I thought so too at first, but then again troy's 3570K isn't hyperthreaded either, and Ivy are definately faster than Sandy, especially when clocked higher. Very bizarre.

Well, as mentioned I think I'm bottlenecking on my gtx260 at higher fps rates. It's a pretty old piece of tech by now.

edit: you could try turning off hyperthreading in the bios matrixi, just to check if you get better results then, I've read that some games do not like hyperthreading at all.

why my results with IGP dissapear from max list, while they are still under shown in "Your benchmarks" section

what version where the results done with? You can select the version on the Min and Max lists.

With 0.6B
see pic

it in list of my benchmarks, but not on max list

There is a limit on how many entries per user can be shown. It was 3, now is 4, making your 4th max result showing up now.

aha, ok, thanks.
I just thought to run benchmark on every possible PC I can access XD. dunno maybe someone want to run LFS with IGP so this will show him what he can expect from IGP

Installed XP SP1 last night, Catalyst 12.6, run benchmark on max and I can't still belive what I m seeing:

my result on Windows 7 was 68 FPS!!!!
Hardware: [email protected], HD6750

This is almost twice better FPS. How is it possible to achive twice better FPS just by changing OS. Most people say Win 7 should do better with games. Is it because LFS isnt so Win 7 optimized? or wat
Can someone else also try to run benchmark on XP and see his result. I mean if you get twice better FPS its worth switch back to XP

Wait until you try it on Win98SE

Every time Microsoft release a new OS you need so much more RAM, HDD space and cpu power to just run it!!!!

SD.

I bet it runs beautifully on Windows 1.0

Unfortenaulty I can't find drivers for anything older than XP

Not sure, but could it be something to do with the DirectX API after 9.0c (last XP compatible release) that causes this? Windows Vista came with DX10 integrated, and I'm not sure what Windows 7 ships with. But could maybe be with some slow backwards compatability on the newer DX APIs. Who knows.

Somebody should delete the top 2 entries in the max and min lists by Taavi(EST), they are clearly faked or user error.

Also, the GTX480 isn't in the list of graphics cards.

Quote from southamptonfc :

Somebody should delete the top 2 entries in the max and min lists by Taavi(EST), they are clearly faked or user error.

Also, the GTX480 isn't in the list of graphics cards.

Nothing's faked there, i can assure you. It's down to how LFS has been optimized. Newer hardware will always cause anomalies on older software.

Quote from Taavi(EST) :

Nothing's faked there, i can assure you. It's down to how LFS has been optimized. Newer hardware will always cause anomalies on older software.

Compaing your result to another with a similar system:

7Specht770.6B 2600k @3.5G 8192 Win7 x64GF GTX 560 Ti - 112FPS

Taavi(EST)0.6B 2500K @3.3G 4096 Win7 x64GF GTX 560 - 160FPS - slower cpu and gfx card, 40% faster. errrr no


There are several Intel cpus running at around 4.4ghz and they all have similar results, your results are WAY off what they should be and cannot be explained by an anomoly!

I wonder if it's lack of hyper-threading, I'm going to turn it off and re-test with my 2600

Quote from southamptonfc :

Sorry to tell you that if not faked, you are doing something wrong. Your result is 2X that of Specht77 with almost the same but better spec:

7Specht770.6B 2600k @3.5G 8192 Win7 x64GF GTX 560 Ti - 112FPS

Taavi(EST)0.6B 2500K @3.3G 4096 Win7 x64GF GTX 560 - 160FPS - slower cpu and gfx card, 40% faster. errrr no


There are several Intel cpus running at around 4.4ghz and they all have similar results, your results are WAY off what they should be and cannot be explained by an anomoly!

While the 2600K is indeed a better CPU it is a HT CPU, which means, it has the same amount of physical cores, but the OS sees them as 8 cores. LFS is only capable of using one core. Which in turn means that LFS is using half of the physical core 0. With the 2500K, LFS gets to use the whole core 0. Find someone else with a 2500K and you'll see.

