- Status
[citation][nom]Why_Me[/nom]The funny part of all this is if the i7 comes out ahead of the AMD again, we will see another thirteen pages of post in regards to what needs to be changed so that the AMD can come out on top.[/citation]
With the same GPU, it should come out ahead. It is a better processor.
It should not be coming out 50% better though, with a GPU bottleneck.
With 4890s vs 4870s, it will be interesting.
With the same GPU, it should come out ahead. It is a better processor.
It should not be coming out 50% better though, with a GPU bottleneck.
With 4890s vs 4870s, it will be interesting.
- Mar 7, 2003
- 12,012
- 0
- 40,790
[citation][nom]jennyh[/nom]Sorry let me be clear.WiC and Farcry 2 favour Nvidia, however with the swap to 4870's (or your underclocked 4890's), the lead for the i7 in WiC should be lessened, and I also believe that the 4890's would be enough to push the Phenom II over the i7 in Farcry 2.[/citation]
So then you believe that 1/2 of the games in our 4-game mini suite *aren't* CPU limited, therefore the games we chose don't "massively favor CPUs over GPUs".
[citation][nom]jennyh[/nom]Also, if I may ask another question - why no AA in Crysis?[/citation]
Good question - we used to use AA in our Crysis benches.
We decided against it because it didn't deliver playable framerates on even high-end rigs. I'm all for stress testing a system, but if it's unplayable at the settings we test at, then it's a useless benchmark IMHO. So we substituted instead for the very high settings, which stress the hell out of GPUs as it is.
So then you believe that 1/2 of the games in our 4-game mini suite *aren't* CPU limited, therefore the games we chose don't "massively favor CPUs over GPUs".
[citation][nom]jennyh[/nom]Also, if I may ask another question - why no AA in Crysis?[/citation]
Good question - we used to use AA in our Crysis benches.
We decided against it because it didn't deliver playable framerates on even high-end rigs. I'm all for stress testing a system, but if it's unplayable at the settings we test at, then it's a useless benchmark IMHO. So we substituted instead for the very high settings, which stress the hell out of GPUs as it is.
balancedthinking
Distinguished
- Apr 28, 2009
- 16
- 0
- 18,510
The 2X4890 Phenom system will NOT win vs. the i7 2X4870 system in FC2.
The far cry 2 benchmark tool, particularly the "small ranch" bench is heavily limited by the cpu and i7 will still win big time even though it only has 2 4870s.
Ever wondered why there is also a special "Intel" bench available? Because Intel paid the developers for the whole far cry 2 benchmark tool!!! This includes optimising the engine AND the benchmark runs to make intel CPU look better!
Make a savegamebench within the normal game and you will see a lot more demand on the gpus and the results will get pretty close. Quiet a feat for a still Intel optimised game.
Use Left4Dead, Fallout3, Race Driver Grid and you will see that a balanced plattform by AMD will beat the i7 Intel plattform with inferior graphics subsystem.
The far cry 2 benchmark tool, particularly the "small ranch" bench is heavily limited by the cpu and i7 will still win big time even though it only has 2 4870s.
Ever wondered why there is also a special "Intel" bench available? Because Intel paid the developers for the whole far cry 2 benchmark tool!!! This includes optimising the engine AND the benchmark runs to make intel CPU look better!
Make a savegamebench within the normal game and you will see a lot more demand on the gpus and the results will get pretty close. Quiet a feat for a still Intel optimised game.
Use Left4Dead, Fallout3, Race Driver Grid and you will see that a balanced plattform by AMD will beat the i7 Intel plattform with inferior graphics subsystem.
jennyh
Splendid
- Mar 3, 2009
- 4,292
- 0
- 22,860
Cleeve, you have to understand that the AMD fanboys and girls are annoyed by this for a reason. While it's easy to rant and rave, pinpointing the reasons aren't quite so easy but I'll give it my best shot.
This review has made the 4890 look worse than it is. Even though you have laid the blame squarely on the 955 BE, you have still shown crossfire 4890's to perform much worse than they do in the majority of games.
This is the 2nd point - many of us believe that the choice of games in the benchmark is very flawed. WiC is a huge lead for Nvidia - it is probably the most Nvidia biased title you can buy right now - a single 260 outperforms a single 4890 in WiC anyway so really it's not a great surprise to see dual 260's in an i7 rig outperform the Dragon. However, this pisses off AMD fans - WiC is an abberation in gaming benchmarks just the same as Grid is in favour of AMD. Most benchmarkers who use WiC would normally have the decency to run Grid as a balance however.
