PSA: AMD XFR Enabled On All Ryzen CPUs, X SKUs Have Wider Range

Subject: Processors | March 4, 2017 - 06:00 AM |
Tagged: xfr, turbo, sensemi, ryzen, overclocking, amd

Following the leaks and official news and reviews of AMD's Ryzen processors there were a few readers asking for clarity on the eXtended Frequency Range (XFR) technology and whether or not it is enabled on all Ryzen CPUs or only the X models. After quite a bit of digging through forums and contradictory articles, I believe I have the facts in hand to answer those questions. In short, XFR is supported on all Ryzen processors (at least all the Ryzen 7 CPUs released so far) including the non-X Ryzen 7 1700; however the X SKUs get a bigger boost from XFR than the non-X model(s).

Specifically, the Ryzen 7 1700X and Ryzen 7 1800X when paired with a high end air or water cooler is able to boost up to an additional 100 MHz over the 4 GHz advertised boost clock while the Ryzen 7 1700 is limited to an XFR boost of up to 50 MHz so long as there is thermal headroom. Interestingly, the Extended Frequency Range boosts are done in 25 MHz increments (and likely achieved by adjusting the multiplier by 0.25x).

View Full Size

How does this all work though? Well, with Ryzen AMD introduced a new suite of technologies it calls "SenseMI" which, despite the questionable name (heh), puts a lot of intelligence into the processor and builds on the paths AMD started down with Carrizo and Excavator designs. The five main technologies are Pure Power, Precision Boost, Extended Frequency Range (XFR), Neural Net Prediction, and Smart Prefetch. The first three are important when talking about XFR.

With Ryzen AMD has embedded a number of sensors throughout the chip that accurately measure temperatures, clock speeds, and voltages within 1°C, 1mA, 1mW, 1mV and it has connected all the sensors together using its Infinity Fabric. Pure Power lets AMD make localized adaptive adjustments to optimize power usage without negatively affecting performance. Precision Boost is AMD's equivalent to Intel's Turbo Boost and it is built on top of Pure Power's sensor network. Precision Boost enables a Ryzen CPU to dynamically clock up beyond the base clock speed across all cores or clock even further across two cores. Lightly threaded workloads will benefit from the latter while workloads using any more than two threads will get the all core boost, so there is not a lot of granularity in number of cores vs allowed boost but there does not really need to be and the Precision Boost is more granular than Intel's Turbo Boost in clock speed bumps of 25MHz increments versus 100 MHz increments up to the maximum allowed Precision Boost clock. As an example, the Ryzen 7 1800X has a base clock of 3.6 GHz and so long as there is thermal headroom it can adjust the clock speed up by 25 MHz steps to 3.7 GHz across all eight cores or up to as much as 4.0 GHz on two cores.

From there XFR allows the processor to clock beyond the 2 core Precision Boost (XFR only works to increase the boost of the two core turbo not the all core turbo) and as temperatures decrease the allowed XFR increases. While initial reports and slides from AMD suggested XFR would scale with the cooler (air, water, LN2, LHe) with no upper limit aside from temperature and other sensor input, it appears AMD has taken a step back and limited X series Ryzen 7 chips to a maximum XFR boost of 100 MHz over the two core Precision Boost and non-X series Ryzen 7 processors to a maximum XFR boost of 50 MHz over the maximum boosted two core clock speed. The Ryzen 7 1700 will have two extra steps above its two core boost so while the chip has a base clock of 3.0 GHz, Precision Boost can take all eight cores to 3.1GHz or two cores to 3.7 GHz. Further, so long as temperatures are still in check XFR can take those two boosted cores to 3.75 GHz.

View Full Size

XFR will be a setting that you are able to toggle on and off via a motherboard setting, and some motherboards may have the feature turned on by default. Unfortunately, if you choose to manually overclock you will lose XFR functionality (and boost). Further, Precision Boost and XFR are connected and you are not able to turn off one but not the other (you either get both or nothing). Note that if you overclock using AMD's "Ryzen Master" software utility, it will also disable Precision Boost and XFR, but the lower power C-states will stay enabled which may be desirable if you want the power bill and room to cool down when not gaming or creating content.

