CPUs and Gaming

[Digital Punk]

In your case, as your a gamer, then its better to spend the cash on a new graphics card.

That would give you more fps than just upgrading your cpu, mobo and ram, and would probably be cheaper.

 

[trojan698]

Anandtech, the fps boost in all games has been quite substantial for me especially at 1600x1200

 

[TaKeN]

I might upgrade to conroe,


Its sooooo tempting

 

[wellibob]

I think the best way to upgrade, is to make sure you sell your current/old gear as soon as. This way, an upgrade is always worth it. I been using AMD for ages, but it isnt the best performance chip, or value for money anymore. Intel Duo 2 is. Upgrade if you can, if not ..plod on with what you have.

 

[TaKeN]

TBH I'm not exactly plodding..

Its fast as a badger in windows.

and Plays games Flawlessly


If i got money for my rig then yeah i would definately got Conroe, but i don't see my dad saying yeah i will give you £400+ lol

 

[mrthingyx]

The biggest factor in performance is the graphics card. IMHO.

I run X3 on my system and it makes little difference if my XP1700 is clocked at 1.4GHz or 2.1GHz (believe it or not, I am aware that it is an old CPU and probably wouldn't make that much difference anyway). However, I swapped my old 9600XT for a 9800Pro and the game went from a stuttering slideshow to something very playable with only occasional stuttering.

IMHO, the [H]OCP and Anand are both correct: more horsepower under the hood makes for improved fps (10-20% in the case of Conroe vs. a similarly priced AMD) but when your pushing out over 80fps anyway, who's really going to notice the difference in the grand scheme of things? I know there are anecdotal stories of massive performance gains, etc., but on a practical level? Is it worth the outlay?

AmaTex: the reason your system chugs in CS:S is sitting pretty with a Geforce 5700LE sticker on it. Your CPU is fine for HL's engine but your graphics card is the main limiting factor in your case.

Easyrider: I'm assuming your AMD system was clocked to around 3GHz? So going to a Conroe at 3.8GHz (which, as you have noticed, is not necessarily a dead cert) is going to make a bit of a difference to your gaming experience. Not everybody will get the same gains as you have out of their new system.

 

[2bullish]

There is no real discernable difference between a high end AMD system(either single or dual core) to a Conroe in games. With these chips the bottle neck is rarely the cpu. Even if a Conroe system does allow for a higher frame rates in some games, the framerates are so high anyway bewteen the systems that the difference is negligible.

The question is does "negligible" warrant a complete system overhall at this time or would it be more prudent to wait for AMD's next gen and the subsequesnt drop in prices of the Conroe system components?

As a games player and office/browsing computer user, I see no reason whatsover in paying for a total upgrade to my system for what I believe is very little gain.

Ofcourse, for those of you that want the best benchmarking scores and use your computers for heavy multitasking and do not care about the cost I say go for it, but even if I used my computer in that way, I still would not upgrade as the benefits would still be negligible compared to a high end dual core AMD system. Being able to encode a few seconds faster or clock a little higher would still not motivate me to spend several hundred pounds.

But when the price drops, thats another story.

 

[TaKeN]

I'm feel exactly the same.


AMD Is more than fast enough for me and others at the moment,


Spending £400-£500 on something that is going to not be noticable in games is not the best way to spend my money at the moment

 

[w3bbo]

Conroe is the fastest CPU for games and applications, simple as. I envy those who have one but I for one simply cannot justify spending £500+ on a new setup when my current one runs the applications/games I use quicky and flawlessly. However, if I were to buy a system from scratch then conroe would be the route I would take.

Buying an FX62 now would be stupid and wasteful when a £150 conroe can easily match it and for the most part beat it hands down - even in gaming. However for a sizeable increase in FPS an updated GFX card would give better framerates than an upgraded CPU regardless of how quick they maybe.
.

