Comparing the 45nm Quad Cores

[OLpal]

Read this review on TechWare Labs , here's link.... Inside HW did the article itself !!

http://www.insidehw.com/Review...CPUs-Side-By-Side.html

Makes for fast comparison of the new Quads & how they do on various tests !!
Hope you find this interesting !!

Ol'Pal

 

[myocardia]

That's a good idea. I just wish a reputable site* would do an article like that.



* A reputable site would 1) know that a QX9650 is not the fastest CPU available today, and 2) wouldn't use a slower video card than you and I have, for their gaming benchmarks.

 

[OCGuy]

Originally posted by: myocardia
That's a good idea. I just wish a reputable site* would do an article like that.



* A reputable site would 1) know that a QX9650 is not the fastest CPU available today, and 2) wouldn't use a slower video card than you and I have, for their gaming benchmarks.

100% agree. How could you ignore a native 1600fsb chip? QUE?

 

[OLpal]

Good of you to point that out Myocardia, as i didn't catch the Gts8800 or the Cas 5 memory till you had me take a 2nd. look @ it !!
Although i feel this is nice real world setup to use just for comparison purposes !! As there are alot of people who don't want to spend $200 + $$$ for a video card.. Of course most of them wouldn't be considering a Q9450 to start with !!
Still i commend InsideHW for the time & effort they exerted to bring this comparo to everyone !!

Ol'Pal

Originally posted by: myocardia
That's a good idea. I just wish a reputable site* would do an article like that.



* A reputable site would 1) know that a QX9650 is not the fastest CPU available today, and 2) wouldn't use a slower video card than you and I have, for their gaming benchmarks.

 

[tamaron]

Thank you man! I have been waiting for this long time ago... It seems the q9300 is doing it very well, almost identical points with the underhertzed qx9650. But not in games 20%-30%!!!!!

 

[Idontcare]

Originally posted by: OLpal
Still i commend InsideHW for the time & effort they exerted to bring this comparo to everyone !!

Ol'Pal

Great find OLpal, I agree at least someone FINALLY got of their ass and did a somewhat-less-than-comprehensive review of these varying cache-sized quads.

Thanks for posting it, I read it and was actually surprised to see that for some things the 12MB cache boosted performance ~10% over the 6MB. To me that is significant.

But I'm with myocardia on this one...wtf are the reviews anandtech? These parts have only been out for how many months now? Maybe they are waiting for P45 or some such?

 

[Foxery]

Actually, I'm impressed by how small the cache size differences are for most tests, purely looking at the columns for the Q9300 and an underclocked QX9650. The games benefitted quite a lot, but media encoding and synthetics pretty close.

I haven't thought much about it since getting Folding@Home stable, but I'd be happy to run a few informal tests next to someone with a Q9450 on a P35-chipset system.

 

[OLpal]

Foxery that would be very cool of you !!! But i'm sure myocardia would agree , there has to be some levelness of the playing field [some don't think a 8800gts or ati 3850 , worthy of high end sys] Thanks for chiming in as your system just goes to Help prove my point to myocardia that not everyone runs a super-duper video card on their systems !!

But my contention is Inside HW did us a great service by doing this comparison & using the same video, memory, MB, etc... on each of the tests !! So we can tell how much the cache difference & multipier makes on some common benchmarks !!

To them i say !! Thanks Inside HW
.
Ol'Pal


ps; if you can run similar tests with your ati 3850, i think theat would be cool.. STd clock & OC'd if you can !! I for one would like to see your results..

Originally posted by: Foxery
Actually, I'm impressed by how small the cache size differences are for most tests, purely looking at the columns for the Q9300 and an underclocked QX9650. The games benefitted quite a lot, but media encoding and synthetics pretty close.

I haven't thought much about it since getting Folding@Home stable, but I'd be happy to run a few informal tests next to someone with a Q9450 on a P35-chipset system.

 

[bryanW1995]

ok, my 9450 is already running at 3.4, so we can compare. It's not a "perfect" comparison as your fsb is obviously higher, but it should be pretty close. for starters, go run graysky's x264 test. I've also run cinebench recently, it's still on my desktop.

took me 1:30 to complete my run, ip35E mobo, mem at 4-4-4-12 ddr2 850.

