K10 Barcelona 50% faster than Kentsfield ?

[JackPack]

Originally posted by: NoobyDoo
AMD press release at techpowerup.com

AMD also disclosed updated performance projections for its upcoming native Quad-Core AMD Opteron? processors, code-named ?Barcelona.? The new Barcelona projections are based on the latest SPECcpu2006 benchmarks and show that AMD expects to have up to a 50 percent advantage in floating point performance and 20 percent in integer performance over the competition?s highest-performing quad-core processor at the same frequency.

1. Projections? Which means no real silicon? They don't even use the word "estimated" which is official terminology for SPEC.

2. Highest-performing? At the same frequency? In other words, if Barcelona only clocks up to 2.5 GHz, this so-called advantage is compared to a Clovertown at 2.5 GHz. Yawn.

 

[imported_UncivilizedAMD]

Originally posted by: rexian96

Originally posted by: MDme
some rumors....(which could be facts)

1. Agena FX (Quad core) can OC to 3.05 Ghz
2. It's TDP is 95w
3. The QX6800 TDP is 130w (intel - typical TDP) or estimated by AMD at 150w (max TDP)

links:
http://www.fudzilla.com/index.php?option=com_content&task=view&id=663&Itemid=1
http://www.fudzilla.com/index.php?option=com_content&task=view&id=658&Itemid=1
http://www.fudzilla.com/index.php?option=com_content&task=view&id=649&Itemid=1
http://www.fudzilla.com/index.php?option=com_content&task=view&id=631&Itemid=1
http://www.fudzilla.com/index.php?option=com_content&task=view&id=470&Itemid=35

If the performance of Barcelona/Agena is really +50% in FP and +20% in integer then intel would need around 3+ Ghz of Conroe to match it (a little bit less if penryn is used for comparison)

The 50% FP and 20% int advantage means that it will outperform it clock-for-clock by at least 20% based on code used. 2.5Ghz Agena=3.0Ghz C2D in int and 3.7Ghz C2D in FP.

Don't forget that Kuma Dual cores (which will launch at 2.0-2.9Ghz) should scale to a higher frequency if the Agena is really able to OC to 3.05Ghz.

Just imagine a 3.2 Ghz Kuma DC....

*disclaimer: these statements assume that item #1-3 are accurate. THe poster does not assume responsibility for the content of the links.

Sounds very interesting. I wonder if these AM2+ CPUs can be overclocked with an AM2 board ? I guess it's too early expect a definitive answer They didn't mention if they used a future board or one that's available in the market when they overclocked that 1.9GHz Agena.

The overclock is fake!!!

Link

Link 2

 

[Viditor]

Originally posted by: JackPack

Originally posted by: MDme
some rumors....(which could be facts)

1. Agena FX (Quad core) can OC to 3.05 Ghz
2. It's TDP is 95w
3. The QX6800 TDP is 130w (intel - typical TDP) or estimated by AMD at 150w (max TDP)

1. The 3 GHz Agena OC screenshot has been proven to be fake. CPU-Z doesn't look for an L3 on the K8, according to the author of CPU-Z.

2. TDP of Agena FX is 120W. Agena might be 95W, but >2.5 GHz is 120W.

3. TDP of Intel QX series is 130W. Refer to reputable sites such as LostCircuits where they have already measured the QX6700 to be 115W.

1. It has not been "proved" to be anything...cpuz is easily modified and those who are assigned to test and write new bioses (for example the type of engineer who would actually have an ES) do this on a regular basis as part of their job. CPUZ gets it's info from the bios...
If an engineer is testing a new bios, you can imagine that he would want to modify CPUZ to actually gather all of the info he needs.
2. AFAIK, standard Agena is still at 95w TDP...even when it's over 2.5 GHz. The TDP number is for the entire line of non-F/X chips.
3. Just as you say, the TDP number is never really reflective of actual power used...however this is even more true for AMD. So while I believe what you say about the QX, I imagine that the Agena will be functioning well below 95w as well in the real world.

