Intel Pentium Dual-Core E5200

[Kougar]

Thanks for the article, the results in the benchmark are good to show the differences of the Processors, but if we talk about budget it might not be the real numbers... Why would anyone who can buy a high-end board, ram, cooler, 64-bit OS, good graphics card, use a $80 Processor? The total price using a high-end Processor in the setup compared to the total price of using an E5200 is not big, considering someone will buy that kind of parts.

The goal of any reviewer is to build a test platform that ensures the component being tested is the bottleneck as much as possible.. It won't honestly make any tangible difference in reality but still it ensures the results are as CPU-dependant as possible.

Even building a budget platform the differences aren't going to be significant, and at least I don't see any reason to do so over simply building a "best-case scenario" platform. At least this way readers would know the upper bounds of the capabilities, which won't differ that much from a budget system as far as CPU testing is concerned.

 

[Rob Williams]

The goal of any reviewer is to build a test platform that ensures the component being tested is the bottleneck as much as possible.. It won't honestly make any tangible difference in reality but still it ensures the results are as CPU-dependant as possible.

It's bizarre... I just ran into this debate in a conversation just the other night. I also agree with Kougar... spot-on.

Throughout all of our CPU testing, we don't change our PC components at all, in order to have all of the results remain comparable. If we swapped out the motherboard just to properly suit the processor being reviewed, then we'd have no right to begin comparing results. When we stick with one machine, we see real performance results that can be compared. We want to be able to use just one platform that can handle any processor we receive. It's that simple.

 

[Unregistered]

Guys, thank you very much for great review!!!

I’m professional photographer and must emphasis that I especially appreciate your Lightroom tests! That is the main reason i visit your site on regular base!

But I have few assistants and can’t provide all of them with i7 920, so I just buy something on the budget and overclock as much as i can for everyday work. (I live in eastern Europe)
I would really appreciate even more, your tests, if you could add overclocked results in your charts! If you can't do it for all charts, do it for real world applications!

Thanks again!
Misha

 

[Rob Williams]

Hi Unregistered:

Thanks for the nice comments! Overclocking in general is one aspect of our content that I'd like to up a little bit, but I'm working to figure out how to best handle this. In some of our recent GPU content, I've included a page dedicated to overclocking, and then showed the differences in real performance between stock and the overclocked state, and I think I'll do something similar for our CPU content going forward.

Here's the problem. What overclock do I stick with for such testing? The CPU might hit 4.0GHz, but very few people are going to go that far, and in some cases, their chip may not even have the capability. It's a difficult task... we'd have to choose an overclock that we'd feel would be achievable by anyone who has the same chip, on most any motherboard. So if we have an i7-920, we'd likely use 3.6GHz, since a 180MHz BCLK is a realistic overclock.

I'm up for ideas on this from anyone who wants to throw their two cents in.

 

[Kougar]

That's a tough call... I'm biased simply because I do run a 4GHz+ overclock, but I can see where you're coming from. Best answer would be to see what most of your readers think on that I guess.

 

[Rob Williams]

I guess if time permits, we could just include performance for "modest OC" and "top OC", but in the end, that does take a fair amount of extra time. It's like benchmarking three CPUs instead of one. I could just focus on the real-world benchmarks though... nothing like SYSmark or the synthetics, since that would save a lot of time.

The most time-consuming thing would be finding a top overclock, and that's something I'll admit I have zero interest in. I'd never find the same top overclock that someone else would, sadly. I'd figure as long as I get a decent clock above stock, we could consider it a "top OC". Only hardcore OCers are going to care to go higher. I'm definitely a casual OCer in every sense of the word.

 

[Misha]

Hi Misha:

Thanks for the nice comments! Overclocking in general is one aspect of our content that I'd like to up a little bit, but I'm working to figure out how to best handle this. In some of our recent GPU content, I've included a page dedicated to overclocking, and then showed the differences in real performance between stock and the overclocked state, and I think I'll do something similar for our CPU content going forward.

Here's the problem. What overclock do I stick with for such testing? The CPU might hit 4.0GHz, but very few people are going to go that far, and in some cases, their chip may not even have the capability. It's a difficult task... we'd have to choose an overclock that we'd feel would be achievable by anyone who has the same chip, on most any motherboard. So if we have an i7-920, we'd likely use 3.6GHz, since a 180MHz BCLK is a realistic overclock.

I'm up for ideas on this from anyone who wants to throw their two cents in.

Hi Rob,

thanks for your respond, and sorry for my very late answer, but you got another reg. member!

I know that testing OC-ed processor would be doubled task, but i would test just serious, money making apps.
Like Adobe Lightroom, Premiere, 3D Studio Max, Cinebench 10 and so.... I really don't care for syntetic software.