 

[myocardia]

Originally posted by: OLpal
As there are alot of people who don't want to spend $200 + $$$ for a video card.

That's not the point. Don't you think all of the CPU's would have had identical FPS, if they had used this card? That's why you always use the fastest card available, when doing CPU comparisons, that include FPS benchmarks.

 

[OLpal]

No really myocardia, i do see & understand your point !!
1st the 8800gts is not that bad of a card as isn't foxerys... [using the cas 5 memory does throw me though] They do however represent what the Avg. builder might put in their systems or a ati 3870, gt8800, 9600gt or the like !! [even the 8800gs isn't too shabby] As for me, i'd probably buy another 8800GT in a Heartbeat...

I guess what i'm saying [while seeing your point] this is a realistic expectation of what most avg users will see with these particular processors.... & i think it's a fair and very HELPFULL comparison guide for us all !! To assist in our decisions based on performance !!
I'd like to see them take this to another level [include the E8400] overclock them to within stock voltages..
ie;
E8400 to 3.6ghz
q9300 to 3.2ghz
Q9450 to 3.4ghz
ect, ect, ect, & see how they benchmark against not only each other but their stock settings could all be compared on this baseline system !!

Sound Good To You !!??? Ol'Pal

@ BryanW1995 this could prove interesting & satisfy myocardias request for better components, i hope you guys post & compare your results here so we can keep track & compare them .. Foxery & the rest of you, does that sound like a good idea to you !!
Who knows anand & gary key might get on board & do 1 of their super comparo articles on these chips !!

Originally posted by: myocardia

Originally posted by: OLpal
As there are alot of people who don't want to spend $200 + $$$ for a video card.

That's not the point. Don't you think all of the CPU's would have had identical FPS, if they had used this card? That's why you always use the fastest card available, when doing CPU comparisons, that include FPS benchmarks.

 

[bryanW1995]

if you want to compare a Q9450@3.6 vs a Q9300 @3.4 I can oblige there, too, though my x3350 is at my house and will be unavailable until sat night. well, unless I can convince the wifey to run the bench for me...hmmmm, I can see how that conversation will go:


bryan: hey, uh, baby, can you, um go run graysky's x264 encoding test for me. And, um, maybe some cinebench?
wifey: what, I can't hear you, both the babies are crying. hey, can you grab some tape, glue, rubber bands, diapers, and tampons on the way home friday night?
bryan: :|:| never mind

 

[OLpal]

Those benchies would be most welcomed !!
Cute conversation with wife !! How old are your young'ns ??
Mine usually says "" My puter is working just fine, so don't mess with it"
Will be interesting to see how many will do these tests , so we can have our own little guide to the universe of Quads !!
Temps & coolers would be appreciated also guys !!
Ol'Pal

Originally posted by: bryanW1995
if you want to compare a Q9450@3.6 vs a Q9300 @3.4 I can oblige there, too, though my x3350 is at my house and will be unavailable until sat night. well, unless I can convince the wifey to run the bench for me...hmmmm, I can see how that conversation will go:


bryan: hey, uh, baby, can you, um go run graysky's x264 encoding test for me. And, um, maybe some cinebench?
wifey: what, I can't hear you, both the babies are crying. hey, can you grab some tape, glue, rubber bands, diapers, and tampons on the way home friday night?
bryan: :|:| never mind

 

[bryanW1995]

ok, my two results on the Q9450 @ 3.4, xp pro sp2 (in signature)


Results for x264.exe v0.58.747
encoded 1442 frames, 74.01 fps, 3904.67 kb/s
encoded 1442 frames, 20.47 fps, 3952.97 kb/s

Results for x264.exe v0.59.819M
encoded 1442 frames, 80.67 fps, 3887.37 kb/s
encoded 1442 frames, 23.05 fps, 3963.57 kb/s


cinebench:
3792: 1 cpu
13205 : all cpus
multiprocessor speedup: 3.48

 

[Foxery]

Initial results, Q9300 @ 3.4 GHz with my normal 5-5-5 timings (I'll try other settings another day)
Long logs chopped to the best of multiple runs
Minimum idle core temp: 40C

Cooler is a Xigmatek S1283, very similar design to Bryan's Tuniq.
Motherboard is an ABit IP35-Pro. Only difference from Bryan's is a few extra SATA ports and a 2nd Ethernet controller. Yay.