 

[hardwareking]

barcelona may well be faster than intel processors(conroe and penryn)
but the crucial factor will be price
with intel making quad core mainstream in Q3 with the $266 Q6600,u have to wonder, does AMD have anything to offer in that region while having some sort of profit margin?

 

[Viditor]

Originally posted by: hardwareking
barcelona may well be faster than intel processors(conroe and penryn)
but the crucial factor will be price
with intel making quad core mainstream in Q3 with the $266 Q6600,u have to wonder, does AMD have anything to offer in that region while having some sort of profit margin?

Actually, for Barcelona price really isn't an issue. Remember that Barcelona is a server-only product, and the price of the CPU is fairly unimportant (relatively speaking).

For the desktop market, it's different (though still not clear cut yet). The important points there are:

1. While quad core desktop sounds cool, there are still very few applications for it. Therefore, it is necessary for Intel and AMD to market chips for which there is not yet a need so that the software community can be enticed to create that need ("if you build it, they will come..."). This is important for both companies because (as Intel found out) there is a point of diminishing returns in growing the power of individual cores, so that the cheapest and easiest way to grow is by adding cores.

2. We should definately expect Intel to be the low cost leader (at least for this year) on desktop quadcore chips...remember that there is very little profit in them and that the major reason for selling them is to build critical mass in quadcore consumer so that the software developers can afford to spend on development for them. Intel can afford this kind of marketing investment far better than AMD can, and I expect AMD will ride Intel's coattails for awhile.

3. One last point...if what we've read actually comes to fruition, AMD will probably have a performance edge on Intel in quad core because of the FSB bottleneck. But again, that point is moot as it will continue to be a fairly small and unprofitable market for the consumer desktop space.

 

[Stoneburner]

okay, when the hell is it dropping? I'm getting tired of waiting.

 

[f4phantom2500]

Originally posted by: destrekor
Hopefully Intel redesigns the marchitecture for their quad-cores, Kentsfield is definitely not the right way to go about it. 2 chips sammiched together? boo

That's how they did the Pentium D lineup. Basically, take 2 of your current processors, throw them on the same chip, call it a (double number of cores in each processor) core processor, then work on making a native design with that number of cores, release it, then later on stick 2 together...that's probably how they're gonna do it since it means they can have chips with more cores (albeit just 2 stuck together) in mass production faster.

 

[coldpower27]

Originally posted by: Viditor

Originally posted by: JackPack

Originally posted by: MDme
some rumors....(which could be facts)

1. Agena FX (Quad core) can OC to 3.05 Ghz
2. It's TDP is 95w
3. The QX6800 TDP is 130w (intel - typical TDP) or estimated by AMD at 150w (max TDP)

1. The 3 GHz Agena OC screenshot has been proven to be fake. CPU-Z doesn't look for an L3 on the K8, according to the author of CPU-Z.

2. TDP of Agena FX is 120W. Agena might be 95W, but >2.5 GHz is 120W.

3. TDP of Intel QX series is 130W. Refer to reputable sites such as LostCircuits where they have already measured the QX6700 to be 115W.

1. It has not been "proved" to be anything...cpuz is easily modified and those who are assigned to test and write new bioses (for example the type of engineer who would actually have an ES) do this on a regular basis as part of their job. CPUZ gets it's info from the bios...
If an engineer is testing a new bios, you can imagine that he would want to modify CPUZ to actually gather all of the info he needs.
2. AFAIK, standard Agena is still at 95w TDP...even when it's over 2.5 GHz. The TDP number is for the entire line of non-F/X chips.
3. Just as you say, the TDP number is never really reflective of actual power used...however this is even more true for AMD. So while I believe what you say about the QX, I imagine that the Agena will be functioning well below 95w as well in the real world.

Source for Agena's 95W TDP across the lineup, as there are 125W TDP processors in AMD's current performance lineup.

