![[TxM]Spitfire](./core/images/default/default_avatar_large.png "[TxM]Spitfire"){kind=avatar}
Re: Intel the King? That's not what it says here.
O here we go again :wall: :blabla:
Originally posted by ibelcaniva
O here we go again :wall: :blabla:
-
-
Re: Intel the King? That's not what it says here.
Hmm, that's a tough one. I'd say they're just about perfect for eachother actually. Which is bad for you since you don't gain any performance from upgrading either one. But look on the bright side - You aren't losing any performance to bottlenecks! Lol.Originally posted by Teh_Spambot
But if you had to upgrade something, I'd say go with a videocard first.Comment
-
Re: Intel the King? That's not what it says here.
How about my rig in my Sig :) ANd im also pissed becuse i got this comp befor i even knew about the new INtel CPUS ****ing im PISSEDbecuse i always liked Intel but AMD was better for gameing so i went with them
Comment
-
Re: Intel the King? That's not what it says here.
You're missing my point. Conroe (unicorn) supposedly has a shared L2 cache. Super PI is a very L2 cache intensive program. You are getting scewed results by only running a single PI because that one instance of PI is able to use all of the L2 cache. If you ran dual then the cache would split it resources to both cores effectively decreasing your PI runs by 4-5 seconds on the 1M test.Originally posted by Flanker
How do you know both would score the same if an intance was run on each core? Would the core split it's L2 evenly? How will this affect multi-threaded applications. I'm tired of seeing benchmarks of a dual core CPU exploiting it's shared L2 in single benches. I want to see some real world dual benches.
Conroe isn't a single core processor with a 4Mb cache so I expect to see some benches that show what the processor is meant for.Comment
-
Re: Intel the King? That's not what it says here.
Thats what my tech savvy friend thought, and what I thought from his input as well.Originally posted by Flanker
For some reason the comp just doesnt seem to perform as well as it should.... many people run medium/ high or even all high with that card and similar processors, yet mine has trouble on medium multiplayer.
Yet in singleplayer, it handles all high pretty good..... for some reason, going from 512 to 1024 mb ram didnt seem to do much. I still have to test to see if 1024 to 1500 did more.Comment
-
Re: Intel the King? That's not what it says here.
More ram helps this game tremendously. Having 1.5Gb + is the sweet spot for BF2.Originally posted by Teh_SpambotComment
-
Re: Intel the King? That's not what it says here.
Definite bottleneck by the videocard there.Originally posted by [STO] Killer DragonHow about my rig in my Sig :) ANd im also pissed becuse i got this comp befor i even knew about the new INtel CPUS ****ing im PISSEDbecuse i always liked Intel but AMD was better for gameing so i went with them
If you go with any modern processor (2Ghz+ A64, 3Ghz+ P4, any Conroe) then it is unlikely that there will not be a bottleneck. For reference, my processor (3200+) can handle a 7900GT (not saying I have one) while at it's stock speeds.
Thats why I say that the videocard is the most important part of a gaming computer. You can get the best videocard (single) and get a relatively weak processor and still get the same gameplay as if you got a FX-62 or a X6800. Which is why you should spend a nice amount on your videocard in a gaming computer.
Originally posted by ThunderPimp
What's wrong with a Dual Core CPU using all of the unused 4MB of cache in single-threaded applications? That's of the advantages of a Conroe. Conroe, in single-threaded applications, IS a single core processor with 4MB of L2 cache.
And all SuperPi records are set on one thread. Unless you want to start a Dual Core SuperPi competition, Conroe will continute to dominate in SuperPi.
I know what you mean since I have the same videocard. I think it depends a lot on the amount of people on the server. I know that sometimes I'm getting maxed out FPS on the same settings, while on a more populated server I'll be hanging around 30-50FPS on Med settings on 1280x1024.Originally posted by Teh_SpambotComment
-
Re: Intel the King? That's not what it says here.
Whats wrong with it is that those results don't corespond to multithreaded app performance which are the future of computing. It's a bunch of hype.Originally posted by Flanker
In applications that use both cores such as 3Dmark 06 the Conroe only leads a comparible A64 by 1-2%. This is with the X6800 compared to the FX62.Comment
-
Re: Intel the King? That's not what it says here.
Hah, show me some benchmarks. Its more like 30%.Comment
-
Re: Intel the King? That's not what it says here.
Most of the new reviews don't show 3Dmark 06... There is a reason for this.Originally posted by Flanker
They only bother running an X2 3800+ in the link below because an FX would show how feeble conroe really is. Don't give me that GPU bottleneck crap since this site actually runs the test at 1024x768:
http://www.tweaktown.com/articles/92...k06/index.html
All of those benches are consistent with each other. Hardly a fluke.
Do I even need to keep going? Owned.Comment
-
Re: Intel the King? That's not what it says here.
Owned? Hardly. You're comparing the overall scores buddy, you don't do that with 3DMark, not even '06.
Look at the CPU only tests. And keep in mind, the E6700 is a 2.66Ghz processor while the FX-62 is a 2.8Ghz processor. And the E6700 still beats it.
Owned by your own benchmarks.Comment
-
Re: Intel the King? That's not what it says here.
Comment
-
Re: Intel the King? That's not what it says here.
Why not compare overall scores? It's the whole system that needs scoring not just the processor. Im not soley using an AMD processor all alone am I? It has other components that work together in a complete system. With those 3Dmark 06 benches you can clearly see a Conroe system is barely better than an X2 system.
In the overal score comparison the Conroes even have the benefit of DDR2 and still barely best a similar AMD system using DDR1.
The X6800 share the same exact components as the FX62 systems except the motherboard and processor in those benchmarks. You can bet you should compare overall 3Dmark scores with those.Comment
-
Re: Intel the King? That's not what it says here.
Freak test. Compare the score of the E6700 in that test to the scores in the other 3 benchmarks. Something went wrong. Its the same processor in the same test, and the scores for the other CPUs in the other benchmarks are fairly close. They messed up something on that test.Originally posted by Sir. MatDef
*Think's about it for a second.*Originally posted by ThunderPimp
*Throws the idea in the trash.*
We don't compare the overall score because we're talking about the processors. And 3DMark06 is 90% videocard, 10% processor. A faster processor makes a fraction of the difference in the overall score than it actually should. Why NOT look at the CPU only scores? They're the ones actually evaluating the CPUs.
If you want to compare overall scores, go look at gaming benchmarks. This is what the complete systems look like;
Comment
-
Re: Intel the King? That's not what it says here.
Lets talk about processors then. Even the CPU benchmarks only show about a 15% benefit with Conroe over the same speed A64.
Where is this mythical 30% you speak of? Maybe the unicorn has it.
There you go pulling single threaded benchmarks out when I'm clearly talking about stressing the processor as it was designed. Go ahead and try to change the subject to take attention away from Conroe's weakness. Where are the 3Dmark 06 benches from anandtech? Thas right they aren't there.Comment
Comment