 

[Vault5ChOnG]

Well ive updated my system from:

AMD64 4000+
2GB Geil PC3200
X1900XT-X 512MB
X1900CF Edtion 512MB

TO:

Intel E6700
2GB G-Skill PC25400
X1900XT-X 512MB
X1900CF Edtion 512MB

And ive almost tripled my fps, if i do com_maxfps 0 in games to see how much fps i can actually get up to i notice a huge difference.

From my last setup i was getting 450 tops and now:

www.v5-gaming.com/shot0000.jpg
www.v5-gaming.com/shot0002.jpg
www.v5-gaming.com/shot0003.jpg

Those are without AA or anything just my regular setup.

This one is with 1600x1200 on DX9 mode and x4AA and everyother setting up full.
www.v5-gaming.com/shot0004.jpg

As you can see its still a smooth framerate

 

[Sparky]

In your case, as your a gamer, then its better to spend the cash on a new graphics card.

That would give you more fps than just upgrading your cpu, mobo and ram, and would probably be cheaper.

Thats just what im about to do now.Was going to get a full Conroe system but after reading a few reviews it would make little difference to my games.A instead for me i think.x1900xt

 

[2bullish]

Well ive updated my system from:

AMD64 4000+
2GB Geil PC3200
X1900XT-X 512MB
X1900CF Edtion 512MB

TO:

Intel E6700
2GB G-Skill PC25400
X1900XT-X 512MB
X1900CF Edtion 512MB

And ive almost tripled my fps, if i do com_maxfps 0 in games to see how much fps i can actually get up to i notice a huge difference.

From my last setup i was getting 450 tops and now:

www.v5-gaming.com/shot0000.jpg
www.v5-gaming.com/shot0002.jpg
www.v5-gaming.com/shot0003.jpg

Those are without AA or anything just my regular setup.

This one is with 1600x1200 on DX9 mode and x4AA and everyother setting up full.
www.v5-gaming.com/shot0004.jpg

As you can see its still a smooth framerate

That looks great but you missed out your mobo. The question is do you see £500+ worth of difference? Also, with respect, changing your chip has given you 600fps increase, I find that somewhat hard to believe.

From all the reviews I have seen from respected sites, Anandtech, Tomshardware and HardOCP they have reported between 10 and 15% at BEST increase in SOME games. You are talking nearly 300% mmmmm I think not.

 

[manveruppd]

Now hang on a minute, I'm confused...

I thought that games became more rather than less CPU-bound as you upped the resolution, yet in those benchmarks by Anandtech they drop the resolution to 640x480 in order to compare pure CPU performance: http://www.anandtech.com/cpuchipsets/showdoc.aspx?i=2795&p=17

Have I got it wrong?

I'm aware that for games with intensive physics calculations and lots of on-screen objects such as flight simulators CPU speed would matter at any resolution - I'm talking about the majority of games here, although I'm guessing new games will be getting more and more CPU-intensive as time passes, since physics seems to be the new trend in games development.

 

[2bullish]

Now hang on a minute, I'm confused...

I thought that games became more rather than less CPU-bound as you upped the resolution, yet in those benchmarks by Anandtech they drop the resolution to 640x480 in order to compare pure CPU performance: http://www.anandtech.com/cpuchipsets/showdoc.aspx?i=2795&p=17

Have I got it wrong?

I'm aware that for games with intensive physics calculations and lots of on-screen objects such as flight simulators CPU speed would matter at any resolution - I'm talking about the majority of games here, although I'm guessing new games will be getting more and more CPU-intensive as time passes, since physics seems to be the new trend in games development.

It is in fact the GPU that is used the most not the CPU and future games will have more GPU power hence physics, where the calcs are done by the gpu. Lowering the resolution in tests ensures that the cpu is not limited by the test model gpu and can get a clearer picture of bench marks that are cpu specific.

