• 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.

ATI cuts 6950 allocation

I agree that the article is unfair to AMD:

The 6870 and 6850 are excellent products, and will have no real competition until the GTX 560 series is released. Also to say that AMD can compete "everywhere but the top" is complete rubbish... They have been the only manufacturer with a dual-GPU card for the past 6 months, and have held the "performance crown" (with the 5970) for around a year.

Still, unless there are some extremely odd benchmarking setups at work, the comments about performance being roughly on-par with the GTX570 are hard to ignore, especially as they fit in line with the leaked benchmarks we've started seeing these past two days.
 
From the article that mack-attack linked to:



So 1536 SP in all likelihood...



Sadpanda.

I suppose I knew it was coming if the card was 1536SP, but I still thought that it might trade blows with the GTX580 in one or two scenarios.

Even 1536SP should be faster than what we've seen in the screenshots purporting to be 6970 scores tho - which does give weight to the "turbo" setting above imo as the scores are about 20% short of what 1536SP should give up.
 
Even 1536SP should be faster than what we've seen in the screenshots purporting to be 6970 scores tho - which does give weight to the "turbo" setting above imo as the scores are about 20% short of what 1536SP should give up.

Well, my estimate for a 1536SP Cayman was around 10% slower than the GTX580. So if the 6970 is performing at roughly GTX570 levels, then that's more like 5% or so behind what I was expecting.

I think some people have been assuming too much from the 4D shaders (in relation to Barts performance), and perhaps too close to linear performance scaling with the increased shader count.

As always though, we'll only know for sure on Wednesday :p
 
Well, my estimate for a 1536SP Cayman was around 10% slower than the GTX580. So if the 6970 is performing at roughly GTX570 levels, then that's more like 5% or so behind what I was expecting.

I think some people have been assuming too much from the 4D shaders (in relation to Barts performance), and perhaps too close to linear performance scaling with the increased shader count.

As always though, we'll only know for sure on Wednesday :p

HD6970 being 25% faster than HD6870? Seriously? What is HD6950 going to be like? :rolleyes:
 
That is a good question, the numbers we've seen so far don't leave much space between the 6870 and the "6970" even with a 20% boost.
 
It seems really strange that all the 'leaked' benchmarks we've been seeing seem to suggest that the 6970 is barely any faster than the 5870 - at this rate I have to wonder how the 6950 is going to be any faster than the 6870!
 
**** poor springs to mind.

So dissapointed. everything might change in the next 3 days though.
We might get a 50% increase in performance, if pigs could fly....
 
Last edited:
Maybe we are seeing a 2900xt situation here

pathetic performance earlier , Then with new drivers very good performance

Sadly not.
2900xt was new aritechture from ground up.
6900 are not. Despite drunkenmaster saying otherwise.
They barts on steroids with bolted on "improved" 4d shaders.
Hardly revolutionary.
 
I can't believe people are actually entertaining the thought that these are going to be about the same performance as 5870s. What the hell? :confused:. People really do go crazy around new releases. For AMD to release such cards would be a complete and utter joke.
 
The implication is that they are around the performance of a GTX570 (or GTX480), which would place them a fair chunk faster than the 5870. I haven't seen anything to suggest they would be only 5870 performance.

Anyway, I suspect they will come out faster than the 570/480 overall, even if it requires the 10.12 driver. The slide for the driver suggested that it was available from Dec 1st, so I imagine that most of the review sites should have had access to it for their reviews.
 
3dm11xtreme.png


look at the top
 
(image)
look at the top

That and also the fact that he's using driver version 8.790.6.0, which is a test driver.

Supposedly, the production driver's version is 8.801:
96904816.jpg


EDIT:
Benchmark Reviews were also using the same old driver as in the German benchmarks in the article where they are saying AMD is losing the GPU battle. I pointed the same out and posted a comment, but they always keep deleting it. Why are they afraid of the new driver?
 
Last edited:
I'll reserve final judgement until the card is actually out but at 1536 shaders this card is starting to look like a joke release.

I'll explain why.

As the fastest single GPU card you can rightly choose a premium, as soon as you drop below that then in my mind you are suddenly competing with all the previous gen.

A 6970 at 25% faster than a 5870 for 100% more price is a joke
 
Back
Top Bottom