal a 3ghz K10." lol)... If intel can make 6ghz C2Q's stock... to ever counter a K10... That will be the day this girl goes bi unless I can OC it to 3.5ghz then intel can't touch this smexy box of mine. lmao

I'm not sure if this has been posted before, but AMD has published some estimated Barcelona benchmarks. The estimates are for a dual processor, 2.6 GHz Barcelona (quad-core) machine. The slides compare these scores to Xeon and dual-core Opteron scores from about 1 month ago, but there are more recent scores available. Below are the most recent scores:

SPECint_rate2006
2 x 2.66 GHz Clovertown: 91.2
2 x 2.6 GHz Barcelona (estimated): 104
2 x 3 GHz Woodcrest: 59.4
2 x 3 GHz dual-core Opteron: 56.6

SPECfp_rate2006
2 x 2.66 GHz Clovertown: 60.9
2 x 2.6 GHz Barcelona (estimated): 92
2 x 3 GHz Woodcrest: 45.5
2 x 3 GHz dual-core Opteron: 52.1

As far as I know, no one has submitted scores for a 2 x 3.0 GHz Clovertown machine.

So what are these benchmarks?

SPEC CPU2006

SPEC CPU2006 is a widely used benchmark suite for high-end (server-type) computer systems. Here is the description of the SPEC suite from the site:

CPU2006 is SPEC's next-generation, industry-standardized, CPU-intensive benchmark suite, stressing a system's processor, memory subsystem and compiler. SPEC designed CPU2006 to provide a comparative measure of compute-intensive performance across the widest practical range of hardware using workloads developed from real user applications. These benchmarks are provided as source code and require the user to be comfortable using compiler commands as well as other commands via a command interpreter using a console or command prompt window in order to generate executable binaries.

int vs fp

The "int" benchmark suite focuses on integer-intensive applications, and "fp" in floating-point applications.

Examples of the applications in the "int" suite are PERL, compression (bzip2, video), C compilation, XML processing, and artificial intelligence.

Examples of the applications in the "fp" suite are quantum chemistry, ray-tracing, and molecular dynamics (protein folding).

rate

Individual SPEC benchmark applications are essentially single-threaded. To test the multi-threaded processing power of a multi-core system, the "rate" benchmark is used. In the rate benchmark, any number of copies of an application can be run simultaneously, and the resulting score is in units of "jobs / time", where a "job" is a completed run of a single application copy. From the SPEC site:

The "rate" calculated for each benchmark is a function of:

the number of copies run *
reference factor for the benchmark /
elapsed time in seconds

which yields a rate in jobs/time. The rate metrics are calculated as a geometric mean from the individual SPECrates using the median result from three runs. As with the speed metric, all copies of the benchmark during each run are required to have validated correctly.

Think of the "rate" benchmarks as measures of performance when all cores are loaded. If a machine is memory-bandwidth constrained, it will show up on this benchmark. Likewise, if a machine has slow cores or fewer cores, it will also show up.

My take

It's pretty obvious why AMD chose these tests: they're widely used and based on real-world applications, and they highlight the greatest strength of AMD's architecture (its memory subsystem). Because the ability to scale is such a big factor in these tests, they don't tell us a lot about the single-threaded performance. Of course, people generally don't buy quad cores for single-threaded performance, but what about Kuma?

If a 2.6 GHz Barcelona is expected to have more than twice the integer performance of a 3.0 GHz dual-core opteron (when they are fully loaded), it seems to me that there's a good chance that Barcelona has a significantly faster core than the K8.

There are a lot of unanswered questions here, perhaps the biggest of which is this: why on earth are they still only giving us estimates???

I know Barcelona is going to support HT3 but anyone know what frequency it will support at launch? I know HT3 supports a maximum of 2600mhz

Quote:

---

Originally Posted by Alxea

I linked to toms hardware when? lol

---

http://farm1.static.flickr.com/207/4...9f518e10da.jpg

tgdaily, was developed from Tom's Hardware. While they go by TGDaily now, they were certainly part of Tom's in the past.

Here's a quote from their website.

Quote:

---

TG Daily has developed out of a small collection of regularly updated "Hardnews" newslinks on the well-known Tom's Hardware Guide website

---

http://www.tgdaily.com/content/view/4/83/

All I ever said was that 2.9GHz was not "confirmed" as you first posted, as I indicated in my reply to your link. This is still the case as your subsequent posts verify that this is only a rumor. Therefore, your first statement that 2.9GHz was confirmed was and still is incorrect as of this time. AMD has confirmed the SSE4 code but nothing about specific operating frequencies other than the ranges previously posted. Yes, AMD may change this in the future, but as of now 2.9GHz is a rumor and not "confirmed". That was my only point.

Last post on this subject, and sorry to everyone for taking this off topic. :slapass:

AMD roadmap for Agena/Kuma (the desktop Barcelona CPUs) clock speeds from this article. Article Link Note the top line FX parts are still only given as ranges, no specific speeds as of yet.

http://images.dailytech.com/nimage/4...stars_list.png

so this folder ? is it the native quadcore?

http://multicore.amd.com/us-en/quadcore/

Yes it is ,tictac.Thats the presentation for the native quad core K10 Phenom/Barcelona MPU.

