Intel Q9450 vs Phenom 9850 - ATI HD3870 X2

[gosh]

Intel Q9450 vs Phenom 9850

EDIT:
GPU = ATI HD3870 X2

Game: Call of Duty 4
Settings: 1920x1200, 4xAA 16xAF 4xAA 16xAF, everything MAX
Q9450 = 65
Phenom 9850 = 69

Game: Crysis
Settings: 1280x1024, 1xAA/1xAF, DX9 everything HIGH
Q9450 = Min FPS 22.46, Max FPS 61.71, Avg 41.96
Phenom 9850 = Min FPS 29.82, Max FPS 63.20, Avg 47.89

Game: STALKER
Settings: 1920x1200, 16xAF everything MAX
Q9450 = 83
Phenom 9850 = 84

Game: TDU
Settings: 1920x1200, 4xAA/16xAF, everything HIGH
Q9450 = 45
Phenom 9850 = 47

Game: Half Life 2
Settings: 1920x1200, 4xAA/16xAF, MAX everything.
Q9450 = 295
Phenom 9850 = 300

[perkam]

M8, this is for XS Benchmarking.

Unless you post some of the results in the post, it will be removed due to Linking to External Websites.

Perkam

[GAR]

How is the phenom beating the Q9450, that doesnt look right.

[stangracin2]

i think its something about the spider platform

[gosh]

How is the phenom beating the Q9450, that doesnt look right.

I/O performance

[RunawayPrisoner]

How is the phenom beating the Q9450, that doesnt look right.

Just reduce LOD when it's Phenom's turn, and voila. (in Crysis' case)

Or... selectively choose where in the game Q9450 will be slightly lower than Phenom.

[gosh]

Just reduce LOD when it's Phenom's turn, and voila. (in Crysis' case)

http://www.overclockersclub.com/revi...el_q9450/8.htm

[RunawayPrisoner]

http://www.overclockersclub.com/revi...el_q9450/8.htm

Benchmark_gpu, sir... benchmark_gpu...

Now please point out to me how Crysis is CPU-bound in a GPU test? And at most, I'd believe that Crossfire support is better on the AMD boards, which is nothing new. Crossfire performance varies with Intel boards as well. But the 6fps difference would lead me to believe that either:

1) Crossfire is better with the AMD board used.

2) Something was wrong... like LOD was turned all the way down (or up), in which case, it can totally happen.

P.S.: So... uhh... maybe some CPU-bound benchmarks would do better? And a GPU without Crossfire or SLI should be used as well... or actually, just benchmark solely the CPUs. This seems more like a Q9450 system versus Phenom 9850 system with a 3870X2 than a Q9450 cpu versus Phenom 9850 cpu.

P.S. 2: By the way, umm... in that test, it's very weird on the Q9450's side.

333 x 8 is default for Q9450. Even if it was 334 x 8, it wouldn't even be 2700MHz, it would be more like 2672MHz. And then... what kind of FSB/RAM ratio would give 1080MHz for RAM when the system bus is 334MHz????

[MAS]

remember, HT 3.0 is a secret Spider weapon

The huge bandwidth is needed when writing or reading textures (that canot fit into local videomemory) to RAM.
Current Intel platform cannot provide such bandwidth, HT 3.0 can.

[oohms]

I remember back in the 939 days running the HT bus at 200 mhz instead of 1000 mhz and how there was no performance difference... Not saying this is fake, but im highly skeptical until other reviewers can show similar results

(I hope it isnt fake )

[informal]

The user who tested both systems is highly respected forum member,the one who brought us the 1st or 2nd user review of Radeon 4850 on the net...He sure didn't fake that one since we know how well Radeon 4850 performed.And why would he fake this now??

[gosh]

I remember back in the 939 days running the HT bus at 200 mhz instead of 1000 mhz and how there was no performance difference... Not saying this is fake, but im highly skeptical until other reviewers can show similar results

I think that the bus hasn’t been any problem until this year. Video Card performance has increased tremendously this year. Games that are using threading more have arrived. Playstation 3 and XBOX 360 works well with threading. That increases the market for threaded game engines that need much more performance when they communicate with external hardware.

[MAS]

the results are right, I know some sources with similar results
if Phenom could reach 4GHz - it would be very good gamer cpu, but unfortunately it couldn't

[Genarator]

I assume they are both at stock speeds, any chance of a benchmark of how they compare when they're overclocked?

Good to see AMD making a comeback!

[gosh]

Here is another interesting list showing scaling in Race Driver Grid

LIST - Processors Bottom to Top

THREAD

[bowman]

remember, HT 3.0 is a secret Spider weapon

The huge bandwidth is needed when writing or reading textures (that canot fit into local videomemory) to RAM.
Current Intel platform cannot provide such bandwidth, HT 3.0 can.