Heres http://www.anandtech.com/cpuchipsets/showdoc.aspx?i=2802&p=11 the Anandtech article using Oblivion to show fps increase non dependant on gpu. It does, however, state that if they could they would up the gpu power rather than up the cpu power. Anyway I will let you see for yourself.

The fact is if you go major high end Conroe extreme you are going to see between 12 and 18% increase in frame rates, however, the cost is going to be astronomical, and can we tell the difference, honestly?

So the arguent is still, is it worth it at the moment, I would contend it is not, as not only would it cost several hundred pounds (in the case of the extreme conroes, yikes even more) to get this increase, if all the user does it play games and surf?

Not to mention what do you do with all your old kit, spend time flogging it on ebay?

In my case it is not, for the time being that is.

 

[Vault5ChOnG]

Motherboard ive got now is the MSI PowerUp Edition. Ive always had a part on the map where i go to test my fps, always the same part ive done to i know how much im gaining on little tweaks i do. the spot i usually goto i used to get 450fps and now with my new setup im getting 900/1000 fps. I did always think my AMD64 4000+ was holding back the power of my x1900 crossfire setup and this result maybe proves that or maybe its just one hell of a cpu lol.

 

[taz488]

well i think a cpu makes quite a bit of difference in games, and also what ive noticed is that comparing a slower clocked amd 64 to a higher but same performance p4 has reversed and its conroe now that has the slower clock speed but higher performance =] comparing conroe and athlon 64 x2's. but when amd bring out a competitor to conroe do you think it will be the same again? will amd bring out a dual core equivalent but better performance than conroe?

 

[manveruppd]

Lowering the resolution in tests ensures that the cpu is not limited by the test model gpu and can get a clearer picture of bench marks that are cpu specific.

Right, so I did have it back-to-front: I thought that the role of the CPU became more prominent at higher rather than lower resolutions, thinking that higher resolutions meant more data that the CPU had to prepare for the graphics card to render. The two anandtech articles reinforced this wrong impression, as, in the Conroe benchmarks, they tested at 1600x1200, whereas in the Oblivion article they tested at 1280x1024, and the differences between CPUs were far more dramatic at 1600x1200.
This is important to get straight because, judging from this thread, a lot of people are upgrading to Conroe thinking it will make a difference, whereas, depending on what resolution they game at, upgrading their GPU might have made a bigger difference. If someone's gaming on a 17" TFT at 1024x768 and can't go above that, the criteria by which they'd judge what's holding them back is different than someone on a 32" widescreen TFT.

 

[Vault5ChOnG]

you think maybe my last mobo wasnt running crossfire properly if im getting that much of an improvment?

Another thing ive noticed.

Remember all the horizontal lines people where complaining about with crossfire and crt moniters.. mines have gone since upgrading :/ lol

 

[ferrettoes]

Chong u foolje ;D
Anything above 333fps in cod2 u can't jump normally its about 3/4s the normal jump height so in reality u cant play without it being capped @ 333

 

[Fr0dders]

So the arguent is still, is it worth it at the moment, I would contend it is not, as not only would it cost several hundred pounds (in the case of the extreme conroes, yikes even more) to get this increase, if all the user does it play games and surf?

just having a read of the Anandtech Core 2 Duo overclocking article

the reason im considering an upgrade to Core, is that ive only got an old Winchester Athlon 64 3200. I cant get a stable frame rate in 1680 x 1050 in BF2 with 4 x AA and 16x AF

But the benchmarks done by others with faster processors but the same GPU clearly show a stable frame rate. So im guessing its my CPU thats holding me back

but have a look at this :

this clearly shows the CPU limitation that every1 is talking about with games. the top 3 processors are all within 2/3 fps of each other. obviously reaching the threshold of the CPUs limitations

the 2.0ghz Winchester isnt tested here, but look what is, a 3800+ 2.0ghz. And thats a full 40 fps behind an overclocked 6300.

Now surely its gotta be worth an upgrade for 40 fps ?