If you look at Santa Rosa vs Windsor their TDP's are pretty synchronized, 120W for Opteron 2222 SE vs Athlon 64x2 6000+ 125W,

If anything I expect AMD to launch 65W, 89W and 125W versions overtime.

Bareclona's TDP is 95W to 2.3GHZ from what we know so far, and 120W to 2.5GHZ.

 

[coldpower27]

Originally posted by: destrekor
Hopefully Intel redesigns the marchitecture for their quad-cores, Kentsfield is definitely not the right way to go about it. 2 chips sammiched together? boo

Not the right way to go about, it's a great way as it gives you Quad Core out in the market that much quicker allowing developers to start programing for them in anticipation of better models in the future.

 

[coldpower27]

Originally posted by: hardwareking
barcelona may well be faster than intel processors(conroe and penryn)
but the crucial factor will be price
with intel making quad core mainstream in Q3 with the $266 Q6600,u have to wonder, does AMD have anything to offer in that region while having some sort of profit margin?

I expect Barcelona to be faster then it's supposed competitors Clovetown and Tigerton in the server range.

Budapest and Agena will have a lot tougher time though going against Kentsfield, where FSB isn't nearly as much of an issue and it won't last very long as by next quarter Yorkfield will be released since Budapest and Agena are Q4 products.

With a 283mm2 die and the need to start recouping R&D, as well as their constrained capacity we don't know for sure, if they will try to trump Intel in the lower end Quad Core space.

Intel is more fortunate with their large amounts of 65nm capacity and their MCM approach they can adjust their Quad Core mix as market requires more easily.

 

[Viditor]

Originally posted by: coldpower27

Source for Agena's 95W TDP across the lineup, as there are 125W TDP processors in AMD's current performance lineup.

You mean in their 90nm lineup without the enhanced power system? Yes there are, but I don't see your point...

If you look at Santa Rosa vs Windsor their TDP's are pretty synchronized, 120W for Opteron 2222 SE vs Athlon 64x2 6000+ 125W,

If anything I expect AMD to launch 65W, 89W and 125W versions overtime.

Bareclona's TDP is 95W to 2.3GHZ from what we know so far, and 120W to 2.5GHZ.

Keeping in mind that TDP is not a measurement but a guideline to OEMs for system building, AMD has been very consistent about having a single number for each line of processers. The one exception is the F/X line which is the only unlocked desktop processer and is built specifically for enthusiasts...

 

[liebremx]

Originally posted by: Viditor

1. It has not been "proved" to be anything...cpuz is easily modified and those who are assigned to test and write new bioses (for example the type of engineer who would actually have an ES) do this on a regular basis as part of their job. CPUZ gets it's info from the bios...

Viditor, that assertion is incorrect. CPU-Z relies mostly on special CPU instructions (cpuid), reading chipset registers and generally collecting information that is available if you know were to look, like the DIMM/SDRAM info stored in the SPD memories.
It might read some info from BIOS-provided data structures like the SMBIOS or ACPI tables but most of the reported CPU info should come directly from queries to the CPU itself via cpuid and/or reading MSR's contents AND also from MCH/IOH registers in Intel systems or from the IMC/IOH registers in AMD's case.

 

[Viditor]

Originally posted by: liebremx

Originally posted by: Viditor

1. It has not been "proved" to be anything...cpuz is easily modified and those who are assigned to test and write new bioses (for example the type of engineer who would actually have an ES) do this on a regular basis as part of their job. CPUZ gets it's info from the bios...

Viditor, that assertion is incorrect. CPU-Z relies mostly on special CPU instructions (cpuid), reading chipset registers and generally collecting information that is available if you know were to look, like the DIMM/SDRAM info stored in the SPD memories.
It might read some info from BIOS-provided data structures like the SMBIOS or ACPI tables but most of the reported CPU info should come directly from queries to the CPU itself via cpuid and/or reading MSR's contents AND also from MCH/IOH registers in Intel systems or from the IMC/IOH registers in AMD's case.

