Upcoming AMD Richland A10-6700 - Radeon HD 8670D tested!

[Cloudfire777]

Hi everyone. Newcomer on this forum, not a stranger to other boards. Got banned on overclock.net (some moderators there didnt like me, lol) so I`m moving in here. Hope you don`t mind

Got hands on the first benchmark scores on the upcoming AMD Richland A10-6700 with the Radeon HD 8670D


CPU-Z of the AMD A10-6700

CPU-Z - Radeon HD 8670D

3DMark Cloud Gate with the Radeon HD 8670D

Comparison against A10-5800K (7660D) which is a 100W APU and the 3570K (HD4000) which is a 77W CPU.

A10-6700 (8670D)
Total score: 6350
Graphic score: 8933

A10-5800K (7660D):
Total score: 5645 (A10-6700 perform 12% better)
Graphic score: 7418 (HD 8670D performs 20% better)

i5-3570K (HD4000):
Total score: 5348 (A10-6700 perform 19% better)
Graphic score: 5545 (HD 8670D performs 61% better)

---


3DMark Fire Strike with the Radeon HD 8670D

A10-6700 (8670D)
Total score: 1131
Graphic score: 1212

A10-5800K (7660D):
Total score: 919 (A10-6700 perform 23% better)
Graphic score: 987 (HD 8670D performs 23% better)

i5-3570K (HD4000):
Total score: 630 (A10-6700 perform 80% better)
Graphic score: 656 (HD 8670D performs 85% better)

'
---

3DMark Ice Storm with the Radeon HD 8670D

A10-6700 (8670D)
Total score: 67462
Graphic score: 86027

A10-5800K (7660D):
Total score: 56098 (A10-6700 perform 20% better)
Graphic score: 65657 (HD 8670D perform 31% better)

i5-3570K (HD4000):
Total score: No data (worse than 5800K)
Graphic score: No data (worse than 7660D)

 

[Phynaz]

Welcome to the forum. Are these your results or are they from elsewhere? Do you have a link to the source?

 

[Durp]

I wish these APU's at least had the GPU power of entry level cards generations ago like the HD6850/HD7770/GTX460.... This is still soooooo slow. Progress is progress i guess.

 

[Homeles]

Got banned on overclock.net (some moderators there didnt like me, lol) so I`m moving in here.

Welcome to the club.

I wish these APU's at least had the GPU power of entry level cards generations ago like the HD6850/HD7770/GTX460.... This is still soooooo slow. Progress is progress i guess.

They're nice for laptops. Especially since they still use crap resolutions.

 

[2timer]

Kind of a baited post for your very first post.

It appears to me as if you posted 3 iterations of the same test, eg 3DMark. Which is fine, except the test results aren't going to show a great difference, since they are more or less similar.

The 2nd to last graph you posted with what looks like German on the bottom, what was the point of that? It doesn't even have the A10-6700 on it.

 

[Phynaz]

What's with the 1/1/2011 date?

 

[Cloudfire777]

Welcome to the forum. Are these your results or are they from elsewhere? Do you have a link to the source?

Not mine
http://www.coolaler.com/showthread.php/301890-%E5%99%93-%E4%BB%80%E9%BA%BC%E6%98%AFHD8670D!-AMD%E6%9C%80%E5%BC%B7%E5%85%A7%E9%A1%AF%E5%8D%B3%E5%B0%87%E7%8F%BE%E8%BA%AB!!-%E4%BD%8E%E8%AA%BF%E6%85%A2%E6%85%A2%E7%9C%8B%E5%B0%B1%E5%A5%BD

Welcome to the club.
They're nice for laptops. Especially since they still use crap resolutions.

Thanks, I`ve seen you at overclock before

Kind of a baited post for your very first post.

It appears to me as if you posted 3 iterations of the same test, eg 3DMark. Which is fine, except the test results aren't going to show a great difference, since they are more or less similar.

The 2nd to last graph you posted with what looks like German on the bottom, what was the point of that? It doesn't even have the A10-6700 on it.

I don`t think you understand what I did. I posted results from a swedish review with scores of the HD4000 (Ivy Bridge) and the 7660D (Trinity) to back up the scores from my difference calculations.

True that the 3 sub scores are from the same benchmark, but I don`t understand why you complain. Should I not post them? It was the only benchmarks I could find, and it shows a real performance gain between Trinity and Richland. The graphic performance of Richland is +25% over Trinity in average for the 3 tests. A 65W APU vs a 100W APU that is...

 

[inf64]

Why you guys always see some conspiracy? I don't see anything strange in these results, they are pretty much in line with clock boost 6700 got in both CPU and GPU parts. Nice thing about this model is the fact that you not only get better CPU and GPU performance than 5800K, you get it at 35W less power draw. That's progress.

Thanks for posting this Cloudfire777! Welcome to the forum.

