• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

*** AMD "Zen" thread (inc AM4/APU discussion) ***

More details about AM4:

http://www.bitsandchips.it/english/...cket-will-be-uopga-and-it-will-have-1331-pins

Future AMD AM4 Socket will be µOPGA, and it will have 1331 pins. A LGA solution will be used for Opteron Zen CPUs.







AM4 solution will have the virtue of putting together the plus points of the three actual AMD sockets: AM3+, FM2+ and AM1.

Like AM3+ Socket, AM4 will be able to support 140+ Watt CPUs and APUs. Like FM2+, AM4 will be able to support powerful integrated GPUs. Like AM1, AM4 will allow to realize budget mainboards, as much as high end mainboards: AM4 CPUs and APUs will have an integrated FCH, but a lot of features can be added thanks to external chipsets (more PCI-E lines for multi GPUs cofigurations, M.2 slots, etc).
Click to expand...
 
Meh, the reality is that due to the way products ship and installation is done, you have a significantly higher chance of damaging pins on the cpu than protected within the socket.

It might be slightly more inconvenient to remove a motherboard but in general 10 mins extra to remove a motherboard compared to the well, lets say minimum of 2 day turnaround for an RMA(though usually much more) is barely noticeable.

It might be that during installation you have a lets say 3% chance to damage the pins on a cpu but 0.3% chance to damage the pins within the socket on a motherboard. Drop a CPU, easy to bend pins, drop a motherboard, almost no chance to damage the pins. It's also easier to drop a cpu than drop a motherboard as it's smaller and more fiddly to deal with. So taking 10 mins more to remove a motherboard for an RMA is inconsequential when the risk of damaging the pins is at least a magnitude smaller than for the CPU.

This is also why LGA remains pin density superior, thinner less strong pins are more likely to be damaged, the same size pins on the cpu would get damaged so often it's not viable where as LGA side with decreased damage risk, the thinner pins make a much smaller difference. You've also got the type of installation. There is very little way to push the pins sideways within the socket, you merely lay the chip down on top. With a cpu as you insert the pins any lateral movement on the chip can bend all the pins in one go.

LGA is like a eleventy bajillion times more sensible than PGA.
 
drunkenmaster said:
LGA is like a eleventy bajillion times more sensible than PGA.
Click to expand...

We'll sensibly agree to disagree :)

Having fitted dozens of one and a dozen or so of the other.

I do have a personal preference for PGA. Practically speaking you would be careful of fitting the CPU anyway and I for one would remove the motherboard to take out the CPU anyway. So taking the mobo out is a given either way.

The only downside for me is if you are unlucky enough to put paste onto the pins from your fingers, it is a ****** to clean off.
 
It says to expect them soon, not that they're already "in the wild".

Here's what actually happened:

AIDA64 put out a new version with preliminary support for Zen. They didn't say what this entails.

Waste Crap Crud Flim-flam
 
http://www.bitsandchips.it/english/...ks-the-next-amd-uarch-is-a-major-step-forward

Interesting times, their source stating that zen should match broadwell-E. Now AMD just has to bring out all parts with SMT enabled and it will force Intel to rethink their entire processor lineup. Since if they gave the i5's SMT, how would they distinguish them from the i7's?

And the AM4 mounting hole changes make sense, they will more than likely make a slightly larger substrate so they can accomodate APU's with HBM.
 
http://www.overclockers.ru/lab/75317/32-norma-pervoe-znakomstvo-s-amd-zen.html

First Zen 1.7GHz 16C/32T server CPU engineering sample tested compared to 5930K 6C/12T. Zen was faster on applications and slower on gaming but games like GTA V, Hitman and World of Tanks are disaster for Zen.

I checked GTA V and Hitman similar numbers on gamegpu site showed Zen numbers sit at 8350 and 2500K Sandy Bridge level.

If Summit Ridge desktop CPU final revision hit retail still show poor gaming performance then I will get Kaby Lake CPU.

Edit: A day later on 2 April, site updated the first page added the line below Introduction "April Fools Day stuff".

:(
 
Last edited:
I think it looks as though that engineering sample is locked at 1.7ghz, the other article I posted they said their samples were frequency locked and Boost was disabled. So it is showing it has far better ipc if it is showing similar cpu results to an 8350 when this part is clocked at 1.7ghz.
 
Last edited:
AthlonXP1800 said:
http://www.overclockers.ru/lab/75317/32-norma-pervoe-znakomstvo-s-amd-zen.html

First Zen 1.7GHz 16C/32T server CPU engineering sample tested compared to 5930K 6C/12T. Zen was faster on applications and slower on gaming but games like GTA V, Hitman and World of Tanks are disaster for Zen.

I checked GTA V and Hitman similar numbers on gamegpu site showed Zen numbers sit at 8350 and 2500K Sandy Bridge level.

If Summit Ridge desktop CPU final revision hit retail still show poor gaming performance then I will get Kaby Lake CPU.
Click to expand...

Mauller said:
I think it looks as though that engineering sample is locked at 1.7ghz, the other article I posted they said their samples were frequency locked and Boost was disabled. So it is showing it has far better ipc if it is showing similar cpu results to an 8350 when this part is clocked at 1.7ghz.
Click to expand...


All the slides have 1700 МГц written next to them, it translates as 1700Mhz

In games at best its 70% the performance of the 5930K. at worst 35%

The 5930K Boosts to 3.7Ghz, so with WOWS being DX9 its very single thread dependant, if we take 1.7Ghz and add 118% we get 3.7Ghz, so add 118% to 21 FPS results in 46 FPS, add 40% to that you get 64, which is the FPS the 5930K has, so that would mean this Zen is about 60% the IPC of Intel, which is about the same as Vishera.

Some what odd if Zen ends up with about the same IPC as an FX-8350, i don't trust it, but if its true AMD might as well not bother.
 
You must log in or register to reply here.
Back
Top Bottom