First off, I want to thank you for the post liebremx...because of you, I now know a heck of a lot more about the way CPUZ works, and I truly appreciate that!
That said, it still doesn't invalidate my assertion...any engineer working on developing a bios would require tools to monitor the same things that CPU-Z does under windows, and patching CPU-Z for the purpose really isn't that difficult for a competent software engineer (it's also smarter than showing a screenshot of a custom tool when you're under an NDA).

 

[coldpower27]

Originally posted by: Viditor

Originally posted by: coldpower27

Source for Agena's 95W TDP across the lineup, as there are 125W TDP processors in AMD's current performance lineup.

You mean in their 90nm lineup without the enhanced power system? Yes there are, but I don't see your point...

If you look at Santa Rosa vs Windsor their TDP's are pretty synchronized, 120W for Opteron 2222 SE vs Athlon 64x2 6000+ 125W,

If anything I expect AMD to launch 65W, 89W and 125W versions overtime.

Bareclona's TDP is 95W to 2.3GHZ from what we know so far, and 120W to 2.5GHZ.

Keeping in mind that TDP is not a measurement but a guideline to OEMs for system building, AMD has been very consistent about having a single number for each line of processers. The one exception is the F/X line which is the only unlocked desktop processer and is built specifically for enthusiasts...

Intel has the same model as well 1 65W TDP for all the E4x00 and E6xx0 lineup, with the exception of the XE and the last time I checked the Quad FX all have 125W TDP's, unless you can source otherwise. Both parties do this, I am quite aware of it.

This doesn't change the fact that both the desktop and server lines have TDP segments, 65W, 89W and 125W desktop 68W, 95W and 120W, server.

My point is that unless you can prove otherwise, the current TDP model pretty much stands. So since you have no source I assume for Agena's 95W TDP at all across the board, so I have to go with what is currently available.

No but they give you a very good idea of what the power level is, and with the 6000+ it's pretty much in line with Intel's TDP even with the assumption that the chipset is drawing a little more power on the AMD side. The 4x4 FX platform is absolutely ridiculous on that front.

http://techreport.com/reviews/2007q1/quad-core/index.x?pg=12 I dunno how your going to account for 216W, at load, in Cinebench.

You also seem to conveniently forget about the 6000+ which I stated has a 125W TDP, and is a non-FX processor. And is actually quite close to it's stated TDP.

Even though K8L/K10 will have better power management features, they don't come into play when your talking about at full load.

TDP's between the Core 2 Duo's and the Athlon 64x2 have been pretty even in terms of comparing, they are exactly where you expect them to be.

From what I see so far, I pretty much expect the server and desktop TDP to be in sync for a given clock frequency.

 

[Viditor]

Originally posted by: coldpower27

Originally posted by: Viditor

Originally posted by: coldpower27

Source for Agena's 95W TDP across the lineup, as there are 125W TDP processors in AMD's current performance lineup.

You mean in their 90nm lineup without the enhanced power system? Yes there are, but I don't see your point...

If you look at Santa Rosa vs Windsor their TDP's are pretty synchronized, 120W for Opteron 2222 SE vs Athlon 64x2 6000+ 125W,

If anything I expect AMD to launch 65W, 89W and 125W versions overtime.

Bareclona's TDP is 95W to 2.3GHZ from what we know so far, and 120W to 2.5GHZ.

Keeping in mind that TDP is not a measurement but a guideline to OEMs for system building, AMD has been very consistent about having a single number for each line of processers. The one exception is the F/X line which is the only unlocked desktop processer and is built specifically for enthusiasts...

Intel has the same model as well 1 65W TDP for all the E4x00 and E6xx0 lineup, with the exception of the XE and the last time I checked the Quad FX all have 125W TDP's, unless you can source otherwise. Both parties do this, I am quite aware of it.

This doesn't change the fact that both the desktop and server lines have TDP segments, 65W, 89W and 125W desktop 68W, 95W and 120W, server.

My point is that unless you can prove otherwise, the current TDP model pretty much stands. So since you have no source I assume for Agena's 95W TDP at all across the board, so I have to go with what is currently available.