I would expect as yields and the binning processes improve for Ryzen AMD may lift or extend the XFR limits either with a product refresh (not sure if a micro-code update would be possible) or maybe only in the upcoming hexa-core and quad core Ryzen 5 and Ryzen 3 processors that have less cores and more headroom for overclocking. That is merely speculation however. Ryzen 5 and Ryzen 3 should support XFR on both X and non-X models, but it is too early to know or say what the XFR boost will be.

XFR is neat though not as big of a deal as I originally thought it to be without limits, and as many expected manual overclocking is still going to be the way to go. This is not all bad news though, because it means that the much cheaper Ryzen 7 1700 just got a lot more attractive. You give up a 50 MHz XFR boost that you can't use anyway because you are going to manually overclock and you gamble a bit on getting a decently binned chip that can hit R7 1800X clock speeds, but you save $170 that you can put towards a better motherboard or a better graphics card (or a second one for CrossFire - even on B350).

I am still waiting on our overclocking results as well as Kyle's overclocking results when it comes to the Ryzen 7 1700, but several other sites are reporting success at hitting at least 4.0 GHz (though not many results over 4.0 or 4.1 GHz which isn't unexpected since these are not the highest binned chips and yields are still young so bins are more real/based on silicon and not just for product segmentation but most can hit the higher speeds at x power, v voltage, and n temperature et al). For example, Legit Reviews reports that they were able to hit manually overclock a R7 1700 to 4.0 GHz on all cores at 1.3875 volts. They were able to keep the non-X Ryzen chip stable with those settings on both aftermarket air and AIO water coolers.

View Full Size

AMD's Ryzen Master overclocking software lets you OC and setup CPU and memory profiles from your OS.

More on overclocking: Tom's Hardware has posted that, according to AMD, the safe voltage ceiling for overclocking is 1.35V if you want the CPU to last, but that up to 1.45V CPU voltage is "sustainable". Further, note that is is recommended not to set the SOC Voltage higher than 1.2 volts. Also, much like Intel's platform, it is possible to adjust the base clock (BCLCK) but you may run into stability problems with the rest of the system if you push this too far outside expected specifications (PC Gamer claims you can set this up to 140 MHz though so AM4/Ryzen may be more forgiving in this area than Intel. Edit: The highest figure I've seen so far is 106.4 MHz being stable before the rest of the system gets too far out of spec and becomes unstable. The main benefit to adjusting this is to support overclocked RAM above 3200 MHz so unless you need that your overclocking efforts are probably better spent adjusting the multiplier. /edit). Finally, when manually overclocking you will be able to turn off SMT and/or turn off cores in 2s (e.g. disable 2 cores or disable 4 cores, you can't disable in single numbers but groups of two).

Hopefully this helps to clear up the XFR confusion. If you do not need guaranteed clocks with a bonus XFR boost for a stable workstation build, saving money and going with the Ryzen 7 1700 and manually overclocking it to at least attempt to reach R7 1700X or 1800X speeds seems like the way to go for enthusiasts that are considering making the jump to AM4 especially if you enjoy tinkering with things like overclocking. There's nothing wrong with going with the higher priced and binned chips if you want to go that route, but don't do it for XFR in my opinion.

What are your thoughts? Are you planning to overclock your Ryzen CPU or do you think the Precision Boost and XFR is enough?

AMD on Extended Frequency Range:

Source: Ars Technica

March 4, 2017 | 06:13 AM - Posted by Master Chen (not verified)

And in other (actually RELEVANT) news - 1800X was just OverClocked to 5.8GHz on all cores & threads.

March 4, 2017 | 06:39 AM - Posted by SheepInACart

The 5.8ghz was a confirmed speed that took the record on HWbot, but it was super flakey, it didn't complete a cinebench successfully run even under LN2 due to its 1.97v insanity.

Interesting to note though, the 1800x (but a different chip+person) did take the record for cinebench back (now set at 2454) while running at 5.36ghz (also on LN2). But just barely, it was was only 9 points ahead of the previous record holder, an i7 5960X overclocked to 6GHz which had reclaimed the record from the Ryzen chips , after it was set at 2449 by a 5.2ghz 1800x on march 2.

