Why is Conroe so great?

Fr0dders · 23 Jul 2006 at 10:02

look at the conroe gaming thread

hes got that oppinion from the [h] review of conroe. that says just that

http://enthusiast.hardocp.com/article.html?art=MTEwOCwsLGhlbnRodXNpYXN0

personally i think that review has to be bs, for the reasons ive put in that thread, how can a 7900 gtx not score more than 60 fps in hl2 except if vsync was on ?

anandtech have a review of the 7950, did a group test, and their 7900 gtx got 88 fps in the same bench. and that was just a graphics card review on a fx57 system. so bias for or against conroe doesnt come into it

what do you reckon to that review ?

Dolph · 23 Jul 2006 at 10:33

Richdog said:
I don't understand why people are saying that at high rezzes you see no performance increases... does the extra power of the Conroe not count for anything towards AI etc even at 1600*1200? Surely a 3200+ wouldn't be able to do it at the same pace if the screen got busy would it, or is a large amount of the processing power wasted?

A system can always be bottlenecked by the slowest componant for a given task. In the case of high res graphics there is a limit to how fast the card can render the information presented to it, so it doesn't matter how quickly the CPU can generate that information, it's limited by the rendering speed of the graphics card. You may notice small fluctuations but nothing major.

Think of it like a small roundabout. It doesn't matter whether you have one lane or ten coming in, only one car could enter the roundabout at a time..

Dolph · 23 Jul 2006 at 10:36

MrLOL said:
look at the conroe gaming thread

hes got that oppinion from the [h] review of conroe. that says just that

http://enthusiast.hardocp.com/article.html?art=MTEwOCwsLGhlbnRodXNpYXN0

personally i think that review has to be bs, for the reasons ive put in that thread, how can a 7900 gtx not score more than 60 fps in hl2 except if vsync was on ?

anandtech have a review of the 7950, did a group test, and their 7900 gtx got 88 fps in the same bench. and that was just a graphics card review on a fx57 system. so bias for or against conroe doesnt come into it

what do you reckon to that review ?

That [H] Review made me want to stab my own eyes out it was so bad... It's a shame to see them dropping so far down as they used to be well respected for well written, relevant and independant reviews.

This conroe review I thought was much better.

http://techreport.com/reviews/2006q3/core2/index.x?pg=1

FreeStream · 23 Jul 2006 at 10:36

Hardop review is rubbish for the most part.

Conroe is good because its cheap (although not the chepeast out of AMD/Intel atm) but still the cheapest one is very close to beating the AMD FX-62, after a mild overclock is whips its backside.

Overclocking: AMDs X2 on air tend to top at 3ghz. At the same clockspeed Conroe is 5-50% quicker, so at 3Ghz conroe still has 500 mhz to go just on air - although retail chips might be rubbish.

Performance: They are VERY good at number crunching, 15/18 sec at stock? 12/13 sec clocked on air. The fastest AMD has done it in what 17s? a 4.2Ghz FX-57. a usual 3Ghz does it in about 23/24 IIRC. And don't say it means F/A, as it shows number crunching power. Anything from Folding to 3D modelling will benefit from this. 3D is usually limited by the graphics card. But in some games (flight sims) Conroe excells as they are all majorly CPU limited.

Until AMD ramp up clockspeeds more (although the FX-62 is getting a very high TDP now) or they go to 65nm, AMD are now behind Intel, for how long? We don't know. K8L in 2008 might briung them ahead again.

CR.

clv101 · 23 Jul 2006 at 10:42

The Conroe is so great because:
Highest performance at the price point.
Low power consumption for the performance (knock on benefits of less cooling, quieter, lower elec bills etc).
Appears to overclock well.
Doesn't need expensive RAM.

In comparison the AMD and older Intel offerings look slow, expensive and hot.

The only downside I can see at the moment is the lack of good cheap motherboards. I'm waiting until early Sept when the CPU price should be down a little (I'm looking for a £200 E6600) and the situation with motherboards will be much clearer.

man_from_uncle · 23 Jul 2006 at 11:00

Conroe is great because:
The overclocking headroom (given that retail performs no different to engineering samples) is such that pound-for-pound a 6300 can score 3 times what a opteron 144 can score in the CPU test of 3dmark06 (assuming both are clocked as far as they go) and 50% more than a 165.

