Who 'really' cares what is actually going on in the processor... it what happends once it's come out ...

It matters because the price/performance ratio of AMD's is much higher in most cases

So what matter in that statment is the price to performance, not the actual design.

But that's just what they can sell them at and still make a profit (or a loss leader). It still does not matter what is actually inside the processor.
If what goes in comes out faster then it is not a bad design, if they can make it cheaper as well then that is a bonus. (and cooler to a lot of people now)

So the only people that really care are the marketing department as
a) does it sound good when they market it and can they put spin on it.
b) make money off it (be that to gain some market share and dependency, or make actual money)

To the condumer if it comes out fastest for what we do at the price we can afford then it doesn't matter what is actually really happening.

---

Team mauisun.org

Intel has been filling a need for efficient powerful CPUs for virtualization. Next month they will be releasing a true quad core @ 45 nano.

I don't think so. The 45nm CPUs to be released next month I believe are still 2 dual-cores sandwiched together. The true quad-cores don't come until the next iteration which is the end of 2008 I believe.

---

It matters to software developers and hardware enthusiasts who, of course, want to squeeze every flop out of their components.
If you look at reviews of current intel "fourcore" chips on such things as ray-tracing, physics and scene rendering. There is only a 3 fold increase, whereas native quad-core's will see a more linear increase in performance of aroud 3.8 times. When they come out this will be clearer. My point was that AMD is more than capable of sticking 2 Athlon 64 X2 dies onto a chip and calling it "quad-core". But they choose not to and concentrate efforts into developing a native quad-core. I believe this concentration of efforts by AMD leads to a more rounded product. Therefore the price paid for an AMD, though equivalant to an Intel, will churn out more performance in "most" cases.
However I think your points about the marketing department are very valid, I have seen HT enabled P4 based PC's marketed (albeit wrongly) as dual-core.

---

The wait is over:
http://arstechnica.com/news.ars/post/20070910-barcelonas-out-and-the-reviews-are-out.html

---

It matters to software developers and hardware enthusiasts who, of course, want to squeeze every flop out of their components.
If you look at reviews of current intel "fourcore" chips on such things as ray-tracing, physics and scene rendering. There is only a 3 fold increase, whereas native quad-core's will see a more linear increase in performance of aroud 3.8 times. When they come out this will be clearer. My point was that AMD is more than capable of sticking 2 Athlon 64 X2 dies onto a chip and calling it "quad-core". But they choose not to and concentrate efforts into developing a native quad-core. I believe this concentration of efforts by AMD leads to a more rounded product. Therefore the price paid for an AMD, though equivalant to an Intel, will churn out more performance in "most" cases.
However I think your points about the marketing department are very valid, I have seen HT enabled P4 based PC's marketed (albeit wrongly) as dual-core.

My point was though it's not he two separate or native quad that really matter, even to the enthusiast. It's how fast it does it.
If Intels is faster (for what they want) they would buy that no matter if it was 4 individual CPU stuck together for the price they wish to pay. so native, none native.. who cares as long as it does it job ;-)
Software developers... well only if they have to change their programming to pull out the performance... but then they are not the consumers :-)

---

Team mauisun.org

It matters to software developers and hardware enthusiasts who, of course, want to squeeze every flop out of their components.
If you look at reviews of current intel "fourcore" chips on such things as ray-tracing, physics and scene rendering. There is only a 3 fold increase, whereas native quad-core's will see a more linear increase in performance of aroud 3.8 times. When they come out this will be clearer. My point was that AMD is more than capable of sticking 2 Athlon 64 X2 dies onto a chip and calling it "quad-core". But they choose not to and concentrate efforts into developing a native quad-core. I believe this concentration of efforts by AMD leads to a more rounded product. Therefore the price paid for an AMD, though equivalant to an Intel, will churn out more performance in "most" cases.
However I think your points about the marketing department are very valid, I have seen HT enabled P4 based PC's marketed (albeit wrongly) as dual-core.

I personaly believe that all the cpu's since the dual core are way ahead of the programers (sofware) already. There are very few programs out there science or not that take advantage of these cpu's. That why virtualization is such a big thing. It is the only way that they can really ultitlize all the cores in dual and quad cores. Once programers write and compile for quad cores that is when we might see a real differnce in AMD vs Intel, but until then which ever run the current versions of software wins no matter what the cpu's design is.

