New E8400

[hooyaa]

I was looking at the new Intel Core 2 Duo E8400 and I have some questions about it. Why is it ~$55 less than the E6850 when they have the same speed and FSB, but the E8400 has a larger level 2 cache? If I read correctly the E8400 has a 6MB level 2 cache while the E6850 has only 4MB?!?

What have I missed here....smaller core....?

[splave]

prices just havent dropped on the e6850 and they are trying to make money on them still.

[supershanks]

Got mine yesterday, spart from issues with the mobo , now sorted it rocks
Intel E8400 Wolfdale Throws out a decent temp yjough, glad i'm on water.

[gfourth]

Nice link supershanks! That makes me feel better about my temps.

EDIT: I am hitting roughly 60c at the same vcore.

[supershanks]

there's a lot of people are carried away by the hype on these and are suprised at how hot they run, i've had people telling me the temps can't be right

luck

[splave]

I idle at 39/42 at 1.4v w/tuniq Nice results and low volts though!

[Zenoth]

@ supershanks

I've just read about all pages of your GA-X38-DQ6 blog, and it is awesome! Thank you very much. There's so much details in there that it over shadows any other "professional" reviews of that board I've read so far. If they could all be like that. Simply great job.

[kntgsp]

Originally Posted by supershanks

there's a lot of people are carried away by the hype on these and are suprised at how hot they run, i've had people telling me the temps can't be right

luck

To which I have no doubt is due to one of the following:

1. Bad chip
2. Poorly applied thermal paste
3. Using a stock cooler while OC'ing
4. Setting the Vcore way higher than it needs to be.

I picked mine up yesterday and have it running at 4.0Ghz 24/7. Idles at 31/27 and after an hour of Prime95 the hottest i saw was 55C. I also run at 1.34V.

[kntgsp]

Also, the E8400 is totally worth the hype. I switched from a P35 board with the Q6600 @ 3.0Ghz to an X-38 with the E8400, and the E8400 blows it out of the water.

There was some review saying that the Q6600 would be faster with WinRAR because the program is multi-core aware. Total opposite. The E8400 even at stock speeds was destroying my Q6600 in WinRAR, let alone when it was clocked up to 4.0Ghz. And it's still the best bang for your buck on the market. $189 for 4.0-4.4 on AIR! Even when the quad 45 nm chips drop, they're going to be priced much higher than the E8400. Shoot, at $189, I can afford to run the thing at 4.8Ghz and blow through one every month. Unless you need the four cores for extremely intense video editing, this is your best bet. Particularly for gaming. Yea there are a few games that "claim" to be four-core aware, but the performance difference they claim are non-existent in real world trials. Dual core is still the best bet for general computing and gaming.

When there actually IS an advantage to having a quad core for a general user/gamer, the current ones will be obsolete and all that money will have been wasted. The E8400 users, they're only out $189

[Zenoth]

There's too few games using three to four Cores, but I know of three that can do it, namely Supreme Commander (not much, but it does), Unreal Tournament 3 (supposedly it does use four Cores constantly, not all of them at 100% though, of course) and finally Lost Planet (the gains for that one cannot be denied, they are significant, measured in the 15+ extra FPS due to the presence of two extra Cores, thanks to a proper use of four threads instead of two on dual Cores).

Despite the fact that games can technically benefit from quad Cores, there's still too few of them, and the ones that do use three or four Cores aren't necessarily the best of the best games out there. So far my E8400 has served me very well, and it's basically 2x to 3x faster in pretty much anything I do with it than it was with my previous X2 4400+. Even in 3DMark06, my CPU score has gone up from 1,796 to between 3,000 to 3,800 (depending on its frequencies). Simply going to a superior architecture while staying in the dual Cores world increased the performance in my games to a tremendous amount (especially Source-based games such as Team Fortress 2 and RTS games in off-line mode against many A.I opponents).

[kntgsp]

But you still have to take into account bang for your buck. For a quad core with a clock speed equal to a dual, you're paying 4x as much. A Q6600 at 2.4 runs Lost Planet slower than an E8400 at 3.0. I've owned both and seen it first hand. Add in the fact that the E8400 is even cheaper than the Q6600 and easily clocks up to 4.0Ghz, there's just no reason to grab a quad core. Even the new 45nm quads are priced far above the E8400. And even if you do shell out the hundreds extra for a higher clocked 45nm quad, the performance gains you see will be negated by an E8400 that's running at 4.2.

Sure there will be faster processors than the E8400 at stock speeds, but you'll be paying out the *** for a negligible performance increase. The 4 frames per second you gain with a quad can be achieved by the E8400 by simply overclocking it. And they run incredibly cool at 4.0+ speeds in comparison to the quads.

[Zenoth]

I must be honest though, If I would have had the money, I would have bought a Q9450 in February instead, because I do lots of Photoshoping, video/audio encoding/decoding, not to mention that I mod a few games like Crysis and Source-based games, and both editors, namely Hammer and SandBox 2 can easily, and I do mean easily use all four Cores at 100% when compiling (which was taking up to an hour with my previous X2 4400+, but I still haven't tested with my E8400).

Basically, I went with an E8400 because it wasn't expensive. The Quads still are, for my wallet that is. Because, let's be honest here, if no one here had any financial limits/problems, I'm pretty sure that they'd all go for a Q9450... heck, why not a Q9650 while we're at it. I think that when it comes to PC hardware, the peace of mind is priceless.

