Q9450

[pipjor]

I'm thinking of building a new workstation centered around the Intel C2 Quad 9450 microcenter is selling it for $299, anyone have experince with this processor??

can you recomend a mobo??

[Slinger]

I saw that price the other day and almsot pulled the trigger, but my Q6600 works great for me for now.

As for mobo suggerstion, any P35chipset will run it wonderful. If you want to get crazy and have the money, you can jump up to a X38/48 chipset, although they are still fairly new.

 

I have used Gigabyte and Abit boards before with not one issue. They have great bios settings for tweaking and overclocking if you wish to do so. Asus make great boards also but are a bit pricier. eVGAs are great but if you aren't gaming, not worth it.

 

I would recommend a Gigabyte P35 board. You can get them for less than $100

[vizfx]

I`m thinking at the same processor for my next system but I have another question:

Is really worth it to have DDR3 RAM or better with DDR2 which is not so expensive?

I`ve found some tests on TomsHardware but they`ve used the Core 2 Duo E8500 processor.

http://www.tomshardware.co.uk/toms-ultimate-ram-speed-tests,review-30648-11.html

From that point of view the increase is really small and translated into is not by far worth it.

„Our conclusion is very simple: you get the best bang for the buck if you stick to the mainstream of the memory market, which currently is still DDR2-800 or 1066, preferably at low latencies.”

But still the question remains since they didn`t use a Quad Core processor.

Best regards.

Mihai

[AJLynn]

I think those numbers sum it up quite well - the Cinema4D test and the Cinema4D animation render on the next page, and the Photoshop and Premiere tests, show very close to zero difference for different RAM speeds and timings. For games, the difference is a lot larger. (Remember that sites like these are usually written for gamers and will place more emphasis on games in their written analysis.)

 

They already tested single and dual CPUs and didn't show a difference between those when it comes to how much the RAM speed matters, and going to a quad isn't going to change things. Stick with the cheap RAM but make sure it's got some clock speed headroom - e.g., it's rated DDR2-800 and all your CPU needs is DDR2-667 or 533, and the latency numbers are reasonably low.

 

The 9450 is quite good, probably worth the $100 over what you can find the 6600 for, BUT under some circumstances it can be more difficult to overclock. The multipier is 8, not 9, so your MB and RAM need to support higher RAM/FSB speeds to achieve the same overclock.

[vizfx]

The processor is pretty good since is near QX6850 in tests that matter for us

http://techgage.com/article/intel_core_2_quad_q9450_266ghz/6

http://techgage.com/article/intel_core_2_quad_q9450_266ghz/7

Some notes about overclocking

http://techgage.com/article/intel_core_2_quad_q9450_266ghz/13

[vizfx]

One note about overclocking the Q9450. They`ve used DDR3 which is a bit expensive but with stable results at 3,2 ghz

[Slinger]

I think those numbers sum it up quite well - the Cinema4D test and the Cinema4D animation render on the next page, and the Photoshop and Premiere tests, show very close to zero difference for different RAM speeds and timings. For games, the difference is a lot larger. (Remember that sites like these are usually written for gamers and will place more emphasis on games in their written analysis.)

 

They already tested single and dual CPUs and didn't show a difference between those when it comes to how much the RAM speed matters, and going to a quad isn't going to change things. Stick with the cheap RAM but make sure it's got some clock speed headroom - e.g., it's rated DDR2-800 and all your CPU needs is DDR2-667 or 533, and the latency numbers are reasonably low.

 

The 9450 is quite good, probably worth the $100 over what you can find the 6600 for, BUT under some circumstances it can be more difficult to overclock. The multipier is 8, not 9, so your MB and RAM need to support higher RAM/FSB speeds to achieve the same overclock.

The problem with those test is they are testing DDR3-1333 and DDR3-1066. Go look at benches with some DDR3-2000. You will see a huge improvement but even with the higher performance, I can't justify dropping $500 for 2gigs of DR3.

 

When it goes main-stream, DDR3 will be well worth the money and the performance boost will be wonderful.

[AJLynn]

What do you mean? They included DDR2 and DDR3 in a wide range of speeds and came up with those numbers. If you look at the spreads in the render tests, you can't see a pattern that says "DDR3 is better when rendering" - the differences you do see are 1-2%, which isn't statistically significant, and it doesn't even seem to correlate to the RAM type.

 

If DDR3-1333 RAM can't beat DDR2-667 RAM by more than 1%, I don't see why going from DDR3-1333 to DDR3-2000 would be significant except that you have to keep increasing the FSB speed to keep pace with the CPU.

[Slinger]

There are no benchmarks yet showing DDR3-2000 running relative to the needs of a renderer...but there are test showing write/read etc speeds with DDR3-2000 and they are alot of faster then DDR2 memory is...go look at xtremesystems site and you will see what I mean.

 

 

My main point is why would you invest in the low speed DDR3 when you will still need a higher=higher priced motherboard? Your average user will not be going this high-end route to begin with...only your true enthusiast. With that said, if someone is going to benchmark DDR2 vs DDr3, get the upper-end speeds of DDR3......why compare the low-end speeds of DDR3 to the high-end speeds to DDR2? The real benefit is that you can get ram that will do 2000mhz now.

[AJLynn]

Right, but what I'm saying is that if going from DDR2 to DDR3 and adding 667MHz does almost nothing, why would we expect that adding another 667MHz will have a large effect? It will give you faster memory write/reads but it looks like what the tests here are showing is that those speeds are not very important for rendering.

[vizfx]

Hi

Finally I have just one question:

Is it worth it to push the system (based on Q9450) with overclocking? The performance is noticeable as seen in the previous comments but how about stability in time and lifetime of the components.

In this case the best memory available is a must.

I`ll keep you informed about my experience with my new system…. It will be mine next week.

Best regards.

Mihai

[pipjor]

i probally wont overclock my new system(should have it in about a week) why push it and risk stability.

 

my 2cents on overclocking is it's cool for gamers with alot of free time to test and crash their systems and to post(pissing contest) their faster speeds and lower temps online but for us folks in production stability is much more important than being .3mhz faster than the next guy...if i really wanted a faster cpu than the Q9450 i would have just bought it...

[Slinger]

Actually you wouldn't b/c there really isn't anything much faster than the Q9450

[vizfx]

Hi again.

The configuration is almost ready:

- processor Q9450

- mainboard Intel Bontrail INBLKDX38BT

- videocard - Gigabyte - PCIEX16, 512MB, GDDR3, GF9600GT

- RAM - 2GB PC8500 DDRII 1066

- HDD – 500GB WD

Unfortunately Nvidia 9800 was a bit too expensive for me.

Best regards.

Mihai