So is the 1800x awesome? Heck yes, but interestingly the 5960x is still fighting tooth and claw for those records despite having half as many threads and being 2.5years old now.

March 4, 2017 | 06:46 AM - Posted by Anonymous (not verified)

Half as many threads? It has exactly the same number of threads as Ryzen, 8c/16t for the 5960x FYI.

March 4, 2017 | 06:50 AM - Posted by Master Chen (not verified)

>6GHz
>5.3GHz
>5.3GHz gets better scores
>MUH INTEL STILL STRONK
Math. Learn 'em.

March 4, 2017 | 08:47 AM - Posted by SheepInACart

Not a fanboy at all, for the money the Ryzen is killer, and it IS currently winning, I just think its interesting that the guys still pushing the 5960x where able to up their game and catch back up for a while.

As for the "maths" part, both chip[s are pushed to the limit, its not like they both can get to the same speed with the same ease, so only the final performance matters. Also because math, bother scores are basically the same, 9 points is literally test variance sort of levels. But I don't doubt with more time though the AMD 1800x will get higher stable records, its not been out long after all.

March 4, 2017 | 10:12 AM - Posted by Tim Verry

Indeed I saw that last night, 1.97 volts is a lot!

March 4, 2017 | 10:45 AM - Posted by Master Chen (not verified)

Try doing that on Intel's 8 core garbage and it'll melt.

March 4, 2017 | 01:00 PM - Posted by Anonymous (not verified)

You couldn't melt it when its under LN2 like the ryzen chip

March 4, 2017 | 01:25 PM - Posted by Master Chen (not verified)

Come on now, play along.

March 4, 2017 | 11:03 AM - Posted by CNote

I got my old 2500k to 5.2 with an Evo 212. Wasn't blender tested but hey....

March 4, 2017 | 11:08 AM - Posted by Master Chen (not verified)

Emphasis on "Sandy"

March 5, 2017 | 12:07 AM - Posted by nobled (not verified)

Yeah, I'm still running a 2500k at 4.8 GHz liquid cooled. I've had it seems as high as 5.4 GHz, but that required higher voltage then I wanted to keep it at for daily usage. The only upgrade I've done on it is replacing my video card with a Nvidia 970.

I haven't seen a big reason to upgrade, but with the 1800X I think I will build a new system keeping the 970 with the 1800X and using it for dev, video transcoding, a bit of gaming, and some server duty. That way I can leave it booted into Linux. And then I'll get an Nvidia 1080 TI for the 2500k to use as a gaming PC with my 4k TV with just Windows on it. Then I won't have to dual boot.

March 4, 2017 | 06:45 AM - Posted by Master Chen (not verified)

And also this http://twitter.com/theoverclocker/status/837760151570403328 happened. I wonder how does it feel, PcPer, to be so heavily exposed for being a bribed Intel slave? Them shekels were truly juicy, eh?

March 4, 2017 | 07:15 AM - Posted by Pholostan

Yes, it does well in synthetic benchmarks like Firestrike. It does worse in actual, real games. Ryzen doesn't run games badly or anything, but an Intel 7700K is almost always better (at stock clocks even). The Physics Score in Firestrike is nice, but the combined score is weird, it is very low. Way behind a stock 7700K and with a Titan X no less.

March 4, 2017 | 07:36 AM - Posted by Master Chen (not verified)

You just got rekt, sonny. See below.

March 4, 2017 | 10:34 AM - Posted by Anonymous (not verified)

Please, the whole review industry is suspect so lay off of this PcPer focus. Your complaints are better said about the whole PC/laptop/mobile/IOT devices ad/review industries. It's been that way from day one with the Internet to this day that is very similar to the first "Great" Radio Days before the FCC as created to fix that and the TV industry afterwords with regards to fair and impartial product reviews.

The Internet is still very unregulated compared to the Radio/TV industry or the Cable TV(TV channel services only) Industry. Everybody and their dog Knows not to trust any reviews from any one review site as there is an intrinsic conflict of interest with the websites that review products getting the ad revenues and review samples(Very dependent on Review samples) from very makers of the products they the review sites are supposed to be impartially reviewing.