[kntgsp]

You upgraded from a 90nm X2 4400+ which runs at 2.2Ghz. to a 45nm chip that runs at 3.0Ghz natively.

You'd be faster upgrading to a banana. What's your point? That's like me saying I upgraded from a Pentium D to a Penryn Core 2 Duo.

You could have upgraded to ANYTHING and it'd be faster. The Toledo's are 3 years old. At least compare the E8400 to something in its class.

[Zenoth]

Originally Posted by kntgsp

You upgraded from a 90nm X2 4400+ which runs at 2.2Ghz. to a 45nm chip that runs at 3.0Ghz natively.

You'd be faster upgrading to a banana. What's your point? That's like me saying I upgraded from a Pentium D to a Penryn Core 2 Duo.

You could have upgraded to ANYTHING and it'd be faster. The Toledo's are 3 years old. At least compare the E8400 to something in its class.

How can I compare it to something I've never owned?

And define "in its class". Well of course I'll compare it to my previous CPU, I can't compare it to another C2D since I never owned any of those until I bought my E8400. And there's no CPU directly within the E8400's class, as far as I know the previous C2D's were 65nm, didn't support SSE4, consumed more energy, are overall 5% to 10% slower at the same frequencies, radiated more heat and are still a little expensive now that the E8400 is there. So there's no technical comparisons possible for "apples to apples", in my opinion.

And let's not exaggerate, upgrading to "anything" wouldn't have meant having a faster CPU. I mean would "upgrading" to a VIA C3 be faster? You're generalizing a little I think. I know what you're trying to say, but when you think about it twice there's really no point to comparing the E8400 to even any of the 65nm C2D dual Cores. In gaming they might well be about the same, but when it comes to SSE4 applications the E8400 gains around 25% to 50% performance increase... and why would we have to ignore SSE4 for comparison purposes? It's there, so let's consider it.

Anyway... All I was saying is that a mere architecture change increased the performance by more than 2x in many applications without increasing the number of Cores. And that's about it, I wasn't trying to prove anything... I don't really get the point of your remark.

[kntgsp]

Anyway... All I was saying is that a mere architecture change increased the performance by more than 2x in many applications without increasing the number of Cores. And that's about it, I wasn't trying to prove anything... I don't really get the point of your remark.

Well in that case, our entire argument is moot since I agree with that. Haha.

And yes I was exaggerating, it's just force of habit having to defend this great CPU against people who claim it runs at 50C, 1.5v, etc, etc. Too many people don't know how to properly apply thermal paste and coolers that a lot of the first impressions of CPUs tend to be a bit skewed.

[splave]

LoL they do run hot. Im idling at 39/42 at 1.4v on a tuniq 120. I lapped the CPU and the HS. I can bench at 4.41ghz and run it at 4.1ghz normally and I cant complain at all. Best chip I have owned but dont go saying they arent hot because its way hotter than the e4400, e2160, e6600, e6420, and e6750 I have used.

[Zenoth]

kntgsp, I understand.

But you should not blindly believe that because, for example, you "properly" applied thermal paste that it automatically makes this chip run cool whatever cooling it uses. Those new 45nm beasts do run hot. Not always, of course, but keep in mind that there's some hardcore over-clockers out there doing everything "properly" and still end up with messy temperatures, some of which far exceed any expectations.

Just take a look at those threads:

http://www.ocforums.com/showthread.php?t=545708
http://www.ocforums.com/showthread.php?t=544848
http://www.ocforums.com/showthread.php?t=544212
http://www.nvnews.net/vbulletin/showthread.php?t=106522

In my case, with "properly applied" AS5, with the stock Intel cooler, running at 100%, my E8400 idles at more or less 40ºC, and reached as high as 56ºC at load. And that was at 3.2Ghz (stock voltage). Then I installed my OCZ Vendetta one day later, same frequencies, I used isopropyl alcohol to remove the paste, and re-apply it for that new cooler, not lapped though. Still at 3.2Ghz, idles at more or less 37ºC, loads at approximately 51ºC (I saw 49 up to 54, it depends on the ambient temperatures).

Yes, people with liquid cooling can have better results, I am aware of that. And yes, people using phase cooling too. And yes, people using high-end air cooling from Thermalright, Thermaltake or Tuniq, etc, indeed can get better results. But, from what I've read, whatever cooling you're using, those chips can run hotter than "expected". In fact a lot of people are very disappointed by the temperatures (primarily idle temps, of course) despite being generally impressed by the over-clocking potential (and even then, you need good Memory and a decent Motherboard).

[comc49]

i upgraded from p4 630 1gigs of ram to core 2 quad 2 gigs of ram. and i don't feel a difference

[BabyGorilla]

Since when do smaller fab chips requiring less power run warmer? Aside from the few anomolies you posted, they seem to run about the same or less than e6750.

Check out XS, I dont believe anyone there had this issue.

WAIT WAIT WAIT...are these shipping with concave IHS again?

Originally Posted by comc49

i upgraded from p4 630 1gigs of ram to core 2 quad 2 gigs of ram. and i don't feel a difference

ummm.....ok. Did you get a e8400 or did you accidently post in this thread??

There is a world of difference between the rig you had to the one you upgraded to...

[MrObvious]

Originally Posted by comc49

i upgraded from p4 630 1gigs of ram to core 2 quad 2 gigs of ram. and i don't feel a difference

There is a difference. Maybe your hard drives are slower or you're swapping a lot or something.