• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

AMD Zen - Key Dates and Information

Page 17 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
On a related note, do we know if the Ryzen turbo is all core or some fraction?
There is both, and something inbetween, likely also depending on the amount of active cores, their actual power consumption and temperatures, application behaviour (e.g. if turbo doesn't make an app faster, the CPU might use the power elsewhere), etc.
 
So it will be a very close contest. Hopefully AMD is sensible enough to price their SR CPUs from USD 200 - USD 500. If AMD can get a 3.4 Ghz base clock 8C/16T with the Wraith cooler at USD 400 and a 3.6 Ghz base clock 8C/16T with an AIO cooler at USD 500 they will do very well.

As posted elsewhere they are still in the decision process when it comes to prices, but that should be about their targeted range, although starting from quite lower than 200$ if we are to interpret the 4C/4T sampled chips, at 120$ such a SKU would still provide 50% gross margin, wich is huge for a fabless company..

At CES, the company openly asked journalists what it should be priced at. Price an eight-core Ryzen at around £400/$500 and you bring down the cost of eight-core chips to that of a six-core—a solid move for consumers, but not one that will greatly affect mainstream performance. Price it the same as a quad-core i7-7700K—about £300/$330—and you dramatically shake up the industry. The decision is yet to be made.

"There are a lot of discussions going on," says Hallock. "We're capturing the feedback. We wanna take share, we wanna be the best price/performance option, we wanna be the first on people's minds. That's part of the bounding box for pricing discussions as well as paying off the R&D investment... We're looking at what Intel does—and we're not gonna do that. We think people want the choice, and need the choice. The market needs the choice—hopefully we can turn it around.


http://arstechnica.com/gadgets/2017/01/amd-ryzen-motherboards-hype/
 
There is both, and something inbetween, likely also depending on the amount of active cores, their actual power consumption and temperatures, application behaviour (e.g. if turbo doesn't make an app faster, the CPU might use the power elsewhere), etc.

I should have clarified, I meant to ask if the 4.0 boost is all core or something less.
 
As posted elsewhere they are still in the decision process when it comes to prices, but that should be about their targeted range, although starting from quite lower than 200$ if we are to interpret the 4C/4T sampled chips, at 120$ such a SKU would still provide 50% gross margin, wich is huge for a fabless company..

$100-120 doesn't sound unreasonable for for a 4C/4T Ryzen. The 880K (2M/4T@4-4.2GHz) retails for ~$100 after all, the 845 is only ~$70 but you do sacrifice some frequency.
 
4Ghz , at least on the 3.6/4Ghz ES, is reserved only for one thread workloads. All core boost and half core boost should be somewhere in the middle, likely 3.7-3.8Ghz range (my guess). Ryzen can adjust Turbo in 25Mhz intervals so it is very precise.
 
I should have clarified, I meant to ask if the 4.0 boost is all core or something less.

Obviously we don't know for sure yet.

In general one would assume max turbo is not for all cores - if they could guarantee all cores at that speed (within temp/power) then it would be the base clock.

You could simplify it and think max turbo is for one core and base is for all cores being used... But this is much too general also: It all depends on what instructions are being run. So who knows? Perhaps all cores could boost to max under certain situations.
 
$100-120 doesn't sound unreasonable for for a 4C/4T Ryzen. The 880K (2M/4T@4-4.2GHz) retails for ~$100 after all, the 845 is only ~$70 but you do sacrifice some frequency.

I agree. 8T Ryzen should be i5 7600K price range and 4T version should be ~20-30% cheaper, depending on relative performance Vs i5 (base/turbo clock is likely going to be lower and not by a small amount).
 
Might be, but:
1. Every Ryzen can be overclocked, even on cheap(er) B350 boards
2. Stock performance will be more than enough, unless you plan to use GTX 1080 in SLI, or make simulation of spaceship entering Earth's atmosphere
 
I should have clarified, I meant to ask if the 4.0 boost is all core or something less.
No problem. As said by inf64, this is for 1C. This boost limit might either be removed for final SKUs with XFR or be a guaranteed lower limit with just 1C.
 
No problem. As said by inf64, this is for 1C. This boost limit might either be removed for final SKUs with XFR or be a guaranteed lower limit with just 1C.
I don't think we need to be concerned at all with turbo mechanics: between XFR and unlocked multi on every Zen SKU, the only limits a DIY customer will face in configuring their Zen setup will be cooling capacity, MB power delivery, silicon lottery for the chip itself. The turbo stuff will likely affect OEM builds.
 