You said yourself that AA is unplayable at high settings on Crysis. Why is this benchmark being used to test high end systems? We're talking about a 2 year old game that still refuses to work properly with high end gpu's. By not using AA, you have opened yourself up to accusations of bias because there was only ever going to be one result. With AA? That i7 system would have died horribly and I bet the Dragon wasn't nearly so bad. In other words, it would have been a 'win' for the Dragon I'm quite sure, and it would have put a whole different complexion on the gaming benchmarks.
I actually think it's a decent review, just the choice of games have made the Dragon out to be much worse than it really is. I know you're doing another one but what would have been much more interesting would be seeing games like Grid and Age of Conan benched on each system. I assure you it would have looked completely different - AMD fans know this, Nvidia and intel fans know this too, but as usual for THG recently, the AMD fans feel like they are getting the short straw once again.
This review has made the 4890 look worse than it is. Even though you have laid the blame squarely on the 955 BE, you have still shown crossfire 4890's to perform much worse than they do in the majority of games.
This is the 2nd point - many of us believe that the choice of games in the benchmark is very flawed. WiC is a huge lead for Nvidia - it is probably the most Nvidia biased title you can buy right now - a single 260 outperforms a single 4890 in WiC anyway so really it's not a great surprise to see dual 260's in an i7 rig outperform the Dragon. However, this pisses off AMD fans - WiC is an abberation in gaming benchmarks just the same as Grid is in favour of AMD. Most benchmarkers who use WiC would normally have the decency to run Grid as a balance however.
You said yourself that AA is unplayable at high settings on Crysis. Why is this benchmark being used to test high end systems? We're talking about a 2 year old game that still refuses to work properly with high end gpu's. By not using AA, you have opened yourself up to accusations of bias because there was only ever going to be one result. With AA? That i7 system would have died horribly and I bet the Dragon wasn't nearly so bad. In other words, it would have been a 'win' for the Dragon I'm quite sure, and it would have put a whole different complexion on the gaming benchmarks.
I actually think it's a decent review, just the choice of games have made the Dragon out to be much worse than it really is. I know you're doing another one but what would have been much more interesting would be seeing games like Grid and Age of Conan benched on each system. I assure you it would have looked completely different - AMD fans know this, Nvidia and intel fans know this too, but as usual for THG recently, the AMD fans feel like they are getting the short straw once again.
- Mar 7, 2003
- 12,012
- 0
- 40,790
[citation][nom]balancedthinking[/nom]Use Left4Dead, Fallout3, Race Driver Grid ...[/citation]
You should be pretty happy then, the games I've added to the suite for the rematch are Left 4 Dead, Fallout 3, H.A.W.X., and Prototype.
Got everything you wanted in there, except GRID.
You should be pretty happy then, the games I've added to the suite for the rematch are Left 4 Dead, Fallout 3, H.A.W.X., and Prototype.
Got everything you wanted in there, except GRID.
jennyh
Splendid
- Mar 3, 2009
- 4,292
- 0
- 22,860
[citation][nom]Cleeve[/nom]You should be pretty happy then, the games I've added to the suite for the rematch are Left 4 Dead, Fallout 3, H.A.W.X., and Prototype.Got everything you wanted in there, except GRID.[/citation]
Thanks for adding those but it would have been a lot fairer had those been used in the i7/260 vs dragon system review.
Thanks for adding those but it would have been a lot fairer had those been used in the i7/260 vs dragon system review.
- Mar 7, 2003
- 12,012
- 0
- 40,790
[citation][nom]jennyh[/nom]WiC is a huge lead for Nvidia - it is probably the most Nvidia biased title you can buy right now - a single 260 outperforms a single 4890 in WiC anyway so really it's not a great surprise to see dual 260's in an i7 rig outperform the Dragon.[/citation]
And that's why the follow up uses the 4870 vs. 4890.
To be honest, I didn't realize that WiC had such a massive Nvidia-slant until I re-did the benches. I had thought the 4890 had enough oomph to overcome any driver preference the game showed. having said that, with the re-test using AMD graphics in both systems, this is no longer an issue.
