VirtualR – Sim Racing News
100% Sim Racing News

phew, thats one anal webiste :) page after page after page of benchmarks with writeups at each resolution. I mean who reads all of that ??
Make a table, put in the numbers and you’re done :) Its all moot anyway because performace will depend on your own hardware and preferred game settings so its still only a guide.

Anyway if I get an improvement with the patch thats great, if not, meh, it was ok anyway.

“The Radeon HD 4000 series has mostly been dropped from this evaluation because any gamers looking to upgrade to an ATI video card should be considering the new Radeon HD 5000 series products for their excellent performance and feature sets.”

Does that mean no performance improvements for my HD 4850?

Of course HD4xxx series will also have performance improvements.. thats just HD5xxx propaganda ^^ no FPS drop anymore ;)

But it seems to have full advantage we will ned both patch and the ati drivers version 9.11 or 9.12..

F1Racer: phew, thats one anal webiste :) page after page after page of benchmarks with writeups at each resolution.I mean who reads all of that ??
Make a table, put in the numbers and you’re done :) Its all moot anyway because performace will depend on your own hardware and preferred game settings so its still only a guide.Anyway if I get an improvement with the patch thats great, if not, meh, it was ok anyway.

Well you have to give it to them for being thorough.

free2game365: Well you have to give it to them for being thorough.

Yeah you’re not kidding :)

for some bizarre reason virtualr is refusing to let me post a comment on the previous FSOne post. It keeps asking me to login even though I just have as you can see from this comment!

The article proves that Shift is even harder on the CPU/memory than it is on the GPU.

At low resolutions, the FPS is very poor considering they use using such a fast CPU. It does reflect my own experience though. We’ll probably have to wait for the next generation of CPU (overclocked) to see Shift running smoothly at 60fps on all tracks with 8+ cars on track.

It runs ok on my MacBook Pro, just at low resolutions. The 9600M GT is a sloth-like video card.

digitalxni: for some bizarre reason virtualr is refusing to let me post a comment on the previous FSOne post. It keeps asking me to login even though I just have as you can see from this comment!

Do the login and then press F5: you should be able to write, then.

Damien Smith: The article proves that Shift is even harder on the CPU/memory than it is on the GPU.At low resolutions, the FPS is very poor considering they use using such a fast CPU. It does reflect my own experience though. We’ll probably have to wait for the next generation of CPU (overclocked) to see Shift running smoothly at 60fps on all tracks with 8+ cars on track.

I have a Q9550 and it still has poor perfomance in some tracks..

This next patch will not only improve GPU performance it will also improve CPU performance :)

@Damien Smith

The CPU utilization looks pretty good in this test, IMO. The average CPU usage is 21% – 27%, which can be considered low, for such a complex application.

Also, bear in mind that in this article the GPUs are tested with the highest possible AA/AF levels – 8x, 12x, 24x AA / 16x AF. That’s why the FPS is lower than 60fps.

I’m playing Shift with 8800GTS (640M) on 1680×1050 with 2xAA/8xAF, everything on “max” except motion blur turned off and shadows on medium. The average FPS is between 40 and 60+ with the maximum number of cars.

khizerk: Does that mean no performance improvements for my HD 4850?

No it just means the article is a bit of a waste of time from our point of view.

The HD5xxx series cards they tested could already play the game fine anyway – at humongous resolutions and, if you want, with 3 screens. As the article says, it makes a fast card even faster.

The irony is, it’s difficult to find stock of HD58xx series cards anyway even if you wanted one. So it’s even more pointless ATI putting out PR telling us how fast they are in this game.

Other articles have shown an improvement in the 4xxx series with the new patch (but they didn’t use a 4850 either)

Bottom line : You are going to see an improvement, but it remains to be seen whether that improvement will be enough.

But, if every other showstopping bug is fixed, I would upgrade anyway if needed. Right now the patch is needed for far more than just the ATI issue though.

It’s a pity they didn’t give the patch / driver early to someone who would have actually tested the cards that had a performance problem with the game. But this is all about PR and Hardcp was where some of ATI’s negative comments about Shift development were published hence they get the patch and, well, as you can see they don’t really live in the mainstream world :)

Also, bear in mind that in this article the GPUs are tested with the highest possible AA/AF levels – 8x, 12x, 24x AA / 16x AF. That’s why the FPS is lower than 60fps.I’m playing Shift with 8800GTS (640M) on 1680×1050 with 2xAA/8xAF, everything on “max” except motion blur turned off and shadows on medium. The average FPS is between 40 and 60+ with the maximum number of cars.

Yeah, but the FPS is virtually the same regardless of GPU settings. There’s a serious bottleneck somewhere in Shift that prevents it from getting above 55-60fps consistently – even on ridiculously fast overclocked systems.

20-30% CPU utilization is a BAD thing. It either means that v-sync is enabled or there’s no multi-threading at all or these’s a massive memory/FSB bottleneck.

