Athlon X4 845 ExcavatorIPC benchmarks

[csbin]

https://www.reddit.com/r/Amd/comments/48wk6q/just_got_my_amd_athlon_x4_845_here_are_some/



it doesn't have an unlocked multiplier

3.5G Base / 3.8G Turbo

 

[Sweepr]

https://www.reddit.com/r/Amd/comments/48wk6q/just_got_my_amd_athlon_x4_845_here_are_some/

40% better than this would still be slower than Intel's architectures from recent years.

 

[csbin]

VS FX-8800P

(15W) 2100 / 3400 Single Channel
​

​

​

​

 

[ShintaiDK]

40% better than this would still be slower than Intel's architectures from recent years.

Yep, it would at best put it around SB performance.

 

[csbin]

40% better than this would still be slower than Intel's architectures from recent years.

Carrizo Not L3 Cache

 

[VirtualLarry]

My 4.455Ghz G4400 scores 2100 / 4000 for CPU-Z benchmark, in 1.75.0.

Edit: More accurately, I just ran the benchmark again, with other stuff running in the background, and I got 2261 / 3956.

 

[ShintaiDK]

Carrizo Not L3 Cache

And this is the part where you add some documentation for what the L3 does in CB11.5 ST scores. The answer seems to be nada.

Lets see how one with 8MB performs.

The 6800K is still king of CB11.5 in terms of AMD APUs.

 

[lixlax]

More benches would be nice and especially if put against other L3'less AMD chips (Trinity/Richland, Kaveri). And keep in mind that Cinebench have been one of the worst case scenarios for BD architectures (perfomance wise vs Intel).

I think L3 does not have much impact in Cinebench?

 

[ShintaiDK]

Is CB the worst case? I thought it was games and browser benchmarks and so on

 

[Arachnotronic]

More benches would be nice and especially if put against other L3'less AMD chips (Trinity/Richland, Kaveri). And keep in mind that Cinebench have been one of the worst case scenarios for BD architectures (perfomance wise vs Intel).

I think L3 does not have much impact in Cinebench?

According to the CPU-Z benchmarks posed by csbin, XV has roughly the ST perf/clock of a Sandy Bridge.

 

[csbin]

And this is the part where you add some documentation for what the L3 does in CB11.5 ST scores. The answer seems to be nada.

Lets see how one with 8MB performs.

The 6800K is still king of CB11.5 in terms of AMD APUs.

http://www.anandtech.com/bench/product/1262?vs=1200





Cinebench 11.5 - Single Threaded/Multi-Threaded


AMD A10-6800K@4.1 GHz, 1.12/3.56

AMD A10-7850K@3.7 GHz, 1.03/3.6

 

[Sweepr]

AMD A10-6800K@4.1 GHz, 1.12/3.56

AMD A10-7850K@3.7 GHz, 1.03/3.6

Lol, you just confirmed what he said.

The 6800K is still king of CB11.5 in terms of AMD APUs.

http://www.extremetech.com/wp-content/uploads/2014/01/CB115.png

Thanks for sharing. So 4.1GHz Piledriver is still faster than 3.8GHz Excavator here. Quite a bit less than the 20% IPC bump mentioned here:

www.planet3dnow.de/cms/wp-content/uploads/2015/08/cinebench115single.png

Edit: Even assuming 4.4GHz (A10-6800K), still less than 10% faster per clock.

>10-20% higher clocked Excavator is slower than Nehalem @ CB 11.5 (another example).

 

[csbin]

Lol, you just confirmed what he said.



Thanks for sharing. So 4.1GHz Piledriver is still faster than 3.8GHz Excavator here. Quite a bit less than the 20% IPC bump mentioned here:

www.planet3dnow.de/cms/wp-content/uploads/2015/08/cinebench115single.png

A10-6800K Richland 4.1 GHz (4.4GHz Turbo) D:

 

[ShintaiDK]

Lol, you just confirmed what he said.



Thanks for sharing. So 4.1GHz Piledriver is still faster than 3.8GHz Excavator here. Quite a bit less than the 20% IPC bump mentioned here:

www.planet3dnow.de/cms/wp-content/uploads/2015/08/cinebench115single.png

>10% higher clocked Excavator still slower than Nehalem @ CB 11.5.

PR stays PR

 

[ShintaiDK]

A10-6800K Richland 4.1 GHz (4.4GHz Turbo) D:

What part is it you dont get? 6800K is faster, yes or no?

 

[csbin]

what part is it you dont get? 6800k is faster, yes or no?

ipc ipc ipc

originally posted by sweepr
lol, you just confirmed what he said.



Thanks for sharing. So 4.1ghz piledriver is still faster than 3.8ghz excavator here. Quite a bit less than the 20% ipc bump mentioned here:

 

[ShintaiDK]

ipc ipc ipc


[/i]

Yes, but reduced clocks. So why would a 6800K owner want to upgrade in the end of the day. Sitting with a faster product.

Whats next, Zen with 3Ghz clocks?

 

[nismotigerwvu]

I think the more interesting chips are the upcoming AM4 models, but it's nice to see Excavator making it to FM2 if only in a single chip. That and few synthetic benchmarks aren't worth reading into that deeply. Besides, we've known for awhile what sort of performance Excavator brings to the table from the Carrizo. No magical scaling to see here.

 

[csbin]

Lol, you just confirmed what he said.



Thanks for sharing. So 4.1GHz Piledriver is still faster than 3.8GHz Excavator here. Quite a bit less than the 20% IPC bump mentioned here:

www.planet3dnow.de/cms/wp-content/uploads/2015/08/cinebench115single.png

Edit: Even assuming 4.4GHz (A10-6800K), still less than 10% faster per clock.

>10% higher clocked Excavator still slower than Nehalem @ CB 11.5.

Even assuming？D:


http://www.anandtech.com/show/7032/...ll-gpu-on-the-desktop-radeon-hd-8670d-hd-4600

 

[lixlax]

Based on XV values that Shintai posted and Richland/Kaveri values posted by csbin.
Points per GHz:

Single thread:
Richland: 0,255
Kaveri: 0,256
Carrizo: 0,282 (10% faster than Richland)

Multithreaded:
Richland: 0,87
Kaveri: 0,97
Carrizo: 1,1 (26% faster than Richland).

I assumed that all chips run at their max turbo clocks in single threaded and at their base clocks at multithreaded.

 

[Sweepr]

Even assuming？D:


http://www.anandtech.com/show/7032/...ll-gpu-on-the-desktop-radeon-hd-8670d-hd-4600

Less than 10% faster per clock? Unless there's something wrong here, the >20% bump in artificial tests failed to materialize in actual desktop products.

MT looks better because it basically eliminated the scaling penalty, but that was expected.

 

[ShintaiDK]

Even assuming？D:


http://www.anandtech.com/show/7032/...ll-gpu-on-the-desktop-radeon-hd-8670d-hd-4600

Funny thing. The 65W Richland will still beat the 65W Carrizo in ST

 

[AtenRa]

A10-6800K = 100W TDP

Athlon X4 845 = 65W TDP.

Energy consumption its not important for ShintaiDK suddenly

Funny thing. The 65W Richland will still beat the 65W Carrizo in ST

How did you come to that conclusion ??

 

[Sweepr]

Throwing some 2009 Phenom II action in the mix:

Still faster than Excavator per clock here.

 

[csbin]

Funny thing. The 65W Richland will still beat the 65W Carrizo in ST

GPU Core 256/384