Are you happy with your upgrade to an E8500 or Q6600? I wanna know

[demiurge3141]

Originally posted by: Gillbot

Originally posted by: myocardia

Originally posted by: Gillbot
I went from a E6550@3.5GHz to a E3110@4GHz to a Q6600@3.4GHz to a E8500@4.25GHz to a Q9400@3.7GHz to a X3330@3.7GHz and in all honestly, I should have stopped at the E3110 or the E8500. I saw little to no gains with the quad core and small gains from the E6550 to the E3110. If I could have, I should have skipped from the E6550 right to the E8500 and stayed there.

Hah, you're pretty much a testament to why not to upgrade.


Sheibler, you're definitely CPU-bound, unless you're running that GTX260 @ 2560x1600.

I've been through the whole line front to back and back to front with certain OC/Performance goals in mind. Needless to say, I have been let down nearly every time with the exception of my trusty E6550 and then the nice E8500. I've had a few very nice quads run through my hands and when compared to the raw speed of the dual, the duals win in everything I need them for. Until more apps become utilized for quads, they are a marketing gimmick.

Originally posted by: OLpal
Unless you have special needs for a Quad , theni'd stick with a E8600 for the nice stuff you allready have in your system !!
Although if you can be assured of getting the newer stepping a E8400 or E8500 would be great also !!
New Stepping is the important part !!

These will run @ a much higher Mhz & cooler than a Quad !!

Exactly!

This is why I get so frustrated when I see people get overwhelmed with the typical "QUAD QUAD QUAD" recommendation. Unless you will really utilize the quad, there is no point in spending the extra $ for it. Do you think every average Joe needs a quad for Solitare, email and internet? Sure, many here are not the typical Joe but then again, many here also don't run F@H 24/7 either. A simple game every now and then, some email and of course, neffing on good old Anandtech!

What extra $? Last time I checked the 8500 and 6600 are the same price. I would recommend the duo a lot more if the quad is actually twice the price.

 

[Gillbot]

But you are comparing a 45nm Dual to a 65nm Quad. Why not compare the 45nm Dual to the 45nm Quad, you'll see that price difference jump. 65nm Chips are being phased out.

45nm pricing
$85 E5200: http://www.newegg.com/Product/...x?Item=N82E16819116072
$120 E7200: http://www.newegg.com/Product/...x?Item=N82E16819115052
$170 E8400: http://www.newegg.com/Product/...x?Item=N82E16819115037
$225 Q8200: http://www.newegg.com/Product/...x?Item=N82E16819115055
$260 Q9300: http://www.newegg.com/Product/...x?Item=N82E16819115043

65nm pricing
$120 E4600: http://www.newegg.com/Product/...x?Item=N82E16819115032
$140 E4700: http://www.newegg.com/Product/...x?Item=N82E16819115045
$190 Q6600 http://www.newegg.com/Product/...x?Item=N82E16819115017

Seems to be a rather nice spread in price to me. And I don't care if there's a sale here or there on a chip. If you want to go that route also, we can dig all day for the best deals in the FS/FT forums and debate until we are blue in the face. The FACT of the matter is, dual cores are CHEAPER.

 

[scheibler1]

not what we were comparing. We were comparing the E8500 and Q6600 which are the exact same price

 

[Gillbot]

And the Q6600 is STILL older technology.

 

[Avalon]

Well I went from an E2140 that was capped around 2.9ish to an E8400 that's sittin @ 4Ghz, and I've definitely noticed the difference, even in some games. It's pretty nice.

 

[soccerballtux]

Originally posted by: Gillbot

Originally posted by: soccerballtux
yeah yeah that's what everybody said back with the FX-55, then 6 months later games came out that made use of dual core, and the FX-55 couldn't keep up. Myocardia has one of those stories, he gave it just recently.

Not everyone upgrades hardware or software every 6 months. I waited till LONG after supreme commanders initial release to buy so the price would drop below the crazy $50+ price point. When it hit $20 on a sale, I finally grabbed it.

Right. I was just giving an example so the OP can decide for himself.

 

[demiurge3141]

Originally posted by: Gillbot
But you are comparing a 45nm Dual to a 65nm Quad. Why not compare the 45nm Dual to the 45nm Quad, you'll see that price difference jump. 65nm Chips are being phased out.

45nm pricing
$85 E5200: http://www.newegg.com/Product/...x?Item=N82E16819116072
$120 E7200: http://www.newegg.com/Product/...x?Item=N82E16819115052
$170 E8400: http://www.newegg.com/Product/...x?Item=N82E16819115037
$225 Q8200: http://www.newegg.com/Product/...x?Item=N82E16819115055
$260 Q9300: http://www.newegg.com/Product/...x?Item=N82E16819115043

65nm pricing
$120 E4600: http://www.newegg.com/Product/...x?Item=N82E16819115032
$140 E4700: http://www.newegg.com/Product/...x?Item=N82E16819115045
$190 Q6600 http://www.newegg.com/Product/...x?Item=N82E16819115017

Seems to be a rather nice spread in price to me. And I don't care if there's a sale here or there on a chip. If you want to go that route also, we can dig all day for the best deals in the FS/FT forums and debate until we are blue in the face. The FACT of the matter is, dual cores are CHEAPER.