The downside of course is that, for most people, it'll require a full upgrade costing anything from £350-£1000. And as clv101 mentioned, the motherboards aran't quite there yet. They still cost a lot, the BIOSs are glitchy and they don't support SLi.

Dolph · 23 Jul 2006 at 11:02

man_from_uncle said:
The downside of course is that, for most people, it'll require a full upgrade costing anything from £350-£1000. And as clv101 mentioned, the motherboards aran't quite there yet. They still cost a lot, the BIOSs are glitchy and they don't support SLi.

Actually, that's not quite true. A 975x chipset based board will happily support SLI, however Nvidia won't release drivers to allow it to do so with Conroe (although drivers can be modified to allow it), as they don't want the NF5's big promotional point taken away

Richdog · 23 Jul 2006 at 11:06

Just read that review and if it holds true Core2 seems to be worthwhile over an X2 only if you're a benchmark freak (which is the most worthless pursuit ever in my eyes) or you do heavy amount of encoding etc which would then save you a lot of time.

For pure gamers, it doesn't seem there's any point in making the switch... I may as well just buy a cheap X2 and clock it to shizz and back and save myself the bother of upgrading for a while.

Dolph said:
That [H] Review made me want to stab my own eyes out it was so bad...

This conroe review I thought was much better.

http://techreport.com/reviews/2006q3/core2/index.x?pg=1

Why was it so bad... it was just testing real-time games at sensible resolutions that mmost people will be playing at and finding barely, if any, difference... something which has been shown before when comparing low and high end CPU's at higher resolutions. All others are testing at 1024*768... and who plays at that res with a modern graphics card?

Dolph · 23 Jul 2006 at 11:38

Richdog said:
For pure gamers, it doesn't seem there's any point in making the switch... I may as well just buy a cheap X2 and clock it to shizz and back and save myself the bother of upgrading for a while.

Well, except the obvious points of higher minimum frame rates (something [H] didn't investigate unlike most of the other reviews as they were altering the settings rather than doing a like for like comparison.

(check the 'playable settings' section on this page, for example http://enthusiast.hardocp.com/article.html?art=MTEwOCwzLCxoZW50aHVzaWFzdA== , the settings used for the Conroe benchmarks are different to the ones used for the FX-62)

There's also the fact that Conroe will prove more future proof as graphics cards improve as the FX-62 will bottleneck quicker than the conroe.

Why was it so bad... it was just testing real-time games at sensible resolutions that mmost people will be playing at and finding barely, if any, difference... something which has been shown before when comparing low and high end CPU's at higher resolutions. All others are testing at 1024*768... and who plays at that res with a modern graphics card?

Firstly, it was supposed to be a CPU comparison, and it's nothing of the sort, with those settings it's a GPU comparison as that's where the system is limited. Surprisingly if you give three different systems the same limiting variable, they'll produce very similar results.

Secondly, as I've already mentioned, they seem to be tweaking the settings to get the same frame rates, rather than choosing a set of settings and getting comparison benchmarks. They are comparing apples and oranges, rather than apples and apples.

Thirdly, as someone else mentioned, at least one of the tests was run with V-sync on... Enough said about that.

Finally, they seem either unable or unwilling to run genuinely CPU limited benchmarks, even in their so called 'apples to apples' test.

All in all, it was a bad set of testing using poor methodology and pretty much useless for the purposes of comparing one CPU to another.

Richdog · 23 Jul 2006 at 11:49

Dolph said:
Firstly, it was supposed to be a CPU comparison, and it's nothing of the sort, with those settings it's a GPU comparison as that's where the system is limited. Surprisingly if you give three different systems the same limiting variable, they'll produce very similar results.

I think you're missing the point and have perhaps got the wrong end of the stick. He stated in that review that he had already done a "proper" CPU test in a seperate article, the point of this review was to show that during gaming at higher resolutions, the benefits of a faster CPU are going to be marginal, which he proved by comparing the different CPU's in the same games at the same settings. There was nothing wrong with his review as he clearly states his intention at the very beginning.