Bandwidth or not, system memory is way too slow for real-time rendering. Maybe it would matter if system memory was GDDR3. DDR2 is too slow to begin with, then you have to account for the hop from system memory to CPU northbridge, via HyperTransport to the motherboard northbridge, then via PCI-E to the GPU.. You can't spin that to be relevant, because it simply isn't in any way.

Games are rarely CPU bound, so I don't see how this is new to anyone. If you play at high resolutions CPU hardly matters.

[gosh]

Here is one explanation why AMD performs better when you run more advanced games and also run them on high res.
It is the same reason as why AMD is strong on servers even if Intel Xeon is higher clocked.
It’s about communication with other hardware (memory and GPU).

When Intel communicates with other hardware it sends data using the FSB. All traffic goes through the FSB. The Latency using the FSB is about 250 clocks. The performance penalty is rather large. If one game is reading memory and/or sends I/O to the GPU using only one thread. Then I don’t think you will see any big differences between AMD and Intel. If the video card is slow you defiantly isn't going see any differences.
Now if the game is using more than one thread the situation changes. If one thread is sending or reading data on Intel, and another thread is sending data to the video card. One of the threads needs to wait. That means that latency goes up. In worst case scenarios it would be double (about 500 clocks).
On AMD this is handled differently. AMD has hypertransport that handles I/O to the video card. If one thread is sending data to the video card it doesn’t compete with memory. On AMD latency is a bit lower too.
Running games on low res or games that isn’t that advanced then this isn’t a problem on Intel. But if you are running a game on high res, have a fast video card (or two). Then this video card is able to handle VERY MUCH I/O. Also if the game is using more memory then more data needs to travel through the FSB on Intel. This situation will lead to more conflicts in the FSB and latency is increased.
Intel is very fast when it isn’t disturbed and is able to use data in the cache. More threads, more memory, more communication with video card and maybe more synchronization between threads. Then performance will shift to favor AMD system design.

[RunawayPrisoner]

*sighs* Do I have to say it again? The two systems used for the test are vastly different, and the numbers on the Intel platform is questionable. You can't say it's a Phenom versus Q9450 CPU test, since many factors are in there. It would be a Phenom system versus Intel system test, though. And even as a platform versus platform test, I still have some questions about the numbers on the Intel platform (like what kind of bus frequency did it take to get a Q9450 to be at 2700 and the RAM at 1080? If the bus frequency budged (for instance... from 333 to 337?), then it wasn't a fair test.

And by the way, Informal, didn't want to say this, but I was the first one ever to 3D bench a X3350/Q9450 at 4GHz. Does that mean I'm credible? No, sir.

[Loser777]

The Phenom 9850 has been beaten by 45nm in many other tests... It probably doesn't show here because the systems are too different.

[JumpingJack]

The Phenom 9850 has been beaten by 45nm in many other tests... It probably doesn't show here because the systems are too different.

Dude....

Gosh = kassler

I thought I had seen the junk before, he picked out a comparison of two quad cores that is GPU limited, then he argues that it is due to Intel's FSB bottlenecking, and argues ad infinitum ... no amount of data, information, or rational explanation will sway him from his message.

He has been spamming other forms with the same post, he hit HardOCP
http://hardforum.com/showthread.php?...hlight=kassler

Here is another one, he is probably bioduken at OCclub:
Read this one, http://forums.overclockersclub.com/i...&#entry1557748
Then read his post just above, number 17.

I found a similar thread with a google over at Tom's, but they apparently deleted it while I was gathering info.

It is best to ignore him.

[gosh]

Dude....

Gosh = kassler

Yes!

[gosh]

The Phenom 9850 has been beaten by 45nm in many other tests...

Do you have any test comparing AMD and Intel on high res and/or advanced settings using fast video cards?

[RunawayPrisoner]

Yes!

Then I guess members on [H] also came to the same conclusion. The fastest (or best) AMD system seems to be a little bit more efficient with Crossfire, no more, no less. The CPUs themselves have nothing to do with the benchmark.

[gosh]

Then I guess members on [H] also came to the same conclusion. The fastest (or best) AMD system seems to be a little bit more efficient with Crossfire, no more, no less. The CPUs themselves have nothing to do with the benchmark.

That could be one reason but it isn't logical.

[RunawayPrisoner]

Then it is logical for Agena to suddenly beat Penryn... out of the blue? And by the way, I think I pointed this out, but Penryn's clocks in this test are not stock, which could mean the tester modified something on the Penryn system.