Because those are the two chips the original poster is talking about.

 

[COPOHawk]

I am running a Q6600 at 3.3 GHZ...and love it. I do some gaming, but mostly a LOT of multi-tasking. My OS on this box is Vista Ultimate and it has a 150 GB Raptor. I have owned a lot of computers over the years, but this is the first one where I am happy with the performance of Windows...and it is Vista to boot...

I bought the Q6600 because it was the same price as the E8500...and don't regret it. When i spec out Dell workstations for customers, I always get the most bang for the buck...and this means the Q6600 these days. I have never had a complaint from someone who thinks that the quad core is "too good".

Who cares if it is older tech? It works well for a good price point. Pretty soon, the 45nm stuff will be "old tech" compared to the Nehalem stuff...

Buy the best tech available at your price point...

HTH....

 

[scheibler1]

Yea I got the Q6600 cause I couldn't really take advantage of the E8500 anyways. I would need a newer mobo and better ram.

The Q6600 should be a nice step up from my 3.33ghz E2180.

 

[apoppin]

i had e4300@3.33Ghz and went to e8600@3.33Ghz

now 3DMark06 will give you a good idea of scaling with a 4870x2 as only the CPU changed

14413 with e4300 @ 3.33ghz
16434 with e8600 @ 3.33Ghz
19395 with e8400 @ 3.99Ghz

you can draw a few conclusions about your own upgrade. e4300 has 2GB of L2 Cache and e8600 has 6GB [and architectural improvements, i believe]

O/C'ing brings further benefits

 

[RussianSensation]

Originally posted by: apoppin

now 3DMark06 will give you a good idea of scaling

I hope you are joking. There is no way your Crysis or World in Conflict gaming experience has improved by 35% those 3dmark scores are showing. 3dMARK is garbage. It bears no reflection on CPU's real world performance. Phenom can keep up with C2D in certain games but its 3dMARK cpu score are much lower...

 

[taltamir]

I always said the same about 3dmark. But lately I am getting the feeling it might be somewhat realistic and just getting bad credit. Actual gameplay is affected by microstutter and instantanous FPS drops, often caused by too weak a CPU. Inaccurage averaging used for even the min FPS (the lowest average of a whole second) is deceptive and tends to overestimate the capabilities of lower end CPUs.
You can always turn off some graphics and get more out of your GPU, you can never turn off some AI or physics to get more out of your CPU. (except with physX, where you run it on a GPU and CPU anyways)

 

[scrubman]

I went with the Q6600 last holiday and love it. I run at 3.5GHz (lucky for me) on air cooling and it feels good knowing that I wont need to upgrade until next year holiday at the earliest. I hope to get more use from the quad core during the next year or so! Actually, I did what my plan was when i built this last holiday and I just bought a second identical 8800GT to go SLI since the prices have dropped to about $100 now. That alone double my Crysis score to put me in the 40's on FPS. In other words, good choice!

 

[yyrkoon]

Originally posted by: soccerballtux
yeah yeah that's what everybody said back with the FX-55, then 6 months later games came out that made use of dual core, and the FX-55 couldn't keep up. Myocardia has one of those stories, he gave it just recently.

I dont know how to break this to you guy, but the reason a dual core performs better is because of all the little behind the scenes stuff the OS does(like copying data from disk to memory etc) is no longer bound to a single core. This means that the OS is no longer starved, or as starved for processor time. [EDIT] What I mean by this, is that the game, and the OS are no longer competing for CPU time(on a single core). Some games benefit more than others here, and games like quake(or anything IDSoft really), are extremely well muti-threaded, and basically an operating system in themselves. Still, even quake in all its multi-threaded splendor does not see a HUGE performance increase using multi-cores.

There are very few games out there right now that use more than 55% 2x CPU cores, and fewer still that will max both cores at 100%.

Also, I would recommend you look into Amdahl's law . . .

Anyways, if the OP is not a serious multi-tasker, he/she will not need 4 cores, and thus buying such will definitely NOT provide better performance compared to a good dual core. Especially if said dual core can clock a good bit higher. Not only that, you'll use less power.