 

[2timer]

Not mine
http://www.coolaler.com/showthread.php/301890-%E5%99%93-%E4%BB%80%E9%BA%BC%E6%98%AFHD8670D!-AMD%E6%9C%80%E5%BC%B7%E5%85%A7%E9%A1%AF%E5%8D%B3%E5%B0%87%E7%8F%BE%E8%BA%AB!!-%E4%BD%8E%E8%AA%BF%E6%85%A2%E6%85%A2%E7%9C%8B%E5%B0%B1%E5%A5%BD


Thanks, I`ve seen you at overclock before



Baited post?

I don`t think you understand what I did. I posted results from a swedish review with scores of the HD4000 (Ivy Bridge) and the 7660D (Trinity) to back up the scores from my difference calculations.

True that the 3 sub scores are from the same benchmark, but I don`t understand why you complain. It was the only benchmarks I could find, and it shows a real performance gain between Trinity and Haswell

Ok I guess I got a little ahead of myself. I apologize. And excuse my rudeness please - welcome to the forums.

But yeah, the charts are a tad redundant, perhaps. The last one doesn't even include the A10-6700.

 

[Phynaz]

Why you guys always see some conspiracy?

If a new reg posted Haswell benchmarks that were 20% higher than IB wouldn't you ask about the source? I would.

 

[inf64]

These results are actually expected since we already know the spec of 6700 model.
For Haswell we only had a few leaked intel roadmaps with supposed clocks,nothing more. And now we have a preview done by THG so no more mysteries there.

 

[strata8]

Nice thing about this model is the fact that you not only get better CPU and GPU performance than 5800K, you get it at 35W less power draw. That's progress.

Remember there was already the 65W A10-5700 with similar GPU performance to the A10-5800K.

It's nice to see AMD's 20-30% GPU improvement claim was true regardless.

 

[Cloudfire777]

Ok I guess I got a little ahead of myself. I apologize. And excuse my rudeness please - welcome to the forums.

But yeah, the charts are a tad redundant, perhaps. The last one doesn't even include the A10-6700.

Thank you.

All the scores are there, except HD4000 in the last one.

Why you guys always see some conspiracy? I don't see anything strange in these results, they are pretty much in line with clock boost 6700 got in both CPU and GPU parts. Nice thing about this model is the fact that you not only get better CPU and GPU performance than 5800K, you get it at 35W less power draw. That's progress.

Thanks for posting this Cloudfire777! Welcome to the forum.

Its ok that they are sceptical, although it is an Engineering sample that have been tested, its probably very close to the truth since like you say, the results are in line. Truly a great feat by AMD to get 25% better IGP performance with a 65W APU since its the same architecture as Trinity

 

[inf64]

Remember there was already the 65W A10-5700 with similar GPU performance to the A10-5800K.

It's nice to see AMD's 20-30% GPU improvement claim was true regardless.

Yes that's correct, but CPU performance was lower than 5800K. Now you get better CPU/GPU performance than both 5700 and 5800K,all that in 65W.

 

[2timer]

Yes that's correct, but CPU performance was lower than 5800K. Now you get better CPU/GPU performance than both 5700 and 5800K,all that in 65W.

Have we seen improved CPU performance in similar TDP ranges for Richland yet? I know GPU performance yes, but I thought no with CPU.

 

[strata8]

Have we seen improved CPU performance in similar TDP ranges for Richland yet? I know GPU performance yes, but I thought no with CPU.

The leaked clocks are higher so you'd expect improved CPU performance

I don't think the 6700 will beat the 5800K in CPU performance, maybe slightly in single-threaded (due to higher turbo) but that's it.

 

[Phynaz]

Looking at that chart what would account for the increased performance? Max clocks are less that 10% higher. Hasn't AMD said there are no architecture changes?

 

[Sleepingforest]

The clock increase can only account for a small amount of the performance increase (as it is roughly 5% higher on both the CPU [with turbo] and GPU side)--the rest is probably due to switching over to Steamroller from Bulldozer "enhanced Piledriver". It's nice to see AMD pushing hard in the budget market. This brings it to roughly GT640 levels, assuming the benchmark scores will scale with the framerate, which is pretty acceptable for budget gaming (assuming the price is around $130).

 

[Olikan]

Looking at that chart what would account for the increased performance? Max clocks are less that 10% higher. Hasn't AMD said there are no architecture changes?

they updated the IMC to support 2133Mhz memories...so, maybe it got improved

i can't see the new turbo improving that much

 

[Phynaz]

the rest is probably due to switching over to Steamroller from Bulldozer.

Richland is using Steamroller cores?

 

[Centauri]

'enhanced Piledriver'

 

[strata8]

Looking at that chart what would account for the increased performance? Max clocks are less that 10% higher. Hasn't AMD said there are no architecture changes?

I think they've said there are some small (on the order of 2-3%) IPC improvements and a rearranged GPU layout which could contribute to the graphics improvements.

 

[monstercameron]

then this does bode well for the 6800k...

 

[Phynaz]

'enhanced Piledriver'

What is enhanced?

 

[Rvenger]

Richland = Piledriver 2.0 with enhanced IMC. Its an improvement on the current architecture. AMD is slowly optimizing the once very slow and power hungry dulldozer. Nice to see them make good improvements!