TechReport on Barcelona

[JumpingJack]

Nice comprehensive review.... he includes top bin Intel, low voltage 2.0 Ghz Intel, as well as 2.0 and 2.5 Ghz Barcey in the data set.

http://techreport.com/articles.x/13176/1

Interesting points.

• The SSE performance, big improvement for Barcey but still way way behind Intel's throughput.
  
• Cache latency is higher for Barcey overall, driven by the L3 cache.
  
• The cinebench data validates the Coolaler's data leaked a week or so ago.

This sorta captures the summary:

Nonetheless, AMD now faces some harsh realities. For one, it is not going to capture the overall performance lead from Intel soon, not even in "Q4," which is when higher-clocked parts like the Opteron 2360 SE are expected to arrive. Given what we've seen, AMD will probably have to achieve something close to clock speed parity with Intel in order to compete for the performance crown.

[Carfax]

Awesome!

Tech Report does the best CPU reviews

Seriously, I'm wondering whether the CPUs are still buggy as the floating point performance isn't what I expected it to be.

In some SSE/FP oriented benches, Barcy is barely beating it's predecessor.

[Shintai]

Reality bites again. No miracle chip. How...unexpected...

[Thesavage]

Actually better than most naysayers where predicting.

[Shintai]

Actually better than most naysayers where predicting.

It is? Most "intel fanboys" said around 65nm Core 2 performance. +0/-5% and it is.

[gallag]

There has been that many people about here saying they have seen k10 first hand and how amazed we were going to be, WTF , only I don't have the time I would trawl through the numerous misleading posts and make a thread to hold these people to account.

[Thesavage]

No a fair number said no improvement over K8 at all so?

[Shintai]

No a fair number said no improvement over K8 at all so?

Show me. because thats not what I ahve seen and read. Everything has gone froma round Core 2 performance to 170% faster.

[Nedjo]

Awesome!

Tech Report does the best CPU reviews

Hope that sarcasm!
They have ONLY ONE server benchmark (SpecJBB), and bunch of pointless and Intel optimized benchmarks (picCOLOR, MyriMatch, Stars Euler...)

For Crying out loud here is the comment for MyriMatch results:

"This application looks to be limited by memory bandwidth or some similar factor; scaling beyond four threads doesn't work out well on any of these systems"
Hi doesn't have bloody idea what does this benchmark doing?!?!
I know couple of people who are developing similar science projects that are optimizing for one or the other CPU architecture, depending what is stacked in their racks…
And what's more important TechReport doesn't have SINGLE ONE Virtualization benchmark!!! That much about professional benchmarking approach of server CPUs!

[Thesavage]

You should really look on more forums Shintai, yes the fanboys on both sides made totally absurd claims but on the whole i think its fair to say it has exceeded the naysayers claims of doom and death.

[Shintai]

So nothing on XS I assume...

[accord99]

You should really look on more forums Shintai, yes the fanboys on both sides made totally absurd claims but on the whole i think its fair to say it has exceeded the naysayers claims of doom and death.

But it's also fair to say it has fallen well short of supporters claims of being 40% better than C2D clock-per-clock.

[LordEC911]

Reality bites again. No miracle chip. How...unexpected...

Right...
No one was expecting a miracle chip.
This is pretty much what everyone was expecting.

But it's also fair to say it has fallen well short of supporters claims of being 40% better than C2D clock-per-clock.

In certain benchmarks.
We also haven't seen the higher clocked chips yet either.

[Thesavage]

But it's also fair to say it has fallen well short of supporters claims of being 40% better than C2D clock-per-clock.

Correct.

[Shintai]

Right...
No one was expecting a miracle chip.
This is pretty much what everyone was expecting.


In certain benchmarks.
We also haven't seen the higher clocked chips yet either.

Only the realists. The fans have claimed both 30% to 170% faster than Core 2.

[BeardyMan]

So nothing on XS I assume...

Barcelona actually owns your level of playing field.

[Thesavage]

Would like to see how Phenom performs on its own platform and ddr2 800+, should be slightly better than anands review. +-7%

[LordEC911]

Only the realists. The fans have claimed both 30% to 170% faster than Core 2.

I would love to see a quote of any somewhat knowledgeable person who was being serious about an increase greater than ~40-70% over C2Q.
I'm guessing some noob in the 30k 3dMark06 K10 thread said the 170%.

I was expecting about equal to C2Q, hoping for a 10-20% advantage but we still have yet to see Phenom on an AM2+ board. Still won't push Phenom way ahead of C2Q but probably equal, possibly slightly ahead.

I'm still holding out to see how Phenom OCs and if Penryn is worth it over a Q6600.

[metro.cl]

I would love to see a quote of any somewhat knowledgeable person who was being serious about an increase greater than ~40-70% over C2Q.
I'm guessing some noob in the 30k 3dMark06 K10 thread said the 170%.

I was expecting about equal to C2Q, hoping for a 10-20% advantage but we still have yet to see Phenom on an AM2+ board. Still won't push Phenom way ahead of C2Q but probably equal, possibly slightly ahead.

I'm still holding out to see how Phenom OCs and if Penryn is worth it over a Q6600.

Someguy form Anand i believe said those numbers.

About OC on Anand they have a cpu overclocked to 2.5 and they need 1.5v for that.

My C2D E6850 does 4.0GHz at same volts so overclock doesnt seem good for AMD.

[WoD]

Are there already any information out about the overclocking performance???


Edit:
Woops, yet read the post just above mine...


So the new AMD CPU will not easily hit the 3Ghz on air like the CPU which was shown some weeks ago?

[LordEC911]

Someguy form Anand i believe said those numbers.

About OC on Anand they have a cpu overclocked to 2.5 and they need 1.5v for that.

My C2D E6850 does 4.0GHz at same volts so overclock doesnt seem good for AMD.

But a C2Q would use ~200w with those voltages, not the ~100w that Barc does.

I also don't believe it was overclocked, no more then any other factory clocked CPU.
AMD had wrote 2.5ghz on the sample and it had different markings then the other chips.

[flopper]

we know to little yet.

[gallag]

But a C2Q would use ~200w with those voltages, not the ~100w that Barc does.

?is this even true? what info did you base this on?

[Hornet331]

Hope that sarcasm!
They have ONLY ONE server benchmark (SpecJBB), and bunch of pointless and Intel optimized benchmarks (picCOLOR, MyriMatch, Stars Euler...)

For Crying out loud here is the comment for MyriMatch results:

"This application looks to be limited by memory bandwidth or some similar factor; scaling beyond four threads doesn't work out well on any of these systems"
Hi doesn't have bloody idea what does this benchmark doing?!?!
I know couple of people who are developing similar science projects that are optimizing for one or the other CPU architecture, depending what is stacked in their racks…
And what's more important TechReport doesn't have SINGLE ONE Virtualization benchmark!!! That much about professional benchmarking approach of server CPUs!

go here if you want server only benches.

[Nedjo]

Someguy form Anand i believe said those numbers.

About OC on Anand they have a cpu overclocked to 2.5 and they need 1.5v for that.

They've received 2.5 GHz! How do you think they could have O/C this CPU on the server mobo?
BTW CPU-Z can't detect good voltage of Barcelona...