If this is your first visit, be sure to check out the FAQ by clicking the link above. You may have to register before you can post: click the register link above to proceed. To start viewing messages, select the forum that you want to visit from the selection below.
![]() |
|
|
#1 |
|
Member
Join Date: Jan 2008
Posts: 460
|
I am using a Phenom 2 ...and the itch to upgrade to i7 (4ghz! 6Gb!) is coming..but i checked a few benchmarks..and most showed that i7 works best at ultra low in game settings...no gpu bottleneck duh! When games are at full settings, why does Phenom2 setup runs a smitten smoother? See below
http://www.tomshardware.com/reviews/...g,2588-13.html Is it something to do with the caches? IPC? triple channel latency? PCIe? and..why does Anandtech conclude that AMD gpu work better on Intel setup and Nvidia gpu work better on AMD setup!? |
|
|
|
|
|
#2 |
|
hardware monkey
Join Date: Mar 2007
Posts: 3,907
|
There was a rumor to this effect prior to the launch of i7, but if you read some decent review sites (i.e. not Tom's) you'll find the i7 does out-perform all other chips on the market, particularly when using multiple GPUs (you know, when actually shifting the bottleneck to the CPU).
|
|
|
|
|
|
#3 | |
|
Red-headed step child
Join Date: Jun 2004
Location: Guess ;)
Posts: 3,084
|
I really think you missed the obvious point here, gongo.
If all six of those processors had an aggregate performance delta of ~3%, then it's obvious that CPU power wasn't the bottleneck of the test. It's even more obvious when you compare something ilke an i3 to an i7 and get a difference of only 0.5%. So, your original concern over of whether an i7 is "good" at games or not is pretty much moot -- it's only as good as the video card. But if you get enough video horsepower attached to that system (ie, multiple cards in SLI / CrossFire) then you'll see the i7 start ripping the Phenom apart. EDIT: They even state just as much at the end of their article: Quote:
__________________
"...twisting my words" |
|
|
|
|
|
|
#4 |
|
Member
Join Date: Jan 2008
Posts: 460
|
I get what you guys meant by multi-GPU scaling...but what i saw around was regarding a single GPU whereby on max in game settings, the i7 seemed to lose a few fps to a similar Phenom 2 build....or was i dreaming things..in any case, could there be some limits wrt to i7 architecture ...which loses steam faster in GPU limited solution...could it be the HT?
|
|
|
|
|
|
#5 | |
|
Regular
Join Date: Mar 2007
Posts: 8,986
|
Quote:
The only test settings posted are those for the i3-530 and the Phenom II x3 720. In which case, are we really splitting hairs over 0.19% difference in speed. Heck we don't even know if the i7-870 or the Phenom II x4 965 are using the same respective MB's and memory as the lower cost parts. So about the only thing we can infer from this review, is that as they said. When your graphics card is the limiting factor, then it's farily moot what CPU you use as long as it is fast enough for the settings you want. Lower the settings if things are too slow for you and you may or may not get a healthy boost in game speed, again, depending on your settings. It is interesting that the i7-870 takes a dive at 2560x1600, however. Indicating a possible issue with memory transfer speeds from main memory. It's possible Clarkdale deals with that better than Lynnfield. Regards, SB |
|
|
|
|
|
|
#6 | |
|
Red-headed step child
Join Date: Jun 2004
Location: Guess ;)
Posts: 3,084
|
Quote:
At the highest graphical resolution and highest graphical settings, what is the most likely bottleneck? The central processing unit, or the graphics processing unit? Here's a hint: it's a rhetorical question. When multiple GPU's are installed (ie, the graphics are no longer the bottleneck) then the i7 series completely destroys the AMD processors. So, when the graphics cards are the bottleneck, then it becomes moot. Make sense?
__________________
"...twisting my words" |
|
|
|
|
|
|
#7 |
|
Member
Join Date: Mar 2002
Location: a vertex
Posts: 354
|
|
|
|
|
![]() |
| Thread Tools | |
| Display Modes | |
|
|