First, we need to figure out which core configurations AMD will be making. Does anyone even have a sample of a 4 or 6 core? And if each processor can overclock by itself, what's the point of having more than one model with the same core count?
 
First, we need to figure out which core configurations AMD will be making. Does anyone even have a sample of a 4 or 6 core? And if each processor can overclock by itself, what's the point of having more than one model with the same core count?

A couple of reasons:
  • Auto overclocking may not be enabled by default intentionally
  • Binning based on auto overclocking potential
  • Segmentation for OEMs who turn the auto-OC off
 
Only a VERY small portion of the population overclocks, thats why.

But isn't AMD releasing these specifically for this type of market and later releasing professional and consumer variants? It just seems strange to offer a CPU at 3.4Ghz and 3.6Ghz when both will likely hit the same speed.
 
It just seems strange to offer a CPU at 3.4Ghz and 3.6Ghz when both will likely hit the same speed.
Will they? 🙂

Here's what siliconlottery.com can tell us about a newly released CPU
74% of tested 7700Ks were able to hit 4.9GHz
56% of tested 7700Ks were able to hit 5.0GHz
26% of tested 7700Ks were able to hit 5.1GHz
5% of tested 7700Ks were able to hit 5.2GHz

If you choose the better performers as 3.6Ghz SKU and the weaker parts as 3.4Ghz SKU, even with some overlapping in overclocking potential the chances still heavily favor 3.6Ghz SKU to clock higher.

PS: imagine one plays Russian roulette with either 1 bullet or 2 bullets in the cylinder. Even if one survives both "games", chances of survival were quite different to begin with. (yes, this was not a car analogy 🙂)
 
Last edited:
Will they? 🙂

Here's what siliconlottery.com can tell us about a newly released CPU


If you choose the better performers as 3.6Ghz SKU and the weaker parts as 3.4Ghz SKU, even with some overlapping in overclocking potential the chances still heavily favor 3.6Ghz SKU to clock higher.

PS: imagine one plays Russian roulette with either 1 bullet or 2 bullets in the cylinder. Even if one survives both "games", chances of survival were quite different to begin with. (yes, this was not a car analogy 🙂)

yeah, but this is different with Zen and the way it auto-OC's right? Point being that a "3.4ghz" Zen should easily hit 3.6 on its own, so what's the point of a 3.4ghz? I guess just more silicon you can sell at a lower price?
 
Point being that a "3.4ghz" Zen should easily hit 3.6 on its own, so what's the point of a 3.4ghz?
On average, at the power required for the lower SKU to hit 3.6Ghz oc, the higher SKU will hit 3.75GHz+ oc. Chip quality is still a deciding factor on what you can achieve in terms of speed even without the multi lock.
 
If AMD wants to sell 500k units for 6 and 8 cores , they should price 8 cores at 600-700$. Intel can easily price 10 cores at 999$ with 8 and 6 cores just 50$ above AMD to just end all the AMD hype.

Why so much? I think they should top out at ~$500. Intel won't want to bring their entire HEDT line (sans 6950x) to $550 or lower. That will kill their margins.

High pricing=tiny market, that's a fact not a supposition.

AMD's current market potential is already tiny for reasons having nothing to do with pricing. Well, maybe not tiny, but still rather small.

On average, at the power required for the lower SKU to hit 3.6Ghz oc, the higher SKU will hit 3.75GHz+ oc. Chip quality is still a deciding factor on what you can achieve in terms of speed even without the multi lock.

Cooling is also a factor. You can push lower bins pretty high, but it'll take more voltage (and more aggressive cooling) to get there.
 
yeah, but this is different with Zen and the way it auto-OC's right? Point being that a "3.4ghz" Zen should easily hit 3.6 on its own, so what's the point of a 3.4ghz? I guess just more silicon you can sell at a lower price?

Will the auto-OC also work on board that don't support overclocking? 😉
 
For whatever reason, even if this has been posted multiple times recently, I only got to find out about it because of your post, so there is no need for you to feel sheepish for posting it.

Mee too, ive been watching like a hawk but also missed this untill that post.
 
Back
Top