Your complaints need to be made with the FTC and state consumer agencies, and any other state and federal agencies concerned with fairness and impartiality in the consumer advertising or product review industries.

Singling out one website that you allege is acting in an unfair matter is not going to do much good when the whole online website technology, review, and Ad industry is still not held to the same high standards as the over the airwaves TV and the classic print media are.

March 4, 2017 | 10:52 AM - Posted by Master Chen (not verified)

It's not specifically about PcPer.
When I say "how does it feel to be exposed?", I imply the majority of tech reviewing outlets out there in general, PcPer just being one of them. But considering that PcPer has pretty ill-known reputation in the hardcore PC enthusiast commune for being a mindbogglingly pro-Intel and pro-Nvidia biased outlet, I just couldn't not add them into those black listings also. PcPer's bias is one of the worst out there, that's why. It's on a level of being outright cringe-worthy, how heavily they schlick for anything-Intel or anything-Nvidia, while heavily downplaying all the time on AMD and anything it does. In such of a particular case, suspecting them of being bribed by the conglomerate monster that is Intel, ESPECIALLY when this company is in state of absolute panic and disarray due to Zen, is really not too farfetched at all.

March 4, 2017 | 11:19 AM - Posted by Anonymous (not verified)

It's the nature of the online PC/related products ad driven and review sample distribution unfairness driven beast! And that can only be solved by the proper amount of regulations to insure fairness and impartiality in the review sample Reviewer selection process and the ad revenue/relationship conflicts of interest transparency process.

Review samples need to be randomly selected from actual retail locations and given to randomly selected reviewers/media outlets to avoid any review sample bias or review sample cheery picking by the makers of the review samples themselves. That whole review sample selection of samples and selection of reviewers in the press needs to be managed by a impartial selection service that is tasked with the random selection of review samples(fomm randomly select retail outlets) and a random lottery for the selection of what press outlets are selected to receive the review sample/s!

Also any ad revenue/other relations that the online press has with the makers of the products that the online press reviews needs to have required transparency statements outlining to the readers if and what are any relationships(Reviewer/website to product maker) that may present any conflict of interest concerns.

March 4, 2017 | 02:43 PM - Posted by Anonymous (not verified)

I don't understand what there is to fuss about.
Ya'll like old ladies crying over how it was better "back in my day".
Fact of the matter, bribery is nothing new.
In the age of Information everybody's got a keyboard. Someone will eventually speak the truth and/or lie.
A 7700k has the better single core performance, yeah maybe. But that's like saying McDonald's is better than a salad because it tastes good. Well it does. Better? Depends who you're asking.
Nothing really changes. The only thing that changes is technology.
And this technology (Zen) is at this point in time the best mass market desktop CPU series in the world. That's not even a question anymore. The only question left to answer,

Is by how much.

March 4, 2017 | 03:18 PM - Posted by Anonymous (not verified)

No it was not better back in the day, Hell the oil business monopolies where just as bad(100+ years ago) until the government and congress created the antitrust laws and broke the oil monopolies up. And Intel and M$/other tech interests should have been broken up many years ago!

The technology industry(PC/related) is on the same level of corruption as the Oil Industry was before there were even antitrust laws on the books. What makes things today more corrupt is that even with 100+ years of antitrust laws on the books the big technology Trusts Intel/Microsoft/Qualcomm/etc are coopting the political process via the lack of proper laws for campaign financing reform and other corporate practices that abuse the fair market system.

So even with the laws already on the books that are supposed to deal with monopoly abuse the corporate/political system is even more corrupt now than ever before. They are not properly enforcing the laws currently on the books for antitrust and there needs to be a new Teddy Roosevelt kind of anti-monopoly populism that is bipartisan/non-xenophobic to fix what is corrupt with the political system with regards to corporate abuses of fair market regulations that are not even being enforced by the government(Politicians too dependent on campaign financing from big monopolistic interests)!

March 5, 2017 | 05:57 PM - Posted by Jeremy Hellstrom

God I love these threads.  The best part is imagining what my life would be like in the reality we are portrayed to live in.