No but they give you a very good idea of what the power level is, and with the 6000+ it's pretty much in line with Intel's TDP even with the assumption that the chipset is drawing a little more power on the AMD side. The 4x4 FX platform is absolutely ridiculous on that front.

http://techreport.com/reviews/2007q1/quad-core/index.x?pg=12 I dunno how your going to account for 216W, at load, in Cinebench.

You also seem to conveniently forget about the 6000+ which I stated has a 125W TDP, and is a non-FX processor. And is actually quite close to it's stated TDP.

Even though K8L/K10 will have better power management features, they don't come into play when your talking about at full load.

TDP's between the Core 2 Duo's and the Athlon 64x2 have been pretty even in terms of comparing, they are exactly where you expect them to be.

From what I see so far, I pretty much expect the server and desktop TDP to be in sync for a given clock frequency.

Of things that are published, the guys at Fudzilla have it pretty much correct...

Text1
"AMD's upcoming Agena FX and Barcelona native quad core CPU's will consume less than 95 Watts"

They also did some sisoft sandra numbers...

Text2
"In Sandra multimedia benchmark Agena / Barcelona K10 2.5 GHz scores over 400.000 it/s (instructions / second) in integer test and about 300.000 it/s in floating point test. Yeah there are some Agena / Barcelona parts runing at 2.5 GHz at press time and we got the chance to share the scores with the world.

Intel's fastest QX6800 Quad core at 3 GHz scores about ~300.000 it/s in integer and ~180.000 it/s in floating point test.

So at least in multimedia stuff and this synthetic benchmark AMD's native quad core is 33.3 percent faster than Intel's fastest Quad core, while in the floating calculations it is a massive 66 percent faster.

The QX6850 with FSB 1333 might improve the situation a bit, but it won't be enough save the day. Intel needs to go much higher than 3 GHZ to match the 2.5 GHz K10 part"

 

[IEC]

Originally posted by: Viditor

Originally posted by: coldpower27

Originally posted by: Viditor

Originally posted by: coldpower27

Source for Agena's 95W TDP across the lineup, as there are 125W TDP processors in AMD's current performance lineup.

You mean in their 90nm lineup without the enhanced power system? Yes there are, but I don't see your point...

If you look at Santa Rosa vs Windsor their TDP's are pretty synchronized, 120W for Opteron 2222 SE vs Athlon 64x2 6000+ 125W,

If anything I expect AMD to launch 65W, 89W and 125W versions overtime.

Bareclona's TDP is 95W to 2.3GHZ from what we know so far, and 120W to 2.5GHZ.

Keeping in mind that TDP is not a measurement but a guideline to OEMs for system building, AMD has been very consistent about having a single number for each line of processers. The one exception is the F/X line which is the only unlocked desktop processer and is built specifically for enthusiasts...

Intel has the same model as well 1 65W TDP for all the E4x00 and E6xx0 lineup, with the exception of the XE and the last time I checked the Quad FX all have 125W TDP's, unless you can source otherwise. Both parties do this, I am quite aware of it.

This doesn't change the fact that both the desktop and server lines have TDP segments, 65W, 89W and 125W desktop 68W, 95W and 120W, server.

My point is that unless you can prove otherwise, the current TDP model pretty much stands. So since you have no source I assume for Agena's 95W TDP at all across the board, so I have to go with what is currently available.

No but they give you a very good idea of what the power level is, and with the 6000+ it's pretty much in line with Intel's TDP even with the assumption that the chipset is drawing a little more power on the AMD side. The 4x4 FX platform is absolutely ridiculous on that front.

http://techreport.com/reviews/2007q1/quad-core/index.x?pg=12 I dunno how your going to account for 216W, at load, in Cinebench.

You also seem to conveniently forget about the 6000+ which I stated has a 125W TDP, and is a non-FX processor. And is actually quite close to it's stated TDP.

