If time permits, a simple compromise would be to bench both kinds of games, the current blockbusters that stress the GPU to their limit and make the manufacturer happy, and a few of the more visually demanding Sims that make us SimHQ readers happy. Or if the Sim results really show no remarkable FPS differences for different cards, resolutions and quality settings, then you could just post them on the forums and not in the main article. Or make a dedicated "Sim round-up" from time to time, after you have gathered a few cards (each of which got its own individual review shortly after release), similar to the "Pixels for Pennies" round-up but using all kinds of simulations as benchmarks. I think this would be a great decision help because people will not only get an idea of what to expect FPS-wise, but they can also compare how well the competing cards handle different titles.

I agree 100% with Acedy. Can SimHQ benchmark both types of titles?

In the end, if benchmarks of sims show no differences between an older and newer GPU, I feel it is SimHQ's duty to tell the readership that they don't need that new GPU for their sims. That's the bottom line. But, I understand that this can't be done at the expense of alienating SimHQ from the review sample pipeline, or else we never have the opportunity to see the results.

Here are some benchmarks that show how using a fast CPU on FSX does indicate a real world graphics card performance eval.

http://www.tomshardware.com/reviews/cpu-gpu-upgrade,1928-10.html

Also, even though IL-2 and older sims still don't benefit as much from newer cards, they do benefit and that perf increase is relevant...

I'd even be willing to pitch in a buck or two for benchmarks with new cards against 5 or 6 sims... on a pay per view basis so SimHQ could support testing the things we all want tested and still be able to maintain a "fastest CPU" system / replace CPU every 6 months or so with mid/high end and O/C, etc...

I'm not sure whether Caveman is agreeing with John/guod's comments about the CPU/GPU's comments or not, but the link above just shows how irrelevant the GPU is on games such as FSX.

I presume the hardware editors have got a balancing act to do here, they'll never please everyone and it's an obvious problem when many 'sim' type games are CPU heavy as opposed to giving the GPU a good workout.

Hehe these are intersting results. IMO you´re both right, in a way: the results show that going from a Geforce 6/7 card to a Geforce 8/9 has a huge effect on FPS (using the same CPU that is), so with the older generations FSX is GPU bound. On the other hand within the Geforce 8/9 generation FSX seems to be CPU bound, as going from a weaker to a stronger card (8800GT/9600GT to 8800GTS/9800GTX) doesn´t really make a difference, what matters though is CPU power.

Thanks for the link btw, good news for me, I´ve still got a 7950GT but plan to upgrade it in the next months.

YES!!! For someone like myself with a 6800 Ultra going to something in the new round of cards from Nvidia and AMD, the benchmark IS relevant. I suspect there are many folks like myself who are ready to jump in with new rigs now that there is so much competition between Intel/AMD and Nvidia/AMD driving the prices down...

Again, my thanks to the folks who took the time to make the review...

Yep, which is exactly why FSX and Lock On aren't used in GPU tests here.



Exactly.

Just picked up a new 4850 a few days ago. I made the jump up from a x850XT I have had for 2 years so I can finally play with some AA and AF turned on.

I haven't really had a chance to play most of my sims and games in depth (BF2 project reality mod, ARMA, GTR2, IL2, LOCKON, but I fired them up real quick and noticed super smooth gameplay with everything on high settings. I am pleased. Can't complain about the relatively low price point on a brand new card either.

We have added some further comments to John's latest article on the 4870 (see the homepage). Read them and comment back in here your thoughts.

We have some further ideas how to address the issue in upcoming reviews.

With so many titles in development, more than we've seen for a decade, it's quite frustrating.

While I understand many of the points that left simulations out from the benchmark testings, there are two sims that could definitly be used and where I'm willing to bet that diference would certainly be noticed (between older and newer video cards and between newer cards), those 2 sims would be:
-> Silent Hunter 4, which with the latest patches (being the latest one, the version 1.4) completly removes Starforce protection.
-> Armed Assault. Some friends on mine that play Armed Assault and that upgraded video cards since, have noticed a very big improvement in terms of performance. For example a friend of mine, which had ATI 2600Pro and upgraded to a Nvidia 9600gt and the incress in perfomance was HUGE!

So, I think that at least those 2 sims should be included in such reviews/benchmarks.