March 6, 2017 | 12:20 AM - Posted by Anonymous (not verified)

Easy for you to say as you don't live in the pyramid scheme of America! But you only need to look at the politicians who get elected south of your border. You are somewhat isolated from all the effects of the constant flipping and flopping that goes on down here in the three rings under the big top with the elephants and the donkeys both beholden to the ringmasters with the deepest pockets. It's really a one party system where the votes that count the most are the paper ballot kind with the images of long gone politicians on them. And both those four legged animals are very happy to feed on the abundant greens filling the trough!

But don't worry too much because the factories will be coming back but it's just that the only jobs created will be in the construction trades short term and for a little while. But those factories are only being brought closer to the markets that they will be serving because the cost of shipping has become a larger expense relative to the cost of any nonexistent two legged labor costs, with the two legged voters made up of DNA/proteins not really needed to make any products. And thus it will be cheaper to locate the factories closest to where the products are needed.

So this time around all that MAGA and wrapped in the flag nationalistic nonsense is only a ruse to get the state and federal factory tax breaks that are supposed to come with a promise of more manufacturing jobs for the two legged voters foolish enough to fall for it all this time around the same as last time around. And this time around even with the metal-plastic-silicon workers busily focusing their silicon brains on making the assembly lines hum on their 24/7 365(+1 extra day leap year) days a year shifts there will also be fellow metal-plastic-silicon workers loading the transportation systems that themselves have similar silicon brains to bring the goods directly to the retail outlets or even directly to some of the very few and very fortunate two legged voters’ front doors.

Oh how this fancy and elaborate pyramid scheme so concentrates those rectangular ballots with images of long gone politicians on them towards the top of that pyramid! Hell they even have an engraved image of that pyramid on back side of those ballots with a big eye on that capstone, as in eye have the most and you have not so much of those ballots with the images of long gone politicians that really are the only thing that matters to the politicians that are not yet long gone.

March 6, 2017 | 02:09 PM - Posted by Jeremy Hellstrom

This is not a political site.  You should head to the forum or another appropriate spot.  We are here for accusations about technology and those who post news or reviews about it.

March 7, 2017 | 12:06 AM - Posted by Anonymous (not verified)

The Monopoly/political part affects the technology part with users not getting their choice of OS or hardware needs met by laptop OEMs that are under the unfair pressure of the abusive CPU/OS/GPU monopolies that make CPUs/OSs/GPUs! PC users are less affected hardware wise but OS wise there still are monopoly concerns. There is also the total lack of regulation of the online news outlets that have that intrinsic review/advertising conflict of interest that was properly regulated and enforced in the traditional over the air Radio and TV industries decades ago.

With people gradually and insidiously loosing control over their very own PC/Laptop hardware via an OS that is forced upon the users at point of sale(Bundled with/pre-installed on the hardware), or by other means with the Laptop OEMs not at all required to provide a complete set of hardware drivers for alternative OSs and it's is getting worse not better

The level of corruption and false and misleading advertising via online news outlets via advertisements thinly disguised as “subjective” articles is endemic with the online news media’s ad driven/review samples needed free of charge dependencies and revenue streams and the lack of any formal regulation, or even proper enforcement of any regulations that may apply.

So yes there are some political issues that can not be avoid even on technology websites as the technology industry that in comparison to the corrupt oil industry of the past gilded age is much worse now than even that past oil industry/Trusts whose monopoly abuses where what necessitated the antitrust laws to begin with. The Antitrust laws that are not at all being enforced today.

But you are the moderator so you can kill the whole comment thread up to the point that you deem appropriate and think about being less flippant with your replies! But some of the comments in the posts and their replies do reflect the truth with which people regard the technology “Press” and the current state of fairness that is not being enforced properly on the WWW/interwebs. Politicians, like the online "Technology" press, are never to be totally trusted and people are free to vote the Politicians out and turn on the ad blockers if they are not happy but the system[ both online review and political election systems ] is totally co-opted by conflicts of interests.

March 7, 2017 | 01:31 PM - Posted by Jeremy Hellstrom

When I am told I am a "bribed Intel slave"; flippant is appropriate.  I am still waiting for the arrival of any of these shekels I am constantly informed I receive.

March 7, 2017 | 05:05 PM - Posted by Anonymous (not verified)

"When I am told I am a "bribed Intel slave"; flippant is appropriate. I am still waiting for the arrival of any of these shekels I am constantly informed I receive.", he typed from his penthouse overlooking Central Park.

