AMD "Zembezi

Page 12 - Seeking answers? Join the Tom's Hardware community: where nearly two million members share solutions and discuss the latest tech.
Status
Not open for further replies.

illfindu

Distinguished
Nov 30, 2009
370
0
18,810
Hey im looking towards the future and I'm eyeing the AMD 8 core bulldozer CPU'S coming down the line. I'v seen some source say there going to use a AM3+ Socket and im wondering if that means youll be able to toss one in a current AM3+ compatible board?
Will my current http://www.newegg.com/Product/Product.aspx?Item=N82E16813130297&nm_mc=OTC-Froogle&cm_mmc=OTC-Froogle-_-Motherboards+-+AMD-_-MSI-_-13130297 msi 870A Fuzion work i noticed just now that its a AM3 not a AM3+ I'm guessing that means ill need a new mother board cause the sockets are comparable?
 
Solution


That "market" is no different from any other market. Not all software makes the best use of 24 cores - some of the tests in that link didn't even make use of 12 cores. How is a lower clocked 24 core server supposed to perform against a higher clocked 12 core server when the workloads are only optimised for 12 cores?

22156.png


The result of this scaling is that for once, you can notice which CPUs have real cores vs. ones that have virtual (Hyper...

Rupees(Indian)45=USD1

And it's available for

Intel Core i7 2600K Sandy Bridge 3.4GHz

Price:16,800.00
 




http://semiaccurate.com/2011/02/03/intels-romley-platform-will-be-available-lga-1356-and-lga-2011/

It doesn't specify. But basically all LGA 2011 CPUs will have support for 3 DIMMs per channel meaning a total of 12 slots. That means if you push 4GB sticks its 48GB. If you push 8GB its 96GB. Or the 16GB sticks (probably not out to the public yet......) thats 192GB.

Of course LGA 2011 is set for entry level workstations.
 
20th June. Way to be slow.

However when it all sorts itself out, an AMD mobo and CPU option may be much cheaper in the long run that a bonkers Intel setup that is unlikely to offer value beyong one CPU.
 
As a result, AMD will be able to create a "halo" effect for its microprocessor lineup with a high-end offering that is projected to outperform existing Phenom II chips by 50%.

Ummm...wasn't it supposed to be "outperform Sandy Bridge by 50%"? Anyone else catch that? And I'm assuming thats the "best case" scenereo, and not general usage.

The more I hear, the more skeptical I am of BD.
 



I am going to guess he is either not a liberty to disclose anything, or he had no discrete knowledge about the "halo" effect thereof.
 
I'm getting so tired off waiting for Bulldozer... I have $800 +/- to spend on an upgrade and would very much like it be my first AMD system but IF its crap I'm going to be very very annoyed i didnt just hop on the SB wagon... All this waiting is doing my head in!
 

If AMD dares to postpone Bulldozer again im going all out i7 2600k and im going to overclock the *** out of it... 31 May is BUY UPGRADE day... Bulldozer or no Bulldozer....
 

Counting Hyperthreading it would be ~10-15% faster than Ph2.

I think the 50% is only based on an 8 core Bulldozer vs the 6 in Thuban. This is most likely not accurate. If Bulldozer is only 50% faster than Ph 2 in this way, then AMD is going under; they will not compete with an 8 core Sandy in any way possible performance wise.

I don't get how throughput and processing speed are two different things(basically where that 50% argument came from); can anyone explain that? I'm only asking, since JFAMD says they can't be compared.

For hard drives, throughput is how many I/Os it does * the size of the file, so why wouldn't that be somewhat like a processor? The amount of instructions processed in a certain amount of time seems like it would be just like throughput, so the I/O=IPC and then there is how big the Instructions are(64 bit).
 


SB is 10-15% over Nehalem. Nehalem is about 10-20% clock per clock over PHII.

Plus you are right. Its 8 core vs 6 core. In their server comparison is was 16 core vs 12 core.

So 6 vs 8 that means already 33% more cores and in synthetics that would give at least 33% more, not to transfer to real world. Maybe in server apps it would but in real world it wont. That could mean 17% core per core and clock per clock over PHII. But who knows. AMD wont tells us a thing.

As for throughput, its like Radeons. A HD5870 pushed out 2TFLOPS of performance. And the HD6990 is 5.1TFLOPS. Well over nVidia. But it never comes out the same in real world performance. So the 50% throughput might refer to how many GFLOPS the CPU can do vs a 6 core PHII.
 
Does anyone here remember that bizarre rumor recently that a few AM3 motherboards could support the upcoming AMD Bulldozer CPU's like the FX -8000's,6000's,4000's etc?

Just tonight I have a friend adamant about building a Phenom II X6 1100 T system.
I have told him to wait for AMD's new Bulldozer CPU's but no he doesn't want to wait for 3 months.
I even advised him to build a Intel Core i7 2600K system before too (certainly significantly more potent than the Phenom II X6 1100T).

Well anyway I was checking some (AM3) motherboards for him and this one caught my eye first because I thought that it was a nice motherboard with good newer features (SATA III and USB 3.0) and requiring no BIOS flashing for his desired Phenom II X6 1100T CPU. Then I just saw something interesting

----Look at the AM3+ CPU ready section on this page-----

http://www.gigabyte.com/products/product-page.aspx?pid=3781#ov

"Get Ready for New Generation of 32nm AMD AM3+ CPU
GIGABYTE's AMD motherboards are ready to support new generation of 32nm AMD AM3+ multi-core processors, delivering the very best platform for multitasking, multimedia and high performance gaming."

Is this deceptive or are they (Gigabyte) advertising that this particular AM3 motherboard could possibly run the upcoming AM3+ FX-8000,6000,4000 (Bulldozer) CPU's with perhaps some kind of performance loss.

Any speculation or opinions about this?
 


Its a AM3+ Mobo. Hence the backwards compatibility to AM3 CPU's.
http://www.gigabyte.com/products/product-page.aspx?pid=3781#sp
 


Ah I see now.It depends upon the motherboard revision (version 3.1 is socket AM3+).Thanks for the clarification.
I wonder if newegg already carry's this or is selling the earlier revision (3.0,2.1,2.0).

http://www.newegg.com/Product/Product.aspx?Item=N82E16813128441

I guess someone would have to contact newegg to see specifically which revision they were currently selling as the one in the photograph on their page is revision 2.0 (looking at the motherboard close up (zoomed in picture).
 
Status
Not open for further replies.