Even though K8L/K10 will have better power management features, they don't come into play when your talking about at full load.

TDP's between the Core 2 Duo's and the Athlon 64x2 have been pretty even in terms of comparing, they are exactly where you expect them to be.

From what I see so far, I pretty much expect the server and desktop TDP to be in sync for a given clock frequency.

Of things that are published, the guys at Fudzilla have it pretty much correct...

Text1
"AMD's upcoming Agena FX and Barcelona native quad core CPU's will consume less than 95 Watts"

They also did some sisoft sandra numbers...

Text2
"In Sandra multimedia benchmark Agena / Barcelona K10 2.5 GHz scores over 400.000 it/s (instructions / second) in integer test and about 300.000 it/s in floating point test. Yeah there are some Agena / Barcelona parts runing at 2.5 GHz at press time and we got the chance to share the scores with the world.

Intel's fastest QX6800 Quad core at 3 GHz scores about ~300.000 it/s in integer and ~180.000 it/s in floating point test.

So at least in multimedia stuff and this synthetic benchmark AMD's native quad core is 33.3 percent faster than Intel's fastest Quad core, while in the floating calculations it is a massive 66 percent faster.

The QX6850 with FSB 1333 might improve the situation a bit, but it won't be enough save the day. Intel needs to go much higher than 3 GHZ to match the 2.5 GHz K10 part"

Very good information. I believe that with Barcelona AMD will be faster almost across the board in terms of performance, while R600's massive shader capability will help it in DX10 games. AMD has been stronger than Intel in FP calc for a long time, so for many applications AMD will still be the better choice. I'm glad I didn't go C2D yet, because K10 is looking good.

 

[liebremx]

Originally posted by: Viditor
First off, I want to thank you for the post liebremx...because of you, I now know a heck of a lot more about the way CPUZ works, and I truly appreciate that!
That said, it still doesn't invalidate my assertion...any engineer working on developing a bios would require tools to monitor the same things that CPU-Z does under windows, and patching CPU-Z for the purpose really isn't that difficult for a competent software engineer (it's also smarter than showing a screenshot of a custom tool when you're under an NDA).

My pleasure.

Now could you clarify your point about CPU-Z? If I understand correctly you are saying that the CPU-Z screenshot of the OC can't be proved to be a fake because the person that ran CPU-Z could have modified it to support the new AMD processor?

So if that's what you are saying, well, I don't think CPU-Z's source code is available is it?

So I'd say is quite possible that you have reliable sources that backup the numbers in the screenshot for a running system w/Barcelona therefore making you, perhaps unknowingly, defend the non-fakeness of the screenshot more than it would be otherwise :beer: Oh, the joys of speculation.

 

[Phynaz]

Originally posted by: Viditor

Originally posted by: coldpower27

Originally posted by: Viditor

Originally posted by: coldpower27

Source for Agena's 95W TDP across the lineup, as there are 125W TDP processors in AMD's current performance lineup.

You mean in their 90nm lineup without the enhanced power system? Yes there are, but I don't see your point...

If you look at Santa Rosa vs Windsor their TDP's are pretty synchronized, 120W for Opteron 2222 SE vs Athlon 64x2 6000+ 125W,

If anything I expect AMD to launch 65W, 89W and 125W versions overtime.

Bareclona's TDP is 95W to 2.3GHZ from what we know so far, and 120W to 2.5GHZ.

Keeping in mind that TDP is not a measurement but a guideline to OEMs for system building, AMD has been very consistent about having a single number for each line of processers. The one exception is the F/X line which is the only unlocked desktop processer and is built specifically for enthusiasts...

Intel has the same model as well 1 65W TDP for all the E4x00 and E6xx0 lineup, with the exception of the XE and the last time I checked the Quad FX all have 125W TDP's, unless you can source otherwise. Both parties do this, I am quite aware of it.

This doesn't change the fact that both the desktop and server lines have TDP segments, 65W, 89W and 125W desktop 68W, 95W and 120W, server.

