Launch of the AMD FX 8 Core processor - Page 3

Christenson · 2011-10-14, 00:06

Quote:

Originally Posted by Dubslow

Even though this would work fine, it still seems you'd get better value/throughput by waiting for Ivy Bridge

You are neglecting to do something important with this old saw:

I'll top up the GPU load with CUDALucas....whatever the base processor!

Cybertronic · 2011-10-15, 16:10

Is there any informations how fast is pfgw on a 8-core bulldozer @ 3.6 GHZ ?

Is the running time of a task at 8 runs the same like 1 run with the same frequenz (without turbo) ?

I heard the perfomance topples to 90% .

Best

kaeptn_kork · 2011-10-16, 20:30

Full p95 benchmarks anyone?
Probably no match for the (cheaper) Thubans though. Maybe next time.

firejuggler · 2011-10-19, 00:01

meanwhile, on Intel Side Ivy bridge *early info*

Dubslow · 2011-10-19, 00:47

Ooooohhhhh.... tasty
Just think about how easy to overclock it will be, especially with current heatsinks

Cybertronic · 2011-10-24, 13:39

Hello people, now I have an benchmark test with PFGW.

The result is devastating.

Per core , the 1090T@3.2 GHz is 35% faster than Bulldozer FX-6100 core @3.3GHz.

Number was 5612052289.14489#/5 - 1:

PFGW v1.2.0

1090T: 1,30s
FX-6100: 2,10s

Best

ET_ · 2011-10-24, 15:44

Quote:

Originally Posted by Cybertronic

Hello people, now I have an benchmark test with PFGW.

The result is devastating.

Per core , the 1090T@3.2 GHz is 35% faster than Bulldozer FX-6100 core @3.3GHz.

Number was 5612052289.14489#/5 - 1:

PFGW v1.2.0

1090T: 1,30s
FX-6100: 2,10s

Best

Did you try the new PFGW here?

Luigi

Cybertronic · 2011-10-24, 16:01

Hello Luigi.

No, the new version is 8% faster. The old file is 2.5 MB large...this was better for downloading by my tester.

I know the reason for this "shit" result.
2 Cores takes one FPU. So the running time is double.

The FX-8150 @3.6GHz have 90% Power of 1090T @ 3.2GHz.

If we have 8 cores and 8 FPUs, than we had have 180%,but so ...?

Bulldozer is not relevant for me.

Here the "first "pfgw-benchmark-test:

System : 1090T@3.2 GHz and FX-6100@3.3GHz

http://www.sendspace.com/file/6rvxr7

Dubslow · 2011-10-24, 20:17

Try running just four FPU intensive threads, and see what the results are. Calling it "8 cores" just does not work for our purposes. The talk has been run 4 LL (FPU) threads, and 4 TF (integer) threads using P95. Maybe something like that will give better performance.

Cybertronic · 2011-10-24, 20:46

I can't test once more but, I believe Bulldozer works with the same frequenz ( one task for one FPU )
18% faster than Phenom II. However, effective the Phenom II X6 1090T can do more than the new FX-8150.

moebius · 2011-10-24, 21:19

The Question for me is simple: If you are running 6 X 8 Threads LL Testing on a FX 8150 processor or 8 X 6 Threads LL Testing on a Phemon X6 1100T which one of these is faster at the end?

2011-10-15, 16:10	#24
Cybertronic Jan 2007 Germany 2CD₁₆ Posts	PFGW on Bulldozer Is there any informations how fast is pfgw on a 8-core bulldozer @ 3.6 GHZ ? Is the running time of a task at 8 runs the same like 1 run with the same frequenz (without turbo) ? I heard the perfomance topples to 90% . Best

2011-10-24, 13:39	#28
Cybertronic Jan 2007 Germany 3×239 Posts	1090T vx Bulldozer FX-6100 Hello people, now I have an benchmark test with PFGW. The result is devastating. Per core , the 1090T@3.2 GHz is 35% faster than Bulldozer FX-6100 core @3.3GHz. Number was 5612052289.14489#/5 - 1: PFGW v1.2.0 1090T: 1,30s FX-6100: 2,10s Best Last fiddled with by Cybertronic on 2011-10-24 at 13:40

2011-10-24, 16:01	#30
Cybertronic Jan 2007 Germany 3×239 Posts	Hello Luigi. No, the new version is 8% faster. The old file is 2.5 MB large...this was better for downloading by my tester. I know the reason for this "shit" result. 2 Cores takes one FPU. So the running time is double. The FX-8150 @3.6GHz have 90% Power of 1090T @ 3.2GHz. If we have 8 cores and 8 FPUs, than we had have 180%,but so ...? Bulldozer is not relevant for me. Here the "first "pfgw-benchmark-test: System : 1090T@3.2 GHz and FX-6100@3.3GHz http://www.sendspace.com/file/6rvxr7 Last fiddled with by Cybertronic on 2011-10-24 at 16:22

2011-10-24, 20:46	#32
Cybertronic Jan 2007 Germany 3×239 Posts	I can't test once more but, I believe Bulldozer works with the same frequenz ( one task for one FPU ) 18% faster than Phenom II. However, effective the Phenom II X6 1090T can do more than the new FX-8150. Last fiddled with by Cybertronic on 2011-10-24 at 20:47

2011-10-24, 21:19	#33
moebius Jul 2009 Germany 1011000010₂ Posts	The Question for me is simple: If you are running 6 X 8 Threads LL Testing on a FX 8150 processor or 8 X 6 Threads LL Testing on a Phemon X6 1100T which one of these is faster at the end? Last fiddled with by moebius on 2011-10-24 at 21:21

2011-10-16, 20:30	#25
kaeptn_kork Sep 2002 5² Posts	Full p95 benchmarks anyone? Probably no match for the (cheaper) Thubans though. Maybe next time.

2011-10-19, 00:01	#26
firejuggler "Vincent" Apr 2010 Over the rainbow 2³·5·73 Posts	meanwhile, on Intel Side Ivy bridge early info

2011-10-19, 00:47	#27
Dubslow Basketry That Evening! "Bunslow the Bold" Jun 2011 40<A<43 -89<O<-88 3·29·83 Posts	Ooooohhhhh.... tasty Just think about how easy to overclock it will be, especially with current heatsinks

2011-10-24, 20:17	#31
Dubslow Basketry That Evening! "Bunslow the Bold" Jun 2011 40<A<43 -89<O<-88 3×29×83 Posts	Try running just four FPU intensive threads, and see what the results are. Calling it "8 cores" just does not work for our purposes. The talk has been run 4 LL (FPU) threads, and 4 TF (integer) threads using P95. Maybe something like that will give better performance.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
(windows) How to launch mfaktO/C with processor affinity	swl551	Software	1	2012-09-24 23:29
Intel unveils 50-core maths co-processor card	stars10250	Hardware	16	2011-06-25 17:24
Prime fatal error on 3D launch	kaeptn_kork	Hardware	11	2009-03-17 01:58
Product launch	xilman	Factoring	12	2006-07-11 16:23
Intel Core Duo processor	drew	Hardware	5	2006-05-29 07:00