• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

[H]Fall 2012 GPU and Driver Comparison Roundup

psychodil said:
There is no such thing as a reference cooled GHZ edition card, all of them have decent coolers and (silicon lottery permitting) decent over clocking headroom.
Click to expand...

They seem to be using one in their review :D.

Even worse if they were using a custom cooled one from a partner anyway as Kepler throttles a bit at 70c and the AMD one will be comparatively cooler.

Unfair to have one throttling and the other going full pelt nice and cool.

More holes! ;)
 
boost.jpg


See how mine boosts?
 
Rusty0611 said:
Roughly. Mine was a good en and boosted to 1110 MHz. Not bad for a reference :D.

@TheRealDeal can you read my post above? There ain't no doubt how bad this review is.
Click to expand...

It is also easy for them to pick a really good booster and most will agree they have been a little anti amd for awhile.

there might not be as much headroom on a ghz edition but its already faster and amd seem to gain more from overclocking.

We never really done stock clocks in the ocuk benchmark thread so who knows how those results would have looked.

I am not disputing when both cards are overclocked they are a match for each other but this was stock clocks and not 1080p.
 
But it was 1080p....

Edit: Also as I said above, they're widely believed to be even at 1080p while overclocked yet the GHz edition according to this review is 15% faster. So are you saying to get even the 680 overclocks 15% more?

Of course it doesn't. That's one big hole in it :D
 
Last edited:
Rusty0611 said:
They seem to be using one in their review :D.

Even worse if they were using a custom cooled one from a partner anyway as Kepler throttles a bit at 70c and the AMD one will be comparatively cooler.

Unfair to have one throttling and the other going full pelt nice and cool.

More holes! ;)
Click to expand...

There are no reference cooled GHz edition 7970's for sale, the reviewers received test mules

Its unfair on us consumers that nvidia gimped their cards to throttle I agree. Its not like throttling 30mhz out of 1100 is going to significantly affect the results anyway.
 
TheRealDeal said:
Missed the 1080p apples to apples. Anyhow as i said this was done on stock cards not overclocked. Amd just got a major boost and clock for clock were similar before the boost in the single player.
Click to expand...

I seriously don't care what is better but I want to know the facts. [H] are not doing this and leaving it to 'guess work', which I can do easy enough.
 
psychodil said:
There are no reference cooled GHz edition 7970's for sale, the reviewers received test mules

Its unfair on us consumers that nvidia gimped their cards to throttle I agree. Its not like throttling 30mhz out of 1100 is going to significantly affect the results anyway.
Click to expand...

I think you've missed the point. It's not about Nvidia gimping their cards if you want to term it that. It's about doing a proper like for like comparison.

Anyway of course it will. Look how much the GHz gained from a small ish overclock out the box.

It's not exactly difficult to keep one under 70c with a custom fan profile but the point was more that they haven't probably done that - who knows they haven't said anything - so it skews the results again if they're comparing against a custom cooled model.
 
TheRealDeal said:
Missed the 1080p apples to apples. Anyhow as i said this was done on stock cards not overclocked. Amd just got a major boost and clock for clock were similar before the boost in bf3 single player.
Click to expand...

Rusty0611 said:
Edit: Also as I said above, they're widely believed to be even at 1080p while overclocked yet the GHz edition according to this review is 15% faster at stock. So are you saying to get even the 680 overclocks 15% more?

Of course it doesn't. That's one big hole in it :D
Click to expand...

:)
 
Rusty0611 said:
According to the benchmark thread they weren't.

12.11 made it swing depending on the OC.
Click to expand...

I'm talking stock 680 vs stock GHz edition (1050mhz). They were very even overall. TPU made the GHz edition 2% faster at 1080p across their broad selection of games. Its not unfeasible that after 12.11, the GHz edition could be around 10% faster than the stock 680 across a broad selection of games. 15% however seems a bit fishy.
 
psychodil said:
I'm talking stock 680 vs stock GHz edition (1050mhz). They were very even overall. TPU made the GHz edition 2% faster at 1080p across their broad selection of games. Its not unfeasible that after 12.11, the GHz edition could be around 10% faster than the stock 680 across a broad selection of games. 15% however seems a bit fishy.
Click to expand...

Bf3 was said to have gains of around 15-20% so its not that surprising to me.
 
TheRealDeal said:
Bf3 was said to have gains of around 15-20% so its not that surprising to me.
Click to expand...

I'm bored pointing out the flaws of the review if you're just going to ignore them. You've not answered my 15% overclock point.

Plus a few little bits here and there such of a low boost, thermal throttling etc and you could quite conceivably reach a 15% difference.

If they're roughly even at 1080 (Lightning vs Lightning) then we can also say the 680 overclocks better :D. What you have said over the last couple of pages doesn't really hold up in the face of the numbers in that review alone let alone what guys in the benchmark thread are seeing.

Anyway this is why the benchmark thread is more interesting to me. No paid off dodgy reviewers who don't really know all that much about the hardware they're testing.
 
Last edited:
The large gains in BF3 for AMD on 12.11 was iirc for the 7870, which hit something like a 16% gain, larger gains then any other AMD card on 12.11.

This reminds me I need to get back to work on the benchmarks thread asap ;)

Edit: To back up rustys 'even at 1080' statement, here is a clock vs clock comparison of two very highly clocked Lightnings from both camps:

680%20vs%207970%20BF3_2.jpg


Difference of around 7% at very best between a 7970 and 680, taking into account the new Nvidia drivers are supposed to give a boost in BF3, this is probably closer to 2%.
 
Last edited:
pgi947 said:
The large gains in BF3 for AMD on 12.11 was iirc for the 7870, which hit something like a 16% gain, larger gains then any other AMD card on 12.11.

This reminds me I need to get back to work on the benchmarks thread asap ;)
Click to expand...

Yes you do! Do you need some 7950 data on some games? Difficult for me to check on me phone :).
 
Rusty0611 said:
I'm bored pointing out the flaws of the review if you're just going to ignore them. You've not answered my 15% overclock point.

Plus a few little bits here and there such of a low boost, thermal throttling etc. Anyway this is why the benchmark thread is more interesting to me. No paid off dodgy reviewers who don't really know all that much about the hardware they're testing.
Click to expand...

Its a stock card which answers all your question and overclocking does not come into. Thermal throttling and again its a stock card with stock fan profiles. This is stock v stock end off no fiddling just plug and play performance. None of the cards tested here were stock models and hence they probably boosted much further due to better cooling and no heat throttling on the gtx680. No review is ever without issue and these guys have been with nvidia all year handing out reward after reward.

As you can see above in the graph the 7970 gained 20%.
 
Rusty0611 said:
Yes you do! Do you need some 7950 data on some games? Difficult for me to check on me phone :).
Click to expand...

Please mate, just bung them in a text file or something for me, I'm going to create charts stacking all cards vs each other for each game/setting/resolution for easier viewing :)
 
You must log in or register to reply here.
Back
Top Bottom