 

[apoppin]

Originally posted by: RussianSensation

Originally posted by: apoppin

now 3DMark06 will give you a good idea of scaling

I hope you are joking. There is no way your Crysis or World in Conflict gaming experience has improved by 35% those 3dmark scores are showing. 3dMARK is garbage. It bears no reflection on CPU's real world performance. Phenom can keep up with C2D in certain games but its 3dMARK cpu score are much lower...

that is a ridiculous blanket statement; you doubt that OCing the CPU brings further performance benefits?
:Q

you seem to ignore the fact that 3DMark06 has 4 game tests that can be broken down into FPS increase and 3DMark is a decent tool to track changes within the same system

Of course the CPU score also increased, but so did the graphics system

CoJ @ 19x12 Fully maxed DX10 bench inc 4xAA/16xAF
x2/e8600@3.33Ghz - 20.2/116.6/53.8
x2/e8600@3.99Ghz - 28.8/128.2/65.4


LP @ 19x12 Fully maxed DX10 bench inc 4xAA/16xAF
Snow
x2/e8600@3.33Ghz - 45.3/23/64
x2/e8600@3.99Ghz - 46.6/24/64

Cave
x2/e8600@3.33Ghz - 57.6/35/71
x2/e8600@3.99Ghz - 61.6/41/71

HL2/LC @ 19x12 - maxed DX9 in 4xAA/16xAF
x2/e8600@3.33Ghz - 217.69/105/301
x2/e8600@3.99Ghz - 236.50/109/301


FEAR: Persius Mandate @ 19x12 - maxed DX9 inc 16xAF
with SS on - no/AA
x2/e8600@3.33Ghz - 72/140/436
x2/e8600@3.99Ghz - 84/145/498

SS off/4xAA
x2/e8600@3.33Ghz - 36/183/659
x2/e8600@3.99Ghz - 41/192/667

Any more questions about CPU scaling?

i have lots more benches to come and they all show it

 

[myocardia]

Originally posted by: RussianSensation
I hope you are joking. There is no way your Crysis or World in Conflict gaming experience has improved by 35% those 3dmark scores are showing.

You need to buy a new calculator, bub. 16,434 is 12.7% faster than 14,413. And just in case you were talking about the difference between 14,413 and 19,395, that's a 25.3% improvement. BTW, games absolutely love a fast CPU, and they love cache. When you add a ~17% higher clockspeed, plus 300% more cache, you're definitely going to have higher framerates in any game, as long as your GPU is fast enough to keep up with it.

Originally posted by: yyrkoon
I dont know how to break this to you guy, but the reason a dual core performs better is because of all the little behind the scenes stuff the OS does(like copying data from disk to memory etc) is no longer bound to a single core. This means that the OS is no longer starved, or as starved for processor time.

I hate to break it to you, guy, but the reason that games benefit from dual-cores has absolutely nothing to do with hard drive access. Nearly every game loads all of the data it needs from the hard drive before you start playing the game. That's why almost all games have "levels". They benefit from more than one core predominantly because video drivers are now multi-threaded, and the other things (besides disk access) that the OS is having to do, like DX calls, etc, can be performed on other cores besides core #0, giving a benefit to even sigle-threaded games.

 

[Petro89]

I upgraded yesterday!

Old system

Athlon 64 3400+ 754
2GB OCZ DDR PC3200
WD 500GB PATA
Radeon 1950pro AGP

To...

Q6600
4GB Corsair
Radeon 4870
Seagate 500 SATA

And I am happy. I dual booted XP and Vista on the old setup. XP was always sweet on that system but Vista was kinda slow. I'm not a huge gamer but even though I've been happy with the 1950, it has begun to show its age a bit.

I set up the 6600 system and installed everything. So far its fast and looks great. Benchmarks are about 30% faster across the board, and when I tinker with the CPU and o/c a bit that will be even better.

Whats funny is this is my first Intel based system in over 10 years since my old 233MMX. I tried out a Phenom and wasn't all that impressed which was disappointing . The Q660 definately seems faster.

So yes, I am happy.

 

[Petro89]

After a bit of tinkering I've got her happily up to 3ghz. No sweat! Looking for more!

Benches are now about 50-60% higher!!!!!

YES I AM HAPPY!

 

[RussianSensation]

Originally posted by: myocardia

You need to buy a new calculator, bub. 16,434 is 12.7% faster than 14,413. And just in case you were talking about the difference between 14,413 and 19,395, that's a 25.3% improvement.

Math 101

Method 1:

(16,434/14,413 - 1) x 100% = 14.02%

Method 2:
(19,395-14,413)/14,413 x 100% = 34.6%

 

[jeffw2767602]

4>2 have a good day

 

[scheibler1]

I'm not happy with the Q6600 I ordered. It's now for sale and I've got an E8500 coming...

 

[apoppin]

Originally posted by: RussianSensation

Originally posted by: myocardia

You need to buy a new calculator, bub. 16,434 is 12.7% faster than 14,413. And just in case you were talking about the difference between 14,413 and 19,395, that's a 25.3% improvement.

Math 101

Method 1:

(16,434/14,413 - 1) x 100% = 14.02%

Method 2:
(19,395-14,413)/14,413 x 100% = 34.6%

Not bad, for a simple OC, huh?



Nice improvement in games also