[wccftech] amd a10-6800k 1st multi-threaded benchmarks?

[IonusX]

http://wccftech.com/amd-a10-6800k-richland-apu-performance-unveiled/
ladies, gents,... trolls the a10--6800k has been benchmarked (supposedly)
here are your multi-threaded scores

AMD A10-6800K:
3DMark 11 Performance - P1773
3DMark 11 Performance OC - P2221
Fritz Chess Benchmark - 7512 Kilonodes/s
AIDA 64 Cache and Memory - 12696/10331/18001/59.3ms (Read/Write/Copy/Latency)
CineBench R11.5 - 3.70 Multi-Threaded

pictures of results are in the link

 

[inf64]

C11.5 scores are in line with 4.4Ghz clock . My 750K Athlon @4.3Ghz (basically 2M PD with deactivated iGPU) scores 3.62pts. So it seems Richland is able to clock up to 4.4Ghz across all 4 cores in MTed benchmark(s) which is a nice boost vs 5800K's 4Ghz in such scenario.

 

[SiliconWars]

Yeah 10% for a basic speedbump isn't bad considering Haswell will average around that.

 

[Enigmoid]

Hmmm... so

CPU

Cinebench R11.5
a10-5800k score is ~3.31
a10-6800k is ~3.70, increase of ~12%.
Frequency increase of ~10%.

(note llano had a score of ~3.5 on this test--however, in many real world benchmarks llano was slower).

GPU

3dmark 11
a10-6800k (2400 mhz RAM)- 1616 GPU score
a10-5800k (1866 mhz RAM) ~1450 gpu score

Looks like a fairly good boost in cpu performance (~12% from that one benchmark) and a very minor jump in igp performance.

(note tom's RAM timings are slightly better)

3dmark score a10-6800k at 2400 mhz ram is 1773, a10-5800k is 1683, performance jump of 5%, right around the igp speed jump (800 to 844 mhz) which is a boost of 5.5%.

Desktop trinity (and I said many times) would be rather underwhelming. Expect the small gains you see in frequency and no more (yet so many thought richland would bring 10-20% CPU and GPU performance gains). Richland is a mobile chip, the power savings will easily be seen in mobile but on the desktop little changes.

Pointless for anyone with trinity. Just do a small OC and you are there cpu wise (5% more gpu power is undetectable, maybe 1-2 fps but also easily OC-able). Most a10-5800k could easily hit that small overclock. Of course richland power consumption might be a little lower.

Of course this is only one review and we will have to wait for other reviews to verfiy this.

 

[Olikan]

CPU

Cinebench R11.5
a10-5800k score is ~3.31
a10-6800k is ~3.70, increase of ~12%.
Frequency increase of ~10%.

well, richland have a better turbo....
so might sustain better the high clocks, the missing 2%

the igp is bandwidth starved anyway....turbo won't help

 

[riotr911]

Did anyone notice the 48 TMUs? I really hope that's not a mistake because I would like to see how it affects real world benchmarks. The OC score is about the same as a stock GTS 450, which is really good.

 

[SPBHM]

it's the same as Trinity, "24 TMUs".
GPUz is wrong.

 

[IonusX]

it's the same as Trinity, "24 TMUs".
GPUz is wrong.

actually this might be wrong, and im not really surprised the performance is so low it isn't a new core, richland is piledriver 2.0 not steamroller
also i thought id live this little nugget from obr. while obr is FoS a lot of the time this is rather interesting that there results match these ones

also here is their pcmark 7 result:

where she basically broke even with a phenom II x4 980