I clocked my E5200 for 24/7 to 4.0 GHz and this was my best bet .

So, i guess you should choose to do some standard OC. I mean best bet for that processor. Or something what most people can do on most common motherboards!

I guess, it would be around 3.6 or 3.8MHz for my E5200!

You have experience more than us, and you could easily tell what can be achieved with 99% proc's on the market, so you could test that value. We would more than happy to see everything you test.

For 24/7 OC, some reasonably, average value is more that enough!

p.s.
English is not my native language, so sorry for typos and grammar I don't have any spell checker at hand by now.

 

[Rob Williams]

Hi Misha:

Welcome to the forums

First off, great overclock on that chip... even unstable that's an accomplishment!

You have experience more than us, and you could easily tell what can be achieved with 99% proc's on the market, so you could test that value. We would more than happy to see everything you test.

That's what I suspect should happen... test an overclocked setting that's easy to apply and uses reasonable bus clocks (~400MHz on Core 2, ~175MHz on Core i7). I would want to stress that they'd be 24/7 overclocks though, because I'm far more of a fan of those than anything on the extreme side.

As for benchmarking scenarios, I'm still in the process of figuring out all we're going to use for upcoming processor content... it's taking a lot longer than I ever anticipated. Adobe Lightroom will most certainly stay, although I wouldn't mind tweaking how we run that test, and perhaps even add a second test (using lots of filters). We also just received our copy of 3ds Max 2010, so it'd be great to find a new workload to use there as well. To date we've just used what was available on the sample's DVD, and it really doesn't appear like Autodesk likes to update those too often.

Adobe Premiere and other such tools... I'm unsure about. Getting the applications wouldn't be a problem, but getting a proper workload would be (I don't even own a camcorder), so I'm not sure what to do there. My immediate thoughts is to use a tool like TMPGEnc Xpress for most of our video tests, because with it, we can use a variety of codecs and filters. One test would be to encode a large video to mobile MP4, another to 720p DivX AVI, and a lossless gameplay video to 1080p WMV (this would be the longest test... WMV offers excellent quality compared to the others). Any further ideas would be great.

Perhaps this should all be moved to that particular thread though

http://forums.techgage.com/showthread.php?t=4922

 

[Unregistered]

E5200 on ASRock motehrboard

Hi all
I am very new to replacing computer components and the compatibility between them. I have the following motherboard:
http://www.asrock.com/mb/overview.asp?Model=775Dual-880Pro

- LGA775 for Intel® Dual Core Extreme Edition / Pentium® D / Pentium® 4 / Celeron® D processors
- FSB 1066/800/533 MHz
- Northbridge: VIA® PT880 Pro
- Southbridge: VIA® 8237R

,2x1GB DDR II RAM and Celeron 2.93MHz processor which I think is old for this age. I read your article about the E5200 and I was wondering if it can be used with my motherboard? I hope it could be, because the MB says that it supports Dual Core on 775 and the processor is of that type, but I need a pro advice before spending 80$ on the chip.
Thank you,
Mite

 

[Rob Williams]

Hi Mite:

Yup, that board is fine. The key feature to look for is that the board uses LGA775 processors. If it does, then the Pentiums will work. You might want to make sure you update to the latest BIOS, though, as those tend to add improved support for newer processors.

 

[therMite]

Thanks for the advice. Greetings from a Man Utd fan

 

[Kougar]

At risk of contradicting Rob, I have to chime in on this one... The last BIOS update predated C2D's, and the board is old enough the VRM's are likely the older versions Intel stated couldn't be validated for C2D's.

The E5200 has a slow 800FSB bus speed so it stands a good chance of working in that board. Just be aware there are no guarantees with that motherboard.

 

[Rob Williams]

You contradicted me, but it's a good thing, because I completely overlooked that. For whatever reason, I read "Intel Quad Core Extreme Edition", and with that, I knew that everything underneath it would work fine. That board doesn't actually state Core 2 support at all (of which Pentium E5200 is a sub-family of), so I'd actually have to recommend <em>not</em> taking the chance on the purchase. If the board doesn't state Core 2 support at all, it might not be worth the chance.

Thanks for the advice. Greetings from a Man Utd fan

Haha, you must be another lurker, huh? ;-)

 

[therMite]

You corrected yourself on time I just came back from searching for a processor from the nearby stores. Looks like nothing from the upgrade. A new system then.
Take care,
Mite

 

[Rob Williams]

You corrected yourself on time I just came back from searching for a processor from the nearby stores. Looks like nothing from the upgrade. A new system then.
Take care,
Mite

Apologies for my mistake... I cannot believe I overlooked that. Thanks to Kougar for noticing it.

Don't make building a new rig sound like a bad thing! Whatever you'll upgrade to will be a huge improvement over what you're using now. You certainly won't be looking back ;-)