A Benchmark with K10 B0 Stepping @ 2,4GHz. Screenshots included.

http://www.fudzilla.com/index.php?op...d=982&Itemid=1

50% faster than QX6700 in Integer and 100% faster in floatingpoint.

Fake: http://www.sweclockers.com/forum/sho...hreadid=680996

http://aycu28.webshots.com/image/177...0283099_rs.jpg

All I can say is (intel is dead in the water. 2.4ghz K10 75% FPU over Intel C2Q 2.66 and 40% ALU.) Damn. A benchmark for youalz oldtimzerz.

Quote:

---

Originally Posted by w0mbat

Fake: http://www.sweclockers.com/forum/sho...hreadid=680996

---

The only thinj thats fake is u :rolleyes:

Alxea, you do know thats the fake screenshots the swedish guy made?

Quote:

---

Originally Posted by Alxea

The only thinj thats fake is u :rolleyes:

---

http://www.dailytech.com/article.aspx?newsid=7291

Quote:

---

False benchmarks published by Fuad Abazovic
Why just FUD when you can FUDZilla!


Former Inquirer writer Fuad Abazovic released benchmarks of the upcoming AMD quad-core Barcelona. Abazovic believes his benchmarks are "an AMD exclusive" of a new processor he has in his possession.

Unfortunately, not only is this tale completely fabricated, but the perpetrators who architected the hoax came forward early this morning on Swedish enthusiast forum, SweClockers.com.

The poster "2good4you" details the process he used to convince Abazovic that his benchmarks were authentic -- which essentially entailed a single email with the image attached. Abazovic then republished the benchmarks as his own, and not for the first time alludes "2good4you."

Rest assured, real Barcelona benchmarks will arrive shortly. AMD typically releases two spins of silicon before retail samples, EVT followed by DVT. As of May 2007, several EVT samples are floating around, though it's difficult to predict the actual retail performance of K10 via these engineering samples. AMD still had serious bugs in its EVT samples before AM2 launched, and the only real change presented on that architecture was the DDR2 memory controller -- for Barcelona we're talking about a dramatically new design.

With Computex coming up in less than a month, it's fairly reasonable to expect a flood of authentic benchmarks to make waves just after or during the show. And even then, don't be so quick to believe the Fear, Uncertainty, and Doubt until we start seeing DVT and Retail samples.

Update 05/15/2007: As of noon today, Abazovic removed the article without an official statement. Abazovic has emailed DailyTech with the following response:

Next time please send me an email and ask me for an explanation will make you more processional

I still stand by this one
http://www.fudzilla.com/.....620&Itemid=35

---

:rolleyes:

Quote:

---

Originally Posted by Alxea

http://aycu28.webshots.com/image/177...0283099_rs.jpg

All I can say is (intel is dead in the water. 2.4ghz K10 75% FPU over Intel C2Q 2.66 and 40% ALU.) Damn. A benchmark for youalz oldtimzerz.

---

This is fake man, look at the CPU-Z, the Selection in the bottom, it is selectable which mean that was s dual socket system(4*2 core??)

K8L/K10/Phenom/ 10h family... whatever

here the software optimization guide...
http://www.amd.com/us-en/assets/cont...docs/40546.pdf

now have to wait for bios optimization guide... :(

don't they release their software/bios optimization guides to insiders first? (those who need them like software devs or mb manufacturers?)
I just wonder why you/we need this bios optimization guide? Ain't it only for writing a new bios?

Quote:

---

Originally Posted by Alxea
The only thinj thats fake is u

---

Quote:

---

Originally Posted by Clairvoyant129
http://www.dailytech.com/article.aspx?newsid=7291

:rolleyes:

---

:owned:

CPUZ can't even read the CPU properly if you didn't notice. Anything could be a error however its not fake, its just a 8 core dual socket system. if it is well then even 2P systems are in serious trouble from intel. Cuz those benchmarks are just owning intel by 50%/75% in all multimedia. And if you didn't know Intel descales with every socket added comepared to AMD that scales with more sockets.

I would hate to see 8 cores trying to run with 133mhz FSB. EWWWWWW!!!! Nothing beats hyper transport when its twice the speed and in the new systems its 3 to 4 times faster then FSB. K8 could never stress the links or use them to their full advantage like K10 can. That still gives AMD a serious Cycles per clock advantage at the same speed. It would still be a nearly 50% lead like they clam if you take way the 2nd chip.

Someone has openly admitted faking that screenshot yet you still claim it's real? :rolleyes:

As for a 50% advantage, I'd really like to see that. People with ESs around here are saying it's more like 10% on average.

Let it go Alxea, the swede who made the screenshots has prooved in the sweclockers forums that he had them before Fuddy-boy posted them with his own logo all over them.