March 7, 2017 | 05:14 PM - Posted by Jeremy Hellstrom

The life I lead in your head sounds so neat.

August 13, 2017 | 09:25 AM - Posted by Anonymouse (not verified)

Valid points on all sides, especially so from the portions that didn't dip into fanatical ranting.

The other person should realize that going a bit too far and lashing out at everyone doesn't enhance the acceptance of their whole position and may cause total rejection from persons not so liberal as yourself (you didn't delete the ranting).

Don't spoil valid points by sugar-coating them with spice. Few want a Birthday Cake with pepper frosting. Had you (the other person, Jeremy) left out a few sentences/paragraphs your whole set of beliefs may have been digested by all.

Up to you, present your beliefs as you see fit and see how they are accepted.

March 6, 2017 | 03:31 AM - Posted by Master Chen (not verified)

"Ignorance is bliss" much there, Jeremiah?

March 26, 2017 | 05:03 PM - Posted by seeratlas d'Tyria (not verified)

Anonymous, I hear your sentiment, but surely,
you didn't just type this with a straight face in the age of Trump....

".... online website technology, review, and Ad industry is still not held to the same high standards as the over the airwaves TV and the classic print media are."

Maybe, I am just too old, but imho there is a certain critical mass reached by anyone/ any organization in media before they acquire the resources and public creds necessary to sustain them in 'telling it like it truly IS' without fear of destruction resulting from releasing a public bad review or critique. WAYYY back there used to be something called Byte Magazine which employed a a rather rambunctious reviewer named Jerry Pournelle.
He achieved his chops by telling the simple truth, good or bad (and often by writing nothing as opposed to trashing a submitted product). There were others almost as influential, who were known by many 'insiders' to be remorseless hacks. Funny how i can no longer remember their names :}

I still vividly recall sitting with Pournelle, the then editor of Byte whose name i forget, the head of a multi-billion dollar electronics corp (lot of money in those days)and a very rich (even then) Bill Gates, over drinks in a house in Vegas on the eve of the industry show. Jerry give Gates an earful: completely honest,direct and heartfelt criticism which seemed well considered, and rather graciously recieved by Mr Gates, whom I believe, had paid for the liquor. lol No rancor, just a pleasant cordial evening.

March 4, 2017 | 11:58 PM - Posted by Anonymous (not verified)

I wonder why the Intel screenshot is cropped, oh wait,
I know, it's all LIES, just like when Wccftech claimed the Rx480 would occur to 1.5 GHz on air, yet it didn't.

So you're saying the whole industry is being paid for by Intel ?
Pathetic little fanboy, no wonder the other fanbases don't take Amd seriously.

March 5, 2017 | 03:07 AM - Posted by Master Chen (not verified)

You're not trying hard enough, kiddo.

March 4, 2017 | 06:55 AM - Posted by Pholostan

PCGamer might talk about 140 MHz base clock, but I'm highly sceptical that it will work at all. According to The Stilt:

https://forums.anandtech.com/threads/ryzen-strictly-technical.2500572/

Overclocking the base clock (BCLK) on AM4 platform is possible, however generally not recommended. This is due to its frequency relations with other interfaces, such as the PCIe. Unlike with Intel's more recent CPUs, there is no asynchronous mode (straps / gears) available, which would allow stepping down the PCIe frequency at certain intervals. The PCIe frequency relation is fixed and therefore it increases at the same rate with the BCLK. Gen. 3 operation can generally be sustained up to ~107MHz frequency and higher speeds will usually require forcing the links to either Gen. 2 or to Gen. 1 modes.

Unstable PCIe can cause various issues, such as system crashes, data corruption (M.2 SSDs), graphical artifacts and various kinds of other undefined behavior.

I seriously doubt that PCGamer has actually run any base clock overclocking, they mention it in passing when talking about memory overclockng.

Looking at their settings when overclocking, they only use the multiplier.

Advanced overclocking has a few other options, like BCLCK adjustment, but unless you're really striving for the final 1-2 percent performance, I'd just go with tweaking the multiplier.