---

Jmarks

It matters to software developers and hardware enthusiasts who, of course, want to squeeze every flop out of their components.
If you look at reviews of current intel "fourcore" chips on such things as ray-tracing, physics and scene rendering. There is only a 3 fold increase, whereas native quad-core's will see a more linear increase in performance of aroud 3.8 times. When they come out this will be clearer. My point was that AMD is more than capable of sticking 2 Athlon 64 X2 dies onto a chip and calling it "quad-core". But they choose not to and concentrate efforts into developing a native quad-core. I believe this concentration of efforts by AMD leads to a more rounded product. Therefore the price paid for an AMD, though equivalant to an Intel, will churn out more performance in "most" cases.
However I think your points about the marketing department are very valid, I have seen HT enabled P4 based PC's marketed (albeit wrongly) as dual-core.

I personaly believe that all the cpu's since the dual core are way ahead of the programers (sofware) already. There are very few programs out there science or not that take advantage of these cpu's. That why virtualization is such a big thing. It is the only way that they can really ultitlize all the cores in dual and quad cores. Once programers write and compile for quad cores that is when we might see a real differnce in AMD vs Intel, but until then which ever run the current versions of software wins no matter what the cpu's design is.

All (well most) the scientific programs can utilise multi cores. Multi processors (or parallel processing as it has been better known) has been around for a long time. It's always been more difficult to program for it, but it's always been used. It's modern programmers who are used to one processing unit, in games, office etc where there has been no target consumer who actually had or needed them.. In science though, they had them for a long time.

---

Team mauisun.org

It matters to software developers and hardware enthusiasts who, of course, want to squeeze every flop out of their components.
If you look at reviews of current intel "fourcore" chips on such things as ray-tracing, physics and scene rendering. There is only a 3 fold increase, whereas native quad-core's will see a more linear increase in performance of aroud 3.8 times. When they come out this will be clearer. My point was that AMD is more than capable of sticking 2 Athlon 64 X2 dies onto a chip and calling it "quad-core". But they choose not to and concentrate efforts into developing a native quad-core. I believe this concentration of efforts by AMD leads to a more rounded product. Therefore the price paid for an AMD, though equivalant to an Intel, will churn out more performance in "most" cases.
However I think your points about the marketing department are very valid, I have seen HT enabled P4 based PC's marketed (albeit wrongly) as dual-core.

I personaly believe that all the cpu's since the dual core are way ahead of the programers (sofware) already. There are very few programs out there science or not that take advantage of these cpu's. That why virtualization is such a big thing. It is the only way that they can really ultitlize all the cores in dual and quad cores. Once programers write and compile for quad cores that is when we might see a real differnce in AMD vs Intel, but until then which ever run the current versions of software wins no matter what the cpu's design is.

All (well most) the scientific programs can utilise multi cores. Multi processors (or parallel processing as it has been better known) has been around for a long time. It's always been more difficult to program for it, but it's always been used. It's modern programmers who are used to one processing unit, in games, office etc where there has been no target consumer who actually had or needed them.. In science though, they had them for a long time.

Oh well just don't tell my wife I was wrong.

---

Jmarks

My point was though it's not he two separate or native quad that really matter, even to the enthusiast. It's how fast it does it.
If Intels is faster (for what they want) they would buy that no matter if it was 4 individual CPU stuck together for the price they wish to pay. so native, none native.. who cares as long as it does it job ;-)

Just take a look at us for example. If you had the choice of a 3GHz Intel non-native quad-core or a 2 GHz AMD native quad-core, what you really care about is if I have 4 BOINC threads running, which one will produce more results in the same time period.

The Intel might be able to finish WUs faster with its 1GHz clock advantage on a single-core system, but if the quad-core could only produce a 3x speed-up vs the AMD with a 3.8x speedup, you would have to see if the extra GHZ can make up for the "loss" of one effective core.

But the difference is not as big with a quad-core system, it is more the 4-socket systems where you get 16 cores on one machine that you would really start to see the scaling difference I would imagine.

---

My point was though it's not he two separate or native quad that really matter, even to the enthusiast. It's how fast it does it.
If Intels is faster (for what they want) they would buy that no matter if it was 4 individual CPU stuck together for the price they wish to pay. so native, none native.. who cares as long as it does it job ;-)

Just take a look at us for example. If you had the choice of a 3GHz Intel non-native quad-core or a 2 GHz AMD native quad-core, what you really care about is if I have 4 BOINC threads running, which one will produce more results in the same time period.