MikeTimbers · 23 Jul 2006 at 11:55

I agree. The HardOCP article must be taken alongside its other reviews of Conroe not in isolation. Thay had already fully reviewed Conroe's ability in encoding etc. and agreed it was way faster than anything else and should absolutely be the processor of choice if you are buying new.

In the gaming article it merely showed that for today's games, if you already have a high-end rig, moving to Conroe will provide limited benefits because it is not the cpu that decides gaming performance once you have a cpu above a certain performance level. They never said that Conroe wasn't faster - merely that it wasn't as huge a difference in gaming as it is in cpu-intensive applications.

w3bbo · 23 Jul 2006 at 12:48

So to cut a long story short any gamer wanting to jump ship and go to conroe from a mid-high end AMD setup will notice little/if any benefit from doing so. If thats the case I'm glad I didn't do it. I can wait a few extra seconds/minutes to finish encoding etc, safe in the knowledge that my fps isn't suffering. However there is also the arguement that when the next gen cards along with dx10 are released the cpu's will be less of a bottleneck than they are now...but by then a new range of cpu's(quad/4x4 core) will no doubt be with us which negates the arguement to upgrade to conroe if you have a recent spec gaming pc imo.

-|ScottFree|- · 23 Jul 2006 at 13:37

w3bbo said:
...but by then a new range of cpu's(quad/4x4 core) will no doubt be with us which negates the arguement to upgrade to conroe if you have a recent spec gaming pc imo.

There are going to be the odd few that are going to upgrade solely to build a system around the Conroe that already have very high specification equipment at there disposal, but they have a very high disposable income or are caught up in the hype of this CPU.

For most enthusiast to cross over to Intel at this moment "IF" they already have a very high standard rig, the gains that would be made from what I can gather would be negligible. You have to upgrade several elements of your machine, motherboard, Ram, to house the CPU. Your talking £600 for the E6600 and decent mobo and 2GB DDR2 ram. That is not cheap for what its bringing to the table at this point with regards to gaming.

Waiting for the release of the Dx10 cards arriving in a few months, and saving your hard earned cash for its arrival, will make the whole experience of crossing over that more pleasurable. More toys, more to play with = more fun. Not only that, a lot of the problems will have been ironed out regarding bios updates and compatablility issues. There will be atleast 20 new motherboards on sale, a little better value for money too. The savings you make will go towards your Dx10 cards!!!

Having said that, most of us will still be using dx9 XP, but there will still be huge improvements made in dx9 with these cards.

Gashman · 23 Jul 2006 at 17:21

Edit: deleted due to being heated post :rolleyes:

NathanE · 23 Jul 2006 at 17:58

Not everyone plays games 24/7, Gashman.....

james.miller · 23 Jul 2006 at 18:03

NathanE said:
Not everyone plays games 24/7, Gashman.....

he seems to have vastly overlooked that point. Encoding and crunching and indeed windows performance in general is very inportant to some people. and what is an fx58?

Richdog · 23 Jul 2006 at 18:28

james.miller said:
he seems to have vastly overlooked that point. Encoding and crunching and indeed windows performance in general is very inportant to some people. and what is an fx58?

I would hazard a guess he means he has an A64 clocked to the speeds of what would theoretically be an FX-58.

james.miller · 23 Jul 2006 at 18:32

oooh like my fx-59 you mean

TooNice · 23 Jul 2006 at 18:32

Indeed. Conroe is good for the reasons why many of the AMD chips were well regarded in the last few years.. even though they carried the same limited benefits in very high resolution gaming. It didn't stop them from being better bang for buck than their Intel counterpart.. and the same applies now. I won't be swapping out my Opteron because it is inconvenient for me to do so right now. But if a friend wants a new PC, I'll nudge them towards a Conroe.. just like I have previously nudged people towards an A64/Opteron.

-|ScottFree|- · 23 Jul 2006 at 18:46

TooNice said:
But if a friend wants a new PC, I'll nudge them towards a Conroe.. just like I have previously nudged people towards an A64/Opteron.

If a friend wants a new PC you should hudge him towards your Opty...then you could go bet yourself the Conroe. It all depends how much you value your friendship