March 4, 2017 | 10:14 AM - Posted by Tim Verry

Right I am guessing in practice the BCLCK is much more limited. I think I read up to 106.4 is stable so far before the rest of the system gets out of wack. Could not remember where i read that though when going over notes :-/.

March 4, 2017 | 10:24 AM - Posted by Pholostan

Yes, it sure looks like it :-/

Would be nice if there was some kind of clocks traps etc, but doesn't look like it so far.

March 5, 2017 | 08:33 AM - Posted by Pholostan

Or have I got things wrong and there is a separate reference clock for memory? Looked at videos overclocking with the Crosshair board, and they seem to regularly set BCLK higer with Asus D.O.C.P for memory overcloking in the EUFI. I don't have any Ryzen stuff, so can't verify.

However looking at Paul's video it seems his system crashes with a BCLK to 112.4 MHz.
https://www.youtube.com/watch?v=IKGJshXgOwU

March 5, 2017 | 10:50 AM - Posted by Tim Verry

hmm im not sure since i dont have hands on with any mobos at the moment. Pretty sure I read the DOCP thing is asus' method for translating RAM XMP profiles to amd compatib.e AMP profiles though.  

March 7, 2017 | 03:27 AM - Posted by Pholostan

Some of the top end motherboards seems to have external clock generators. That's probably it :)

March 4, 2017 | 10:52 AM - Posted by Anonymous (not verified)

There are also questions regarding the Zen/Ryzen and the inter-CCX unit Infinity connection fabric topology that AMD still has not discussed in a thorough enough manner to this day. So hopefully that Strictly Technical forum thread and the various forum posters can help with getting that Infinity fabric and Zeppelin platform CCX to CCX unit connection fabric topology IP information properly sussed out.

I suspect that because of the Server nature of the Zen server IP and Ryzen(Desktop Consumer) variants derived from this Zen server IP, that AMD is holding off on any full information on the Infinity connection fabric topology discussions until the server SKUs are properly released to market. That CPU core to CPU core coherency across the CCX units IP may have some issues, software/OS/firmware/otherwise, that still need to be ironed out before some gaming regression issues at lower resolutions can be fixed. There are still too many unknowns this early in the optimization process to come to any full conclusion.

March 4, 2017 | 07:06 AM - Posted by Master Chen (not verified)

Oh look, but there's MORE:
http://www.imagebam.com/image/9d63f5536139632

March 4, 2017 | 07:14 AM - Posted by Pholostan

What is that image supposed to tell me exactly? It is two different places in the game to begin with. What resolution? What image quality settings? Stock clocks? Systems? Etc.

That image doesn't tell me anything, it's just signal noise.

March 4, 2017 | 07:35 AM - Posted by Master Chen (not verified)

It's stock 1700 VS stock (+ Turbo) 7700K after AM4 motherboard BIOS update. More to come soon, don't worry. But please, do cry harder and never stop. Your tears are truly DELISHUUUSSSSS!

March 4, 2017 | 07:51 AM - Posted by PixyMisa

Grow up.

March 4, 2017 | 08:06 AM - Posted by Mike S. (not verified)

Good luck with that. Master Chen picked the truth and doesn't pay any attention to trivial things like evidence. The best we can do is ignore him (or her).

March 4, 2017 | 08:28 AM - Posted by Master Chen (not verified)

Did you even read what you wrote?

March 4, 2017 | 08:28 AM - Posted by Master Chen (not verified)

Nice try, but not cigar.
You need to try way harder than just that, kiddo.

March 4, 2017 | 10:26 AM - Posted by Pholostan

Sure thing. Then it would be easy for other people, say PCPer, to verify those findings. Just a random picture is not worth anything.

March 4, 2017 | 10:59 AM - Posted by Master Chen (not verified)

I've already asked of Elric to do a retest once he makes BIOS updates and all the other things. After what has been happening for the last four days, I wouldn't trust for a PcPer to even boil me a pot of water. So far, Elric has shown to be one of the very few completely unbiased people out there, in regards to all of this. After all, he did buy all of the needed hardware for his own money, not via InHell's paychecks or "presents". I believe that he could do a proper retest just fine, after the so-called "problems" will be ironed out.