*****
Graysky's X264
*****

Results for x264.exe v0.58.747
encoded 1442 frames, 73.07 fps, 3904.55 kb/s
encoded 1442 frames, 20.43 fps, 3952.83 kb/s

Results for x264.exe v0.59.819M
encoded 1442 frames, 79.15 fps, 3887.61 kb/s
encoded 1442 frames, 22.96 fps, 3962.92 kb/s

Highest core temperature recorded: 57C, per RealTemp v2.48


*****
CineBench R10
*****

Rendering (Single CPU): 3746 CB-CPU
Rendering (Multiple CPU): 13008 CB-CPU
Multiprocessor Speedup: 3.47
Shading (OpenGL Standard) : 6621 CB-GFX

Highest core temperature recorded: 55C, per RealTemp v2.48

*****
Additional System Details
*****
Specification Intel(R) Xeon(R) CPU X3320 @ 2.50GHz
Codename Yorkfield
Core Stepping M1
Technology 45 nm
Stock frequency 2500 MHz
Core Speed 3405.0 MHz (7.5 x 454.0 MHz)
FID range 6.0x - 7.5x

Northbridge Intel P35/G33/G31 rev. A2
Southbridge Intel 82801IR (ICH9R) rev. 02

CAS# 5.0
RAS# to CAS# 5
RAS# Precharge 5
Cycle Time (tRAS) 15
Command Rate 2T [ed: what? I set this to 1T in the BIOS!]
Memory Frequency 454.0 MHz (1:1)
Memory Type DDR2
Memory Size 2048 MBytes
Channels Dual (Symmetric)

Windows Version Microsoft Windows XP Professional Service Pack 2 (Build 2600)

Voltage sensor 0 1.31 Volts [0x83] (CPU Core)

 

[bryanW1995]

don't see much difference there...

edit: just edited my x264 logs to only take highest # as foxery did...earlier I was taking just the first # of each run. slight diff now ~ 1.5% on both tests, but I'm running 4-4-4-12 ddr2 850. wait, he's gotta be running at least ddr2 908...

we really need aigo or yoxxy to clock his 9650 down to the EXACT same speeds as foxery (7.5x, 454fsb). only problem there is that my mobo is nearly identical, and my x3350 mobo IS identical...grrrrr...

edit: @foxery, you can't run 1t on a p35 mobo. I tried with my ip35 pro with the same results

 

[OLpal]

C'mon Aigo & Yoxxy join the Comparo with Foxery & BryanW1995 to give us some more to compare to & analyse !!
You guys did good !! Hope more chime in to contribute !!
Do any of you guys have a E8400 OC to 3.6ghz that can run this X264 & cinbench ??

Couple of questions for you guys though !!
is this X264 bench something you have to buy to run !??? will it run on vista64 ?
Is it a suitable tool for us to be using to make decisions on which processor to buy ??
Why newegg & other main stores don't have Q9450 but they're on Ebay from places you've never heard of ???

Look forward to more benchies !! Ol'Pal

Originally posted by: bryanW1995
don't see much difference there...

edit: just edited my x264 logs to only take highest # as foxery did...earlier I was taking just the first # of each run. slight diff now ~ 1.5% on both tests, but I'm running 4-4-4-12 ddr2 850. wait, he's gotta be running at least ddr2 908...

we really need aigo or yoxxy to clock his 9650 down to the EXACT same speeds as foxery (7.5x, 454fsb). only problem there is that my mobo is nearly identical, and my x3350 mobo IS identical...grrrrr...

edit: @foxery, you can't run 1t on a p35 mobo. I tried with my ip35 pro with the same results

 

[OLpal]

Hey Guys here is the official Graysky X264 page & it has lots of users reporting their results..
You'll have to explain to me how your results fall into which column though !!

http://www.techarp.com/showart...tno=442&pgno=0#Results

Ol'Pal

 

[Foxery]

Both programs output their results to a text file, so I just copied'n'pasted, trimming out some irrelevant stuff.

We only ran 2 programs, but I'm intrigued that the difference is so trivial. (~2%) We could try to line up our RAM timings and multipliers, but frankly, this is a direct comparison between the CPUs at the same frequency. The rest of the settings are simply what you have to use to get that CPU clock.