My point is that unless you can prove otherwise, the current TDP model pretty much stands. So since you have no source I assume for Agena's 95W TDP at all across the board, so I have to go with what is currently available.

No but they give you a very good idea of what the power level is, and with the 6000+ it's pretty much in line with Intel's TDP even with the assumption that the chipset is drawing a little more power on the AMD side. The 4x4 FX platform is absolutely ridiculous on that front.

http://techreport.com/reviews/2007q1/quad-core/index.x?pg=12 I dunno how your going to account for 216W, at load, in Cinebench.

You also seem to conveniently forget about the 6000+ which I stated has a 125W TDP, and is a non-FX processor. And is actually quite close to it's stated TDP.

Even though K8L/K10 will have better power management features, they don't come into play when your talking about at full load.

TDP's between the Core 2 Duo's and the Athlon 64x2 have been pretty even in terms of comparing, they are exactly where you expect them to be.

From what I see so far, I pretty much expect the server and desktop TDP to be in sync for a given clock frequency.

Of things that are published, the guys at Fudzilla have it pretty much correct...

Text1
"AMD's upcoming Agena FX and Barcelona native quad core CPU's will consume less than 95 Watts"

They also did some sisoft sandra numbers...

Text2
"In Sandra multimedia benchmark Agena / Barcelona K10 2.5 GHz scores over 400.000 it/s (instructions / second) in integer test and about 300.000 it/s in floating point test. Yeah there are some Agena / Barcelona parts runing at 2.5 GHz at press time and we got the chance to share the scores with the world.

Intel's fastest QX6800 Quad core at 3 GHz scores about ~300.000 it/s in integer and ~180.000 it/s in floating point test.

So at least in multimedia stuff and this synthetic benchmark AMD's native quad core is 33.3 percent faster than Intel's fastest Quad core, while in the floating calculations it is a massive 66 percent faster.

The QX6850 with FSB 1333 might improve the situation a bit, but it won't be enough save the day. Intel needs to go much higher than 3 GHZ to match the 2.5 GHz K10 part"

Wow, you've sunk to quoting Inq rejects now.

That's really sad.

 

[Viditor]

Originally posted by: liebremx

Originally posted by: Viditor
First off, I want to thank you for the post liebremx...because of you, I now know a heck of a lot more about the way CPUZ works, and I truly appreciate that!
That said, it still doesn't invalidate my assertion...any engineer working on developing a bios would require tools to monitor the same things that CPU-Z does under windows, and patching CPU-Z for the purpose really isn't that difficult for a competent software engineer (it's also smarter than showing a screenshot of a custom tool when you're under an NDA).

My pleasure.

Now could you clarify your point about CPU-Z? If I understand correctly you are saying that the CPU-Z screenshot of the OC can't be proved to be a fake because the person that ran CPU-Z could have modified it to support the new AMD processor?

So if that's what you are saying, well, I don't think CPU-Z's source code is available is it?

So I'd say is quite possible that you have reliable sources that backup the numbers in the screenshot for a running system w/Barcelona therefore making you, perhaps unknowingly, defend the non-fakeness of the screenshot more than it would be otherwise :beer: Oh, the joys of speculation.

I do have reliable sources that back up the data, but I also know a software engineer that does exactly what I described. I don't know if/how he got the source code (maybe he licensed it or develops for them), but he's the one that first explained this to me...
His suspicion is that whoever leaked the shots is an engineer for one of the mobo manufacturers.
To be clear, I'm not saying that the shot isn't a fake, I'm saying that it might not be...not knowing where it came from, I can (as you say) only speculate.

 

[coldpower27]

Originally posted by: Viditor

Originally posted by: coldpower27

Originally posted by: Viditor

Originally posted by: coldpower27

Source for Agena's 95W TDP across the lineup, as there are 125W TDP processors in AMD's current performance lineup.

You mean in their 90nm lineup without the enhanced power system? Yes there are, but I don't see your point...

