I took my 920 all the way to 4ghz but it made no difference on my score, so I dropped down to 3ghz. I might just set it back to stock settings because really I saw no benefit from it at all, and after monitoring CPU usage for everything else I do, there is no need to OC.
I got a 4911 on high with a 5870 with full GPU usage and little CPU load. I'll run again and take some screens.
Shader clocks on Fermi cards are locked to the core, 1:2 ratio. Its really irrelevant regardless, its like asking why an i7 @ 2.8ghz is faster than your q6600 @ 3.0.
I7's are safe to run but best under 80C
50-60 = OC it!
70C = fine
80C = not good to run long periods of time
90C = dangerous/going to crash at some point
100C = boil some water n make coffee on your cpu
These are at load temps, right?
Also, here are the screens of my most recent benchmark on high.
Spoiler: show
That is actually my highest score yet, and that is after I took my CPU down from 4ghz back down to stock. The GPU is OC'ed as you can see though. I tried taking it to 1ghz but it glitched out almost immediately.
Hopefully on load, if your idling near 80C... somethings wrong haha
I got a crappy chip because no matter what I try I can't get 4ghz without it reaching 80C load in prime95. Not new to OCing either. But I never keep a CPU for more than 2-3 years so even if I'm cutting its lifetime in half, doesn't really matter to me.
Say it's to last 10 years, say with my abusive temps it'd last 5, but I upgrade after 2-3 years, so yeah.
I got D3DOverrider and it lets you force it on directx, I used it for assassin's creed II and it worked well. I get alot of split second fps hiccups in that game, I'll be at 70+, drop to 40-something for a split second then go back up to 70, with the games in-game vsync it made my fps annoyingly low, with triple buffering I'm always 55-60 which is perfect.
The weird thing is, I thought with normal vsync fps could only be 60-30-15-etc. but I get ~46fps with normal vsync.
oops double post
So if I can trust that test I can definitely go for 5770 1GB instead of GTX 460 768MB? Also 5770 is ~40e cheaper.
Is there any other test between 5770 1gb and 460 768Mb?
Which one I should go? I'm building new computer for this game. And I use 1920x1200 resolution and maybe second 19" monitor.
I would wait before picking the 5770 over a 460, as the 460 is a good bit faster than the 5770 in essentially every other test. This points to some funky shit going on with nvidia/se/drivers. If you have to order now, then shit I dunno what to tell you.
I gambled with the 460 myself lol.
http://www.overclock.net/nvidia/7781...ck-review.html
XFX cards have been on sale alot on newegg lately, they had a 5850 with -$50 off a few weeks ago, which made it only $20-30 more expensive than a 460 1GB.
460 is faster than a 5850 most of the time, let alone a 5770.
Also, glad I did my testing when I did. Seems like we're capped at 60 FPS now.
Temperatures still outrageous when GPU gets to full load. Lowering the cap seems like a bandaid fix for this since you probably need AA/AO settings to get there now.
Hope it doesnt cause the feedback to go away, it's still a major issue.
This is for the new PC I just made.
CPU: I7 930 @ 3.885 GHz (My everyday stable OC)
GPU: Gigabyte Radeon 5870 Super Overclock 1GB
RAM: 12 GB Corsair Dominator 1600MHz 7-8-7-20, running at @1480MHz
HDD: 10000rpm 600GB Western Digital Velociraptor
High: 5318
http://img844.imageshack.us/img844/5...885cropped.jpg
Low: 7524
http://img16.imageshack.us/img16/1324/7524low3885.jpg
Nice voltage for your processor - should take it higher!
Also, we need to stop having the fanboism with video cards running rampant around here. Cel - I'm not sure where you get your info talking about how Nvidia cards are better than their ATI counterparts. The only one I see that is consistently better is the GTX480>5870 but the 460/470 is not better than that 5850/5870 at all except maybe in metro. The truth is that Nvidia and ATI cards trade blows but you can't sit here and say that 460 and 470 are a good deal better than a 5850/5870 consistently.
Here is recent read-up about the DX11 games being benched by a plethora of cards
http://techreport.com/articles.x/19404/1
Just updated today.
http://www.videocardbenchmark.net/high_end_gpus.html