[citation][nom]jennyh[/nom]You said yourself that AA is unplayable at high settings on Crysis. Why is this benchmark being used to test high end systems? We're talking about a 2 year old game that still refuses to work properly with high end gpu's. By not using AA, you have opened yourself up to accusations of bias because there was only ever going to be one result. With AA? That i7 system would have died horribly and I bet the Dragon wasn't nearly so bad. In other words, it would have been a 'win' for the Dragon I'm quite sure, and it would have put a whole different complexion on the gaming benchmarks.[/citation]
Here's where you and I think very differently. 9 fps vs. 15 fps isn't a win for anybody. It's a loss for everybody though, including the reader who has wasted time on useless information that has no impact on reality.
Who gives a rat's bunghole if either AMD or Intel 'wins' at a benchmark nobody can use? I guess if the point is to 'keep score' for fans of either team, like some bizarre hockey game, it might be of some use.
I won't be a party to that kind of a waste of time though. If I run a benchmark, I at least try to make sure it has an impact on someone's reality, not a scoreboard. If someone wants to spin that as a bias, well, I'll admit to having a bias for real-world results. If they want to spin that as a platform preference, I don't think it holds up, but people tend to see what they want.
[citation][nom]jennyh[/nom]I actually think it's a decent review, just the choice of games have made the Dragon out to be much worse than it really is...
...AMD fans know this, Nvidia and intel fans know this too, but as usual for THG recently, the AMD fans feel like they are getting the short straw once again.[/citation]
Well, I'm hoping the elimination of the GeForce vs. Radeon variable will clear a lot of things up in the follow up and make everyone more confortable with the results.
And that's why the follow up uses the 4870 vs. 4890.
To be honest, I didn't realize that WiC had such a massive Nvidia-slant until I re-did the benches. I had thought the 4890 had enough oomph to overcome any driver preference the game showed. having said that, with the re-test using AMD graphics in both systems, this is no longer an issue.
[citation][nom]jennyh[/nom]You said yourself that AA is unplayable at high settings on Crysis. Why is this benchmark being used to test high end systems? We're talking about a 2 year old game that still refuses to work properly with high end gpu's. By not using AA, you have opened yourself up to accusations of bias because there was only ever going to be one result. With AA? That i7 system would have died horribly and I bet the Dragon wasn't nearly so bad. In other words, it would have been a 'win' for the Dragon I'm quite sure, and it would have put a whole different complexion on the gaming benchmarks.[/citation]
Here's where you and I think very differently. 9 fps vs. 15 fps isn't a win for anybody. It's a loss for everybody though, including the reader who has wasted time on useless information that has no impact on reality.
Who gives a rat's bunghole if either AMD or Intel 'wins' at a benchmark nobody can use? I guess if the point is to 'keep score' for fans of either team, like some bizarre hockey game, it might be of some use.
I won't be a party to that kind of a waste of time though. If I run a benchmark, I at least try to make sure it has an impact on someone's reality, not a scoreboard. If someone wants to spin that as a bias, well, I'll admit to having a bias for real-world results. If they want to spin that as a platform preference, I don't think it holds up, but people tend to see what they want.
[citation][nom]jennyh[/nom]I actually think it's a decent review, just the choice of games have made the Dragon out to be much worse than it really is...
...AMD fans know this, Nvidia and intel fans know this too, but as usual for THG recently, the AMD fans feel like they are getting the short straw once again.[/citation]
Well, I'm hoping the elimination of the GeForce vs. Radeon variable will clear a lot of things up in the follow up and make everyone more confortable with the results.
jennyh
Splendid
- Mar 3, 2009
- 4,292
- 0
- 22,860
It will help yes. I fully expect the i7 to be a decent amount ahead. It's a better chip than any Phenom II will ever be and I don't think anyone would really believe otherwise.
Your point on WiC though - I realise it's hard to believe it but every single WiC bench around is showing up how much it massively favours Nvidia hardware. I've even seen it perform better on a gts250 compared to a 4870 in one benchmark.
Hopefully this will be an eye-opener of sorts for future reviews. WiC will always favour Nvidia hardware, Crysis will always favour i7 and Grid will always favour ATI. I'm quite interested in how well the i7 with 4870's competes with the Dragon however - If that one ends up a decent win for the i7 then that would be real proof of how much faster the i7 is than the 955.
My reckoning is, it will be a tie. 😀
Your point on WiC though - I realise it's hard to believe it but every single WiC bench around is showing up how much it massively favours Nvidia hardware. I've even seen it perform better on a gts250 compared to a 4870 in one benchmark.
