Why is Conroe so great?

[Gashman]

pushed wrong key, hey, why not and sure i get your point, hell i don't play games all the time, i spend lot of time with my girlfriend, i play guitar a lot, use studio max 8 and listen to music/chat, but yeah conroe is definately immensely powerful in number crunching, could come in handy for studio max in all truth. but pentium 4s are powerful at encoding and what not as well, so its not the only processor good at it, it is in a league of its own mind. im planning on using my 3200+ until AMD release 65 nano chips and see what there like, then decided on upgrade PS. my CPU runs at 2.85Ghz at the moment so it should be faster than FX5*. FX is in brother-in laws system with 7600GT replacing his dead card, its overclocked as well to 3.1Ghz i think but its not entirely stable in benchmarking, but no problems in gaming and such, we got same frame rates in doom 3, far cry, americas army and ghost recon 3 +/- one or two frames/second and for purpose of comparison my 3200+ was running at FX55 speed (2.6Ghz), so that review shows the graphics card as the obvious bottleneck.

another problem with conroe is expensive motherboards at the moment, other than that, keep the forum informed on performance and clocking, especially the E6400 and E6600

 

[drunkenmaster]

a bunch of different things to point out. dolph, the title and every page of the review states its a gaming performance review, so saying its supposed to be a cpu review is completely and utterly wrong. one benchmark was run with only one option dropped down slightly as they , and other sites have also, try to show what is playable options, both options would have looked 99.9% the same and the different option was probably not that much slower. that was only oblivion, the other tests use the same results. they do infact do min/max/average framerates you so again you are wrong there.

for 99% of computers bought for home use gaming will be very likely to be the most intensive apps used. of course conroe is faster in encoding and 3d modeling and other very cpu intensive stuff, but honestly think of every family member and friend who has a computer and then ask yourself how many of them do anything more intensive than gaming, then ask yourself how many of them do anything more intensive than writing an e-mail.

as for heat and power, your average ath 64 doesn't use that much more power, and aren't far away from all dropping to a 65w tdp. which is a maximum theoretical(basically impossible/improbable to ever hit tdp), that it will not be at all the time even when the cpu says its at 100% encoding it will still get branch mis predictions and flush out the pipeline, and not prefetch stuff all the time and basically won't have every single part working at every point in time. intels tdp is a average(and god knows what that is) number, and doesn't include the mem controller wattage while amd's on die chips obviously do, and the mem controller does use another chunk of wattage. real usage numbers, ath 64/s conroes are very very very close now, when amd drop to 65nm then will more than likely be lower power in real world use.

not a single site except those very early preview numbers that were unofficial from some guy on a forum, that 100's of reviews/previews have proved wrong, got hugely higher min fps in any game at any res, let alone the resolutions he claimed the numbers at. they were fake, wrong, lies, bs, not real, unbelievable, disproven and crap. the guys a muppet basically.


oh, and [h] benched half life 2:ep 1, not half life 2, it has improved textures and is harder on graphics, and i think he had hdr enabled aswell, though i'm not sure about that tbh, the max fps is clearly stated as 96fps, the average is the same the minimum is the same, there is no v-sync enabled.

 

[Gashman]

well i think its like the performance between A64 and netburst, A64 beats netburst pretty much constantly but not by a huge amount most of the time. in this case conroe is the A64 and AMD are netburst, so its just swapped sides until AMDs release with there competitor to conroe, which could possibly be when they switch to 65 nano. there aren't many peoblems with A64s architecture, they should improve there existing design for the time being to make it closer clock for clock to conroe, as well they should clock well being 65nm

 

[Fr0dders]

oh, and [h] benched half life 2:ep 1, not half life 2, it has improved textures and is harder on graphics, and i think he had hdr enabled aswell, though i'm not sure about that tbh, the max fps is clearly stated as 96fps, the average is the same the minimum is the same, there is no v-sync enabled.

lol can you explain why the 7900 GTX only achieved 60 fps, where on numerous other systems, the Same HL ephisode 1 benchmark has been run and achieved higher results

like this ?

this is the same test, the HL EP1 stress test, run at 4xaa and 16xaf at 1600 x 1200

and the 7900 GTX is seen here scoring 88 fps in this benchmark. And this was a 7950 review, that group tested some GPUs. So theres no bias here towards core, it was released back on the 5th of june long before any1 had seen conroe.

so why are all the [H] benchmarks at 60fps ?

i mean all getting 70 would be weird, but plausable. all 60 fps ? how else can you explain that result except if they had Vsync on ? they say they didnt. but in light of that benchmark above what else can it possibly be ?

and that system was only an FX57. whats they're FX62 doing getting 60 fps with the same vid card ?

 

[manveruppd]

Conroe sucks because I just got a Core Duo laptop.

ConrPWNED!

Sorry, couldn't resist!

Back on topic, regarding the [H] review, at what resolution would you all say they SHOULD have tested so as to not be limited by the graphics card and provide a more accurate comparison of the CPUs under test? 1600x1200 seems pretty high to me!

 

[fornowagain]

This made for interesting reading.

http://www.anandtech.com/memory/showdoc.aspx?i=2800

This consists of running both processors at the exact same speed - 2.93GHz - achieved at the same ratios -

We can now say with authority that Core 2 Duo is the faster performer clock-for-clock across the board. At the same 2.93GHz Far Cry is 27.7% faster, Half-Life 2: Lost Coast is 12.4% faster, and Quake 4 is 22.2% faster on Core 2 Duo. Of course AMD does not currently have a 2.93GHz CPU, so we tested by overclocking FX62. This suggests that FX64, or whatever it will be called, will not help much at 3.0GHz with a 200 clock speed.