Damien Smith:
Yeah, but the FPS is virtually the same regardless of GPU settings.

Yes, but the resolutions are different. Although, the GPUs are also different (from high end to mainstream), so I am agree with you to some degree.

Damien Smith:
20-30% CPU utilization is a BAD thing. It either means that v-sync is enabled or there’s no multi-threading at all…

As I said before, this a complex multithreading software application, and 21% – 27% CPU utilization is normal, IMO.

Those benchmarks show that the game isn’t actually very demanding by today’s standards. There are older games more demanding than that.

The AA comparisons were interesting. Especially that 24x doesn’t make a noticeable visual difference from 12x on ATI (yet it has a huge performance impact). The game does look pretty damn good with 12x AA, and my 4870 does an ok job handling it except when many cars are on screen. The patch and new drivers should help with that :)

Those benchmarks show that the game isn’t actually very demanding by today’s standards.

If that were true, Shift would get 100+fps on very high end hardware but it most certainly doesn’t.

Damien Smith:
If that were true, Shift would get 100+fps on very high end hardware but it most certainly doesn’t.

In your opinion. Not mine.

This test fails in many ways to show what the patch will improve in performance, because

- they don’t use current mainstream ATI hardware (38xx and 48xx cards)
- they go for maximum FSAA, which isn’t necessary at all with the high resolutions that were tested but which surely kills performance
- they tested only one track, and one of the lesser demanding tracks at that
- they tested with a (beta) driver that ATI users don’t have access to at the moment, and besides the patch was supposed to improve ATI performance regardless of the driver so why not test that with currently available drivers?

On top of that, EA/SMS claimed that the patch would also improve performance for nVidia users and the test saw none. Hence, either something’s wrong with the patch or with the test set-up. Either way, it means that what the patch shows in this review is probably not what we will see when it is finally released.

redi: This test fails in many ways to show what the patch will improve in performance, because- they don’t use current mainstream ATI hardware (38xx and 48xx cards)- they go for maximum FSAA, which isn’t necessary at all with the high resolutions that were tested but which surely kills performance- they tested only one track, and one of the lesser demanding tracks at that- they tested with a (beta) driver that ATI users don’t have access to at the moment, and besides the patch was supposed to improve ATI performance regardless of the driver so why not test that with currently available drivers?On top of that, EA/SMS claimed that the patch would also improve performance for nVidia users and the test saw none. Hence, either something’s wrong with the patch or with the test set-up. Either way, it means that what the patch shows in this review is probably not what we will see when it is finally released.

Agree but from what I’ve read in order to achieve the best performance we will need this patch+the drivers 9.12 (supposedly those beta drivers they refer to) released in december.

What really matters is that the performance will be better + a whole alot of bug fixes

f0xx: Agree but from what I’ve read in order to achieve the best performance we will need this patch+the drivers 9.12 (supposedly those beta drivers they refer to) released in december. What really matters is that the performance will be better + a whole alot of bug fixes

Well my problem is that they tested the difference between patch 1 and patch 2 and found a performance decrease, which goes against everything that was promised for patch 2. Hence, I am skeptical to say the least about this whole test/review.

“Fixed controller lag caused by ATI drivers and slower systems,” is what stood out to me. That probably answers most of the people that have controller lag issues and find the game unplayable. The new ATI drivers has solved that problem. Even if the FPS doesn’t go up significantly, smoother running should stop the controller lag. Nvidia will soon also come out with a new driver, which hopefully will help Nvidia in Shift on slower systems. I have a 260GTX. Runs very good with no controller lag. AT 2560×1024, with 16X quality filtering, 8xAA, every thing high in game, except shadows. Which are on medium. 25 low, 60 high 45-50 avg.
Out standing image, smooth as silk, best physics ever experienced.
The higher use of multi-core CPU’s is due to the PhysX use. Which is calculated by the CPU and not a GPU. GPU’s are 100 times faster at those calculations. But then ATI users would be left out.

This patch may make me think of the PC version of Shift as a retail product and not a beta

Howie47: “Fixed controller lag caused by ATI drivers and slower systems,” is what stood out to me. That probably answers most of the people that have controller lag issues and find the game unplayable.The new ATI drivers hassolved that problem. Even if the FPS doesn’t go up significantly, smoother running should stop the controller lag. Nvidia will soon also come out with a new driver, which hopefully will help Nvidia in Shift on slower systems. I have a 260GTX. Runs very good with no controller lag. AT 2560×1024, with 16X quality filtering, 8xAA, every thing high in game, except shadows. Which are on medium. 25 low, 60 high 45-50 avg.
Out standing image, smooth as silk, best physics ever experienced.The higher use of multi-core CPU’s is due to the PhysX use. Which is calculated by the CPU and not a GPU. GPU’s are 100 times faster at those calculations. But then ATI users would be left out.

please, please, please, show us a no controller lag video!