Hopefully this will be an eye-opener of sorts for future reviews. WiC will always favour Nvidia hardware, Crysis will always favour i7 and Grid will always favour ATI. I'm quite interested in how well the i7 with 4870's competes with the Dragon however - If that one ends up a decent win for the i7 then that would be real proof of how much faster the i7 is than the 955.
My reckoning is, it will be a tie. 😀
The reason i loved this article was actualy couse it was rigs vs rigs. A porly optimized dragon rig, but still. There been 1000 thousends of cpu vs cpu reviews, and gpu vs gpu reviews. The benches was to few and i did as most of us AMDs user feel a bit supprised. Picking benches does change the picture a lot.
http://www.neoseeker.com/Articles/Hardware/Reviews/pii955
This review even show that in some benches 955 trails the i7 940.
But i do hope Cleeve and Toms will contuinu this rig vs rig, a nice twist for their build off would be building one rig with AMD and one with Intel/nvidia for the same amount. Well the high end is a no brainer, thats intel/nvidia.
But i give Toms cudus for the first time since i start reading it, at that is a lot of years ago, this article, was truly interesting, especially for us that love building riggs, i buildt round 100 riggs since i started, and yes i built 1 Intel rig, my first
.
And i like that Cleeve is in the forum reasoning with hes audience. Well kyle on Hardcop also is, but Hardcop really is a , how to put it nice, they sure know how put the phenom in a bad light.
And i hope Cleeve is bold enough to try and put up an real Dragon built rigg for the same amount as the I7-260 sli rigg. Not for only games, but for all purphose use.
http://www.neoseeker.com/Articles/Hardware/Reviews/pii955
This review even show that in some benches 955 trails the i7 940.
But i do hope Cleeve and Toms will contuinu this rig vs rig, a nice twist for their build off would be building one rig with AMD and one with Intel/nvidia for the same amount. Well the high end is a no brainer, thats intel/nvidia.
But i give Toms cudus for the first time since i start reading it, at that is a lot of years ago, this article, was truly interesting, especially for us that love building riggs, i buildt round 100 riggs since i started, and yes i built 1 Intel rig, my first
. And i like that Cleeve is in the forum reasoning with hes audience. Well kyle on Hardcop also is, but Hardcop really is a , how to put it nice, they sure know how put the phenom in a bad light.
And i hope Cleeve is bold enough to try and put up an real Dragon built rigg for the same amount as the I7-260 sli rigg. Not for only games, but for all purphose use.
jennyh
Splendid
- Mar 3, 2009
- 4,292
- 0
- 22,860
Osse pretty much said it in 2 words - 'picking benches'.
There is such a huge disparity between websites benches these days..it wasn't like that 3-5 years ago. I could reel off 100 benches on Anandtech showing how the Phenom II's are a good opponent for i7.
I know it takes a lot of time, but the simplest rule now is 'the more the merrier'. More benchmarks are needed because all it takes is one or two games which are 'optimised' for certain hardware and you could end up with what is a very false picture overall.
There is such a huge disparity between websites benches these days..it wasn't like that 3-5 years ago. I could reel off 100 benches on Anandtech showing how the Phenom II's are a good opponent for i7.
I know it takes a lot of time, but the simplest rule now is 'the more the merrier'. More benchmarks are needed because all it takes is one or two games which are 'optimised' for certain hardware and you could end up with what is a very false picture overall.
Twoboxer
Splendid
- Jan 21, 2008
- 4,208
- 0
- 23,460
This got me thinking, albeit in a different direction. I wondered what would be an appropriate, real-world selection of games?
So I started googling for top ten PC games, by hours played, by units sold, 2008, 2009, UK, US, etc . . . lol, got an interesting array of results. Suffice to say if one were to use *those* lists as a guide, the selection of games to benchmark would look quite a bit different.
Now, I'm not suggesting a change in thinking here, but it did make me wonder who and what these benchmarks are for?
Once you elect to benchmark the "challenging" games, and understand that both cpu and gpu challenges as well as preferences exist, what the hell do you do? Arbitrarily select an even number of each so the benchmarks are "fair"? That's rigged to be even, not fair. Use "units sold" or "hours played" and select by consumer importance? Maybe, where's the cut-off? Or are these benchmarks just to satisfy "us"?
As I tried to make sense of all I have read for my *own* use, with my own (intentionally unspecified) priorities, I had arrived at an i7 cpu and ATI graphics as the best combo available today. My own, anecdotal experience with ATI graphics and drivers has not been good, so I am still ATI-shy. But that does not stop me from advising folks who are not similarly biased to use ATI graphics.