 

[Gashman]

you do realise you won't notice the increase in the slightest, unless your sitting there watching the FPS counter, so how is it so superior if you won't even notice?

Edit: heres an anology for you, whats the difference between a ford focus and a lamborghini murchialargo if both are restricted to 60Mph, theres no difference at all, even though the lamborghini is potentially mind bogglingly faster, it makes no difference if you cannot use that speed. i.e. the human eye won't notice the improvement cause its restricted to around 60 frames/second

 

[maddness]

you do realise you won't notice the increase in the slightest, unless your sitting there watching the FPS counter, so how is it so superior if you won't even notice?

Edit: heres an anology for you, whats the difference between a ford focus and a lamborghini murchialargo if both are restricted to 60Mph, theres no difference at all, even though the lamborghini is potentially mind bogglingly faster, it makes no difference if you cannot use that speed. i.e. the human eye won't notice the improvement cause its restricted to around 60 frames/second

well the lambo will get to the 60 a hell of a lot faster

 

[Bomag]

Not if you drive both of Beachy head- they will do 0-60 and eventually back to zero in exactly the same time!!

Given that I have 7800 GTX in SLI my enthusiasm for upgrading to a conroe after the holidays is declining rapidly. Now a low watt 939 x2 which can do 3ghz for £200 (a possibility in September) could be more intersting.

 

[NathanE]

you do realise you won't notice the increase in the slightest, unless your sitting there watching the FPS counter, so how is it so superior if you won't even notice?

Edit: heres an anology for you, whats the difference between a ford focus and a lamborghini murchialargo if both are restricted to 60Mph, theres no difference at all, even though the lamborghini is potentially mind bogglingly faster, it makes no difference if you cannot use that speed. i.e. the human eye won't notice the improvement cause its restricted to around 60 frames/second

It's funny but there was many AMD fans saying the exact opposite when the NetBurst chips started falling behind the game, just as the A64's are now...

You're exactly right though really. Games remain GPU limited. But people don't just play games...

 

[Bennah]

Do we really need to actually answer the original question.

Conroe is the next big thing.

GET ONE!!!

You will be stupid not to...

 

[BillytheImpaler]

It's funny but there was many AMD fans saying the exact opposite when the NetBurst chips started falling behind the game, just as the A64's are now...

You're exactly right though really. Games remain GPU limited. But people don't just play games...

It goes beyond that even. You're more apt to upgrade a GPU later to one that does not bottleneck as much. The Conroe, whose performance is superior to that of any of AMD's offerings, will last longer before it is wholly obsolete and needs replacement.

Add to that that it is cheaper than the AMD Uber-chips against which it is being benchmarked and it is easy to see why it is winning praise from me and others.

 

[Intel Guy]

You can swap in a Merom. They're pin compatible and output less heat than the Yonah you've got.

Not as standard, it requires a bios update, and i dont know if these will be released to the public domain.

 

[Gashman]

It's funny but there was many AMD fans saying the exact opposite when the NetBurst chips started falling behind the game, just as the A64's are now...

You're exactly right though really. Games remain GPU limited. But people don't just play games...

thing is though, netburst chips weren't 'that' much slower than A64, sure A64 beat them at everything, but usually by a reasonable margin, matt still plays on netburst, smithfield to be exact, its now running fine thanks to new heatsink

 

[silversurfer]

i.e. the human eye won't notice the improvement cause its restricted to around 60 frames/second

Jet pilots have shown they are able to detect differences upto 400fps but thats using specialist software that actually puts a difference to each of those 400 frames. I think most games dont update this fast anyway so its only repeating the same frame as the previous, ie pointless.


Conroe is good for making dvds quickly, apart from that most people wont use its power for years. You can be sure though that any slowness in the system is not down to the cpu in the slightest, which is nice
It allows you to turn on compression on ntfs drives with no worry and various other things

 

[Gashman]

Jet pilots have shown they are able to detect differences upto 400fps but thats using specialist software that actually puts a difference to each of those 400 frames. I think most games dont update this fast anyway so its only repeating the same frame as the previous, ie pointless.


Conroe is good for making dvds quickly, apart from that most people wont use its power for years. You can be sure though that any slowness in the system is not down to the cpu in the slightest, which is nice
It allows you to turn on compression on ntfs drives with no worry and various other things

thats the exact reason i don't understand why overclockers/gamers get so excited about 2 frames/second better than someone elses system

 

[HangTime]

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.

Many years ago I used to find [H] quite interesting, but for the last couple of years their method of benchmarking has been utterly, UTTERLY appalling. The final straw for me was them explaining how they benched Doom3 for one particular review, by running through a level with fraps logging the framerate, and then doing the same with another card - which of cause is a completely flawed test, because no two plays through a game are identical (by their own admission). Not only that, but they weren't comparing apples to apples, they do some stupid thing where they change the settings based on what they think is the 'max playable setting' for the card in question. So you end up with conclusions being drawn from results which based on different benchmarks using different settings... nice one.

 

[HangTime]

I think most games dont update this fast anyway so its only repeating the same frame as the previous, ie pointless.

Some games do. Quakeworld for example can update many hundreds of times per second provided you have the right client/server settings. And even in other games where the server is only updating at, say 60hz, you still get a smoother 'feel' with say 120fps.

The real problem is that very few monitors have refresh rates over 240hz, and most are 160hz or less.

 

[Spyhop]

the human eye won't notice the improvement cause its restricted to around 60 frames/second

I can't believe how many times I have read people come out with rubbish like this. You can't put a number on it, since it's always going to be dependant on the nature of motion/video you're observing and the environment you're doing it in. If there is an absolute upper limit, it's going be a lot higher than that.

 

[Cyber-Mav]

conroe is great because it is fast and cheap.