• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

FEATURED AMD ZEN Discussion (Previous Rumor Thread)

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.
MM ..... nobody can give you a true answer for your questions right now ....... at least till people get their hands on one to see what is going on under the hood. Once the NDR is lifted by AMD maybe there will be some answers kicking around.

I agree on my questions being to early :)

Dolk made a commit earlier in the thread. He said that a 16 phase VRM for the CPU was marketing and was just a weaker 8 phase X2 VRM??? If this is true, a 4 phase VRM for the memory is just a weaker 2 phase X2 VRM. Correct???

- - - Updated - - -

:rofl::rofl::rofl::rofl:

They may get this right, but dear lord man.... don't say that again like it means anything!! :p

Didn't you KNOW..........If it's on WCC, It's been written in stone :)
 
Dolk made a commit earlier in the thread. He said that a 16 phase VRM for the CPU was marketing and was just a weaker 8 phase X2 VRM??? If this is true, a 4 phase VRM for the memory is just a weaker 2 phase X2 VRM. Correct???
No. That particular motherboard happens to use doublers. Not all do. I am sure Dolk can explain in details way above your (mine/most of us) head though. :)
 
No. That particular motherboard happens to use doublers. Not all do. I am sure Dolk can explain in details way above your (mine/most of us) head though. :)

What level would you like? :D

Wccftech has been surprisingly on point for Ryzen. Not always there but to a higher degree than some of the older rumor sites. Now, if the reviews guys can comment than we will know if its true or not.

Now for another question, which RAM to grab before this chip comes out. I was thinking of playing it safe until more Memory guys get ahold of the CPU. Get like a 2400MHz DDR4 sticks just to get me through the initial release.
 
What level would you like? :D

Wccftech has been surprisingly on point for Ryzen. Not always there but to a higher degree than some of the older rumor sites. Now, if the reviews guys can comment than we will know if its true or not.

Now for another question, which RAM to grab before this chip comes out. I was thinking of playing it safe until more Memory guys get ahold of the CPU. Get like a 2400MHz DDR4 sticks just to get me through the initial release.

I don't think you'll have to play it that safe Dolk. I would go with the usual recommendation 3200 CL14 or 15
I plan on seeing if it'll boot with some 4266 sticks I have here
 
I havent seen an "official" statement mentioning the max RAM speed just yet. Is 3200 right now the "hypothetical tolerated" max?
 
I heard from good sources the mobo manus were testing well over 4000
 
Have a look here at some of the "phasing" scheme http://sinhardware.com/index.php/vrm-articles/82-vrm-guide

Excellent reading :thup: I would have bought a 16+4 motherboard on the 24-25th and passed on anything less.

- - - Updated - - -

I havent seen an "official" statement mentioning the max RAM speed just yet. Is 3200 right now the "hypothetical tolerated" max?

That was a question that I posted earlier. I was afraid that my 3200 memory would limit me in some of the benchmarks.
 
I agree on my questions being to early :)

Dolk made a commit earlier in the thread. He said that a 16 phase VRM for the CPU was marketing and was just a weaker 8 phase X2 VRM??? If this is true, a 4 phase VRM for the memory is just a weaker 2 phase X2 VRM. Correct???

I don't think they make a VRM Driver past 8 phase so they use Doubler for 16 phase and really it's only 8 phase that delivers more amps with doubling, however has the same clean power as 8 phase . They can use a 8 or 6 phase driver for 4 phase just by not using all the driver connections.:cool:
 
Question : a heavily overclocked 8350 even on LN2 only needs around 8-10 phases (world records were made in a Crosshair Z if memory serves), why would they make 16+ for Ryzen that supposedly uses less power ? What's the point, higher OC potencial or redundancy ?
 
Not all phases are created equal. It's also better, in many cases, to spread that load across more phases for more efficient use/cleaner power/lower temps...


....and I'll leave the details to Dolk again. :)

- - - Updated - - -

I hope you're right......but I don't think so, at least not right away.
well, if it's only extreme scaling... that isn't too bad for it. If these don't break 4.5ghz on water... :(
 
Mr. Scott, get with me for some hackery if you can bring me on board with Zen (Dell isn't letting me have fun with those chips yet :/ )

The number of phases do not correlate to the power plane strength. In fact I can say that even with high end Skylynx server chips (the big nasty 200w+ ones) they have 4 phases on servers. This is pretty common for server because they don't like noisy power planes, and the fewer phases the better. The key to a great VRM power plane is in the architectural design. High end FETs but cut cost on inductors: gonna need more phases. Great inductors but meh FETs, gonna need good drivers to keep FETs happy. There are so many different parts to the VRM design that I could go on a week long describing it all.

If people would like, I could make a quick blurb about power planes and how they all work, but there is a lot of material out there on this. For the basic concept look up Buck MOSFET Power Systems
 
Mr. Scott, get with me for some hackery if you can bring me on board with Zen (Dell isn't letting me have fun with those chips yet :/ )
Lol. I was just in the right place at the right time......as an observer.
 
Mr. Scott, get with me for some hackery if you can bring me on board with Zen (Dell isn't letting me have fun with those chips yet :/ )

The number of phases do not correlate to the power plane strength. In fact I can say that even with high end Skylynx server chips (the big nasty 200w+ ones) they have 4 phases on servers. This is pretty common for server because they don't like noisy power planes, and the fewer phases the better. The key to a great VRM power plane is in the architectural design. High end FETs but cut cost on inductors: gonna need more phases. Great inductors but meh FETs, gonna need good drivers to keep FETs happy. There are so many different parts to the VRM design that I could go on a week long describing it all.

If people would like, I could make a quick blurb about power planes and how they all work, but there is a lot of material out there on this. For the basic concept look up Buck MOSFET Power Systems

Then why do people say more phases provide clean power?
 
Then why do people say more phases provide clean power?

Is a V12 engine always more powerful than a V8 engine?

Marketing is at play here. BlueFalcon has it correct. It is not always completely true that motherboard manufactures use cheaper parts to create power planes. The high ends at ASUS, Gigabyte and the rest can contend in price of a server power plane. But now that the crowed believes that more = better, why move from it? Furthermore, the technique of VRM hasn't changed, its just been tweaked to the nth degree as of late. We have smarter and faster chips due to lower feature size (transistor size). We are getting better performance per a phase with each generation. Early 2000s, we were still mastering all the things we have today in advanced home computers. So quality wasn't as good, so more did in fact help. But now, its all marketing and expectations.

24 shades phase of BS
 
Back