A few answers for OLpal:

Originally posted by: OLpal
Do any of you guys have a E8400 OC to 3.6ghz that can run this X264 & cinbench ??

You do

is this X264 bench something you have to buy to run !??? will it run on vista64 ?

CineBench is also free. Download from http://www.maxon.net
I think I saw a 64-bit executable in the same package.

Is it a suitable tool for us to be using to make decisions on which processor to buy ??

Only if you happen to want this chip for image rendering! Otherwise, it's pretty limited in scope. It's just one example of a fairly taxing application that review sites like AnandTech sometimes use.

Why newegg & other main stores don't have Q9450 but they're on Ebay from places you've never heard of ???

EBay is always full of people who hoard hot items and resell them for a profit when the supplies to real retail stores dry up.

 

[aigomorla]

Originally posted by: myocardia
That's a good idea. I just wish a reputable site* would do an article like that.



* A reputable site would 1) know that a QX9650 is not the fastest CPU available today, and 2) wouldn't use a slower video card than you and I have, for their gaming benchmarks.

actuallly....

ruby and i can go 1 on 1 on this statement with you.


The QX9770's have been at best "disapointing". Of course were factoring in average overclocking.


Anyhow, these girls scaled @ 4ghz is no joke. Heres my HD Benchmark, however my system was running a tad bit slow by looking at the delta between the first and second test.

x264 HD BENCHMARK RESULTS
Please copy/paste everything below the line into the forum post to report your data
^^^^^^^^^^^^^^^^^^

Results for x264.exe v0.58.747
encoded 1442 frames, 83.15 fps, 3904.62 kb/s
encoded 1442 frames, 82.99 fps, 3904.62 kb/s
encoded 1442 frames, 83.29 fps, 3904.62 kb/s
encoded 1442 frames, 83.44 fps, 3904.62 kb/s
encoded 1442 frames, 24.11 fps, 3953.18 kb/s
encoded 1442 frames, 24.11 fps, 3953.18 kb/s
encoded 1442 frames, 24.04 fps, 3953.18 kb/s
encoded 1442 frames, 24.03 fps, 3953.18 kb/s

Results for x264.exe v0.59.819M
encoded 1442 frames, 90.04 fps, 3889.32 kb/s
encoded 1442 frames, 90.30 fps, 3887.16 kb/s
encoded 1442 frames, 90.48 fps, 3889.28 kb/s
encoded 1442 frames, 90.31 fps, 3888.84 kb/s
encoded 1442 frames, 27.04 fps, 3962.83 kb/s
encoded 1442 frames, 27.08 fps, 3963.41 kb/s
encoded 1442 frames, 27.07 fps, 3962.58 kb/s
encoded 1442 frames, 27.06 fps, 3963.07 kb/s


System Details
--------------
Name Intel Core 2 Extreme QX9650
Codename Yorkfield <--- NO its called persephone!
Specification Intel(R) Core(TM)2 Extreme CPU X9650 @ 3.00GHz
Core Stepping C0
Technology 45 nm
Stock frequency 3000 MHz
Core Speed 4250.2 MHz (10.0 x 425.1 MHz)
FID range 6.0x - 10.0x

Northbridge Intel X38 rev. 01
Southbridge Intel 82801IR (ICH9R) rev. 02

CAS# 5.0
RAS# to CAS# 5
RAS# Precharge 5
Cycle Time (tRAS) 12
Command Rate 2T
Memory Frequency 425.1 MHz (1:1)
Memory Type DDR2
Memory Size 3070 MBytes
Channels Dual (Symmetric)

Windows Version Microsoft Windows XP Professional Service Pack 2 (Build 2600)

max VID 1.250 V
Voltage sensor 0 1.41 Volts [0x58] (CPU VCORE)
Number of processors 1
Number of threads 4
Number of threads 4 (max 4)
L2 cache 2 x 6144 KBytes, 24-way set associative, 64-byte line size
Instructions sets MMX, SSE, SSE2, SSE3, SSSE3, SSE4.1, EM64T
Package Socket 775 LGA (platform ID = 4h)

Temperature sensor 0 38°C (100°F) [0x26] (TMPIN0)
Temperature sensor 1 44°C (111°F) [0x2C] (TMPIN1)
Temperature sensor 2 27°C (80°F) [0x1B] (TMPIN2)
Temperature sensor 0 52°C (125°F) [0x35] (core #0)
Temperature sensor 3 49°C (120°F) [0x38] (core #3)
Temperature sensor 1 49°C (120°F) [0x38] (core #1)
Temperature sensor 2 46°C (114°F) [0x3B] (core #2)
Temperature sensor 0 25°C (76°F) [0x19] (GPU Core)
Temperature sensor 0 24°C (75°F) [0x18] (GPU Core)



errr looking at the other test, i think my system is running a little slow. :T

need do a reboot and repost later. Also tigthen up a bit on my ram timings.