If you look at Santa Rosa vs Windsor their TDP's are pretty synchronized, 120W for Opteron 2222 SE vs Athlon 64x2 6000+ 125W,

If anything I expect AMD to launch 65W, 89W and 125W versions overtime.

Bareclona's TDP is 95W to 2.3GHZ from what we know so far, and 120W to 2.5GHZ.

Keeping in mind that TDP is not a measurement but a guideline to OEMs for system building, AMD has been very consistent about having a single number for each line of processers. The one exception is the F/X line which is the only unlocked desktop processer and is built specifically for enthusiasts...

Intel has the same model as well 1 65W TDP for all the E4x00 and E6xx0 lineup, with the exception of the XE and the last time I checked the Quad FX all have 125W TDP's, unless you can source otherwise. Both parties do this, I am quite aware of it.

This doesn't change the fact that both the desktop and server lines have TDP segments, 65W, 89W and 125W desktop 68W, 95W and 120W, server.

My point is that unless you can prove otherwise, the current TDP model pretty much stands. So since you have no source I assume for Agena's 95W TDP at all across the board, so I have to go with what is currently available.

No but they give you a very good idea of what the power level is, and with the 6000+ it's pretty much in line with Intel's TDP even with the assumption that the chipset is drawing a little more power on the AMD side. The 4x4 FX platform is absolutely ridiculous on that front.

http://techreport.com/reviews/2007q1/quad-core/index.x?pg=12 I dunno how your going to account for 216W, at load, in Cinebench.

You also seem to conveniently forget about the 6000+ which I stated has a 125W TDP, and is a non-FX processor. And is actually quite close to it's stated TDP.

Even though K8L/K10 will have better power management features, they don't come into play when your talking about at full load.

TDP's between the Core 2 Duo's and the Athlon 64x2 have been pretty even in terms of comparing, they are exactly where you expect them to be.

From what I see so far, I pretty much expect the server and desktop TDP to be in sync for a given clock frequency.

Of things that are published, the guys at Fudzilla have it pretty much correct...

Text1
"AMD's upcoming Agena FX and Barcelona native quad core CPU's will consume less than 95 Watts"

They also did some sisoft sandra numbers...

Text2
"In Sandra multimedia benchmark Agena / Barcelona K10 2.5 GHz scores over 400.000 it/s (instructions / second) in integer test and about 300.000 it/s in floating point test. Yeah there are some Agena / Barcelona parts runing at 2.5 GHz at press time and we got the chance to share the scores with the world.

Intel's fastest QX6800 Quad core at 3 GHz scores about ~300.000 it/s in integer and ~180.000 it/s in floating point test.

So at least in multimedia stuff and this synthetic benchmark AMD's native quad core is 33.3 percent faster than Intel's fastest Quad core, while in the floating calculations it is a massive 66 percent faster.

The QX6850 with FSB 1333 might improve the situation a bit, but it won't be enough save the day. Intel needs to go much higher than 3 GHZ to match the 2.5 GHz K10 part"

I am sorry, this is hardly a source. Fuadzilla is about as reliable as the Inquirer without knowing where they get their results from. It's hit or miss, so you will forgive me for the moment if I will remain skeptical of these results.

http://www.techreport.com/reviews/2007q2/core2-qx6800/index.x?pg=12

Even assuming the results were accurate compare the 6000+ and the E6600 above which are overall equal processors in performance.

E6600 vs 6000+

Integer 128000 vs 44800
Floating Point 99300 vs 68700

These are great synthetic scores, but they hardly are representative of reality.

 

[RoyG]

Bulldozer is said to be 50% faster than i7 again

 

[OCGuy]

lol @ this thread....seems to be the same pattern as the BD hype build-up.

 

[frostedflakes]

Interesting necro.

 

[piesquared]

Interesting flamebait.

 

[OCGuy]

Interesting flamebait.

Im guessing you were one of the people saddened that Fudo was incorrect on this a few years back?

This is a perfectly valid necro, and it is actually quite relevant...threads here don't vanish.