A while ago I've read some text about HT vs non-HT CPUs, they stated that HT shouldn't affect 1 core performance (eg superPI), but when using 2 cores from same CPU performance should be increased by 30-50%. There was superPI and wPrime benchmarks to approve this

Quote from Taavi(EST) :

Which in turn means that LFS is using half of the physical core 0. With the 2500K, LFS gets to use the whole core 0. Find someone else with a 2500K and you'll see.

That's not how it works at all. Anyway, I am able to turn off HT to test this, it gives me a few fps improvement but nothing like 40%.

Your results are totally inconsistent with everyone elses. If they were 40% slower than they should be, I could believe it was a driver quirk or similar.

I'm sorry but it's impossilbe for a slower computer to put in so much faster results, impossible! The only explanation is the testing or recording method was wrong or broken somehow.

He may have tweaked his graphics driver settings for minimum quality/max performance, so it looks like crap but it's fast. I generally have mine set for maximum quality. I don't know how that might affect CPU performance, though.

I complained about this score inflation years ago on the first page of this thread.

Quote from southamptonfc :

I'm sorry but it's impossilbe for a slower computer to put in so much faster results, impossible! The only explanation is the testing or recording method was wrong or broken somehow.

My age old Pentium 4(3 ghz) machine with a Club 3D Radeon X550 and Win XP could run LFS at 40-60 fps constantly after replacing the thermal paste on both the GPU and CPU. Comparing that to DANIEL-CRO's specs it was pretty unbelievable to me considering it's a 7-8 year old rig. I would assume it's because Pentium 4 is single core.

Quote from hp999 :

My age old Pentium 4(3 ghz) machine with a Club 3D Radeon X550 and Win XP could run LFS at 40-60 fps constantly after replacing the thermal paste on both the GPU and CPU. Comparing that to DANIEL-CRO's specs it was pretty unbelievable to me considering it's a 7-8 year old rig. I would assume it's because Pentium 4 is single core.

Which version of LFS and which benchmark? The more recent benchmarks are a lot more demanding, you cannot compare results like that. You would need to run this benchmark on your P4.

I upgraded from a P4 (3.2 EE) to a C2D 8500 (dual core) and my minimum frame rate was higher on the C2D than the average framerate of the P4

Actually I wouldn't be surprised if the last Pentium 4s delivered better performance than SU7300. They have more than twice as high clock speed, same FSB speed and LFS doesn't make any use of multiple cores or SSE4 instruction set.

Quote from hp999 :

My age old Pentium 4(3 ghz) machine with a Club 3D Radeon X550 and Win XP could run LFS at 40-60 fps constantly after replacing the thermal paste on both the GPU and CPU. Comparing that to DANIEL-CRO's specs it was pretty unbelievable to me considering it's a 7-8 year old rig. I would assume it's because Pentium 4 is single core.

Comparing to which benchmark?
I think that even stock E5200 should be atleast twice faster than pentium 4 in single core performance. My current superPI result is about 15s and Pentium 4@3GHz is about 45s.
Tell us exact results scored in benchmark. also keep in mind that my average FPS online is about 150-200.

Quote from MadCatX :

Actually I wouldn't be surprised if the last Pentium 4s delivered better performance than SU7300. They have more than twice as high clock speed, same FSB speed and LFS doesn't make any use of multiple cores or SSE4 instruction set.

Clock speed is irrelevant for comparing CPU speeds between different generation/types of CPU's.

As an example is aforementioned P4 @3GHz compared to core2duo CPU at 1.8GHz (forgot which model it was).
Well.. P4 was SLOWER some 20-30%.

Quote from DANIEL-CRO :

Comparing to which benchmark?
I think that even stock E5200 should be atleast twice faster than pentium 4 in single core performance. My current superPI result is about 15s and Pentium 4@3GHz is about 45s.
Tell us exact results scored in benchmark. also keep in mind that my average FPS online is about 150-200.

Notice how I used the past tense in my previous post - it was taken apart last year, meaning no accurate benchmark numbers, sadly. Doubt I could even find the parts to put it back together either, more than half of the PC is some 100km away from me. If anyone here has a spare P4 and an X550 lying around then give it a go.