Maybe i'll go for a suicide @ 4.6ghz. :T

 

[myocardia]

Originally posted by: aigomorla

Originally posted by: myocardia
That's a good idea. I just wish a reputable site* would do an article like that.



* A reputable site would 1) know that a QX9650 is not the fastest CPU available today, and 2) wouldn't use a slower video card than you and I have, for their gaming benchmarks.

actuallly....

ruby and i can go 1 on 1 on this statement with you.


The QX9770's have been at best "disapointing". Of course were factoring in average overclocking.

How does that make the QX9650 "the top model of the current lineup"? The QX9770 is the fastest processor available today, and I would think that if I know that, a "review" site should also. Wait, you don't also agree with them using the 12th fastest video available today for their FPS tests, do you?

 

[Foxery]

They're both unlocked processors, and the same core; there's no practical difference besides the freakish price tags. We use higher FSBs than 1600 anyway.

aigo: The two tests are always different. The benchmark includes 2 versions of the program, for some odd reason.

 

[myocardia]

Originally posted by: Foxery
They're both unlocked processors, and the same core; there's no practical difference besides the freakish price tags. We use higher FSBs than 1600 anyway.

aigo: The two tests are always different. The benchmark includes 2 versions of the program, for some odd reason.

Are you saying that the QX9650 is Intel's best performing processor? If so, you should consider going to work for the linked website. You can then give us CPU comparisons, showing FPS in the latest games, using my old GF2 video card, like they've done in this article.:laugh:

 

[aigomorla]

Originally posted by: myocardia

Originally posted by: Foxery
They're both unlocked processors, and the same core; there's no practical difference besides the freakish price tags. We use higher FSBs than 1600 anyway.

aigo: The two tests are always different. The benchmark includes 2 versions of the program, for some odd reason.

Are you saying that the QX9650 is Intel's best performing processor? If so, you should consider going to work for the linked website. You can then give us CPU comparisons, showing FPS in the latest games, using my old GF2 video card, like they've done in this article.:laugh:

what im saying is ruby and i both noticed that QX9770's arent all thats cracked up to be. This is all i am gonna say into this matter. But ruby brought it to my attention. My friends QX9770 is very disapointing at best.

Stock speed, yeah the QX9770 is better. But once you factor ocing in, the average QX9650 will win.

Ummmm actually i had a grip of things loaded in the background. Thats why my first score is slow compared to the second.

The second test i started turning stuff off. :T

 

[myocardia]

Originally posted by: aigomorla
Stock speed, yeah the QX9770 is better. But once you factor ocing in, the average QX9650 will win.

Since you don't seem to have clicked on the link yet, I'll just quote it for you:

This time we?ve compared two CPUs: Core2Quad Q9300 (this is the bottom model of current lineup) and Core2Extreme QX9650 (the top model of current lineup).

Now, this was posted on May 5, 2008. My point all along had nothing to do with which I would buy, because I'd never buy either, and it definitely had nothing to do with overclocking, which they didn't do.

My point was, I'd never trust a website that A) doesn't even know what the fastest CPU a given company makes, and much more importantly, B) uses a midrange video card (that's actually slower than my video card) as the deciding factor between which CPU is faster. Like I keep saying, if they'd used this ATI 7000, or my old GF2, don't you think that every single FPS benmark would have been identical, between the two CPU's?

Or, to phrase it differently, gaming CPU benchmarks are absolutely worthless, if you don't use the fastest video card available at that time. I.E., if the video card can't keep up with the processor, of course you'll get identical FPS, but not because the two CPU's are in fact identical in speed, but only because both are faster than said video card. I never would have mentioned the CPU, which I only brought up because I thought it funny, if they had used a video card that was appropriate.