The Intel might be able to finish WUs faster with its 1GHz clock advantage on a single-core system, but if the quad-core could only produce a 3x speed-up vs the AMD with a 3.8x speedup, you would have to see if the extra GHZ can make up for the "loss" of one effective core.

But the difference is not as big with a quad-core system, it is more the 4-socket systems where you get 16 cores on one machine that you would really start to see the scaling difference I would imagine.

if i remember correctly...you can't compare amd and intel clock speeds because of the way they process data. so amd's 2 ghz is just as fast as intel's 3ghz, but rated slower due to the way it processes data.
read this article to see my point

---

My point was though it's not he two separate or native quad that really matter, even to the enthusiast. It's how fast it does it.
If Intels is faster (for what they want) they would buy that no matter if it was 4 individual CPU stuck together for the price they wish to pay. so native, none native.. who cares as long as it does it job ;-)

Just take a look at us for example. If you had the choice of a 3GHz Intel non-native quad-core or a 2 GHz AMD native quad-core, what you really care about is if I have 4 BOINC threads running, which one will produce more results in the same time period.

The Intel might be able to finish WUs faster with its 1GHz clock advantage on a single-core system, but if the quad-core could only produce a 3x speed-up vs the AMD with a 3.8x speedup, you would have to see if the extra GHZ can make up for the "loss" of one effective core.

But the difference is not as big with a quad-core system, it is more the 4-socket systems where you get 16 cores on one machine that you would really start to see the scaling difference I would imagine.

if i remember correctly...you can't compare amd and intel clock speeds because of the way they process data. so amd's 2 ghz is just as fast as intel's 3ghz, but rated slower due to the way it processes data.
read this article to see my point

Well apart from the Pentium-M & even more the Core2 cahnged that again and it constantly changes... but I think you should read it as a hypothetical situation ;)

---

Team mauisun.org

if i remember correctly...you can't compare amd and intel clock speeds because of the way they process data. so amd's 2 ghz is just as fast as intel's 3ghz, but rated slower due to the way it processes data.

You can't compare clock speeds directly which is why I said, "might be faster" but with such a high clock-speed difference (50% in the case I was talking about) it could push it ahead.

---

Not bad, Francois :P



Didn't imagine you'd actually print them tho!

:-)

---

Faust.

"Intel Senior Performance Engineer Francois Piednoel shows off Cinebench running on a 3.4GHz overclocked Skulltrail dual quad-core CPU system."

32-bit Windows, one G80GTX, 2x 3.4GHz quad cores

http://www.youtube.com/watch?v=s5pFoPg7Jm8

Around 41 secs
Multi CPU: 21989 CB CPU


Skulltrail is running 2x QC Penryn Harpertown socket 771 at 4GHz.

http://www.youtube.com/watch?v=9UkAX5Ojn9s


Oh my god... Who? put this new toys in rosetta and show us what is RAC!!!

---

http://amd-member.com/campaigns/black/

Does anyone know what they will launch?

TriFire or phenom??

---

http://amd-member.com/campaigns/black/

Does anyone know what they will launch?

TriFire or phenom??

Maybe this.
'AMD will offer a tri-core Phenom processor for desktops to bridge the gap between dual- and quad-core models.'

http://www.eweek.com/article2/0,1895,2184471,00.asp

---

Jmarks

The tri-core will be the quad-core processors with problems. It is a way to make some money with a bunch of "broken" processors. To be honest, i think its a smart move, and the consumer will prefer a 3 core processor to a dual core processor with overclock (Intel)...

---

http://amd-member.com/campaigns/black/

Does anyone know what they will launch?

So it is the 25th and still the same message, did they launch anything today?

---

They launcheg a (new? :/) processor...


http://game.amd.com/us-en/unlock_athlonblack.aspx

Things aren't quite good with amd...

---

They launcheg a (new? :/) processor...


http://game.amd.com/us-en/unlock_athlonblack.aspx

Things aren't quite good with amd...

Huh??? Cant see whats negative on that, unless u sing someone elses song.

They release a line of CPUs with free multiplier - thats a great thing for a desktop CPU and people that like to overclock!

---

http://www.MIAteam.eu