So while I am really, really, really interested in the outcome of the benchmark, I'm not sure how useful it will be in shaping my opinion. Who knows, maybe it'll be magically definitive. Maybe it's just one more data point.
Cleeve
Well same as fps at 5-15 is useless, that goes for fps way over 60, there u have to crank up all settings to see how far u can go and still stay round 60 fps. Ok do u have a screen that can do 100 hz then the relevant top fps is 100, even thou the eys cant see it, it is been told that it is better for the eys, was one of the big gimmicks with 100 hz tv vs 60 hz tv.
Well same as fps at 5-15 is useless, that goes for fps way over 60, there u have to crank up all settings to see how far u can go and still stay round 60 fps. Ok do u have a screen that can do 100 hz then the relevant top fps is 100, even thou the eys cant see it, it is been told that it is better for the eys, was one of the big gimmicks with 100 hz tv vs 60 hz tv.
For some reason my oppinion of what a good buildt rig is didnt show. But in short.
For a gamer rigg, it is 2 major things to consider, it is a combination of cpu/gpu that ensure round 60 Fps in games, especially first shooter games, second and as importent, it is a build that does have huge framedrops, in a first shooter game a framedrop in the wrong place will cost u the victory.
There are sevral things that can reduce framedrops.
A good SSD disk is one
8 gb memory and disk swap file to 0
Not using a Core2 cpu
Tweaking ur cpu use for background tasks
For a gamer rigg, it is 2 major things to consider, it is a combination of cpu/gpu that ensure round 60 Fps in games, especially first shooter games, second and as importent, it is a build that does have huge framedrops, in a first shooter game a framedrop in the wrong place will cost u the victory.
There are sevral things that can reduce framedrops.
A good SSD disk is one
8 gb memory and disk swap file to 0
Not using a Core2 cpu
Tweaking ur cpu use for background tasks
- Jul 24, 2006
- 22,918
- 290
- 57,490
[citation][nom]balancedthinking[/nom]Use Left4Dead, Fallout3, Race Driver Grid and you will see that a balanced plattform by AMD will beat the i7 Intel plattform with inferior graphics subsystem.[/citation]
Don't use GRID or the NVIDIA fanboys will bitch and whine as well. GRID is so shader-heavy that NVIDIA cards don't stand a chance :lol:
Don't use GRID or the NVIDIA fanboys will bitch and whine as well. GRID is so shader-heavy that NVIDIA cards don't stand a chance :lol:
sandmanwn
Distinguished
- Dec 1, 2006
- 915
- 0
- 18,990
BS.
You've been called out. Gather all your THG accounts to help the voting out so you don't look so bad. Regardless the numbers don't match and the benchmarks are biased. You are full of it.
Don aka Cleeve's review is garbage and now so are you.
buckinbottoms
Distinguished
- Jul 1, 2009
- 66
- 0
- 18,630
Good to see the crap article is no longer on the front page. Testament to how much garbage was in it.
Why_Me
Titan
- Jan 17, 2008
- 16,874
- 2,654
- 72,340
[*** up. You are adding no useful input, well, have you ever? Btw, did you ever manage to get that i7 chip out your ass?[/quotemsg]
"Don't hate the playa's Kev, hate the game." 😛
tbo Kev, I used to own a couple AMD rigs back in the day, but that was before I moved out of the trailer park and sold my 1984 Dodge with the 8-track player and fuzzy dice.
"Don't hate the playa's Kev, hate the game." 😛
tbo Kev, I used to own a couple AMD rigs back in the day, but that was before I moved out of the trailer park and sold my 1984 Dodge with the 8-track player and fuzzy dice.
- Status
TRENDING THREADS
-
Review Nvidia GeForce RTX 5080 Founders Edition review: Incremental gains over the previous generation- Started by Admin
- Replies: 151
-
News Reviewer reports RTX 5080 FE instability — PCIe 5.0 signal integrity likely the culprit
- Started by Admin
- Replies: 39
-
News Trump to impose 25% to 100% tariffs on Taiwan-made chips, impacting TSMC
- Started by Admin
- Replies: 37
-
Review Nvidia GeForce RTX 5090 Founders Edition review: Blackwell commences its reign with a few stumbles
- Started by Admin
- Replies: 225
-
-
News DeepSeek's AI breakthrough bypasses industry-standard CUDA, uses assembly-like PTX programming instead
- Started by Admin
- Replies: 39
Facebook
Twitter
Instagram