March 4, 2017 | 11:47 AM - Posted by zgradt

XFR is a huge disappointment for me. AMD made it sound like it was monitoring a ton of parameters to allow it to automatically overclock to the limit of stability.

Instead, we get a 0.1Ghz bump? Lame.

March 4, 2017 | 12:16 PM - Posted by Tim Verry

Yeah, I'm not sure what happened on that front. We may have to wait for Zen+ "RyzeAgain" processors for an unlimited XFR. :-)

March 4, 2017 | 12:44 PM - Posted by Master Chen (not verified)

Update your BIOSes already, FUD-spreading crybabies.

March 4, 2017 | 12:33 PM - Posted by Anonymous (not verified)

AMD was under a lot of pressure to get Ryzen to market at the promised time and there are legal(SEC)/financial(Stock market) issues with any delays! That whole XFR IP is still needing some tweaks software/Firmware/others before it can be properly judged as a success or a failure. Intel has had issues with its similar new CPU IP also so there are teething problems that come as part of the process for any CPU designer/maker, or for any other types of processors also!

AMD has been in that hanging by a thread mode for so many years and still doing a damn good job of engineering CPUs and GPUs over the past few years with the limited resources that AMD has had for even more years!

Ryzen is a brand new almost clean sheet design as far as the power subsystems are concerned as well as the Zen micro-architecture that Ryzen is based on even more so. Nothing that is as complex of an undertaking as the Zen/Ryzen CPU/any CPU to design/develop is ever just released to market without some issues to be worked on. AMD has done more than was expected this time around with Zen/Ryzen even besting that promised 40% IPC gain with an actual 52% IPC improvement over its Excavator micro-architecture.

Zen/Ryzen is not so bad for a design that AMD bet the whole company on! And AMD has never been so all in as it has been with this Zen/Ryzen project.

March 4, 2017 | 06:20 PM - Posted by zgradt

The marketing promises are just rubbing me the wrong way. Touting overclocking on a chip with barely any headroom, showing off carefully selected gaming demos before launch (one even showed them trouncing a 7700k), and pushing pre-orders before the NDA's are up seem like not the best way to go about it.

Still, I'm totally in on Zen. I'm just waiting on a MB and cooler brackets to get my 1700X up and running. I've been waiting for a reasonably priced 8 core CPU for a while.

March 4, 2017 | 10:22 PM - Posted by Anonymous (not verified)

Marketing is the liar's profession that traces its roots back to the snake oil salesman and the very first fib ever told.

Overclocking was mostly done with the lower cost/lower binned SKUs in an attempt to find that golden lower cost/lower binned SKU that just barely failed the top end binning shakedown cruse. And these Golden lower binned parts could with a little overclocking headroom be as good as a top binned part, with some minor stability issues, if the end user could find the proper mix of overclocking settings to make the part perform like a top/higher binned part at a great cost savings.

Overclocking was never intended to be an industry feature in the first place! It was only because more people started to try overclocking that the marketing folks started to push that unofficial part of the consumer buying decision process onto the gaming/hardware market. Then and now as far as the engineering reasons that CPU parts are binned in the first place is because of whole groups of stability and power/process variability issues inherent in the making of any CPU, and other processor, parts. The higher binned parts have to pass all of the tests to make the grade while the lower binned part may only have to fail one of the many binning tests to be binned lower.

Overclocking is still not officially supported and there have been only some engineering/approved official methods of clock boosting that has come into the industry via scientifically approved methods of binning parts that can operate over a greater clock frequency range and still meet all the other binning tests for thermal/stability and CPU part durability metrics like MTBF and other measures needed to meet advertised and implied warranty periods.

If Intel’s parts are more overclockable while still meeting their other binning metrics then why is Intel selling these parts that may be able to be higher clocked out of the box officially. Is Intel doing some form of intentional under specifying of their binning regimens in order to foster a mind share in their customer base that Intel’s parts on average have more overclocking headroom when in fact it may be that Intel may actually be engineering/binning on purpose in order to make it artificially appear that Intel’s parts on average are better at “overclocking”.

Intel does still have a sizable 14nm process node tweaking/generational update advantage compared to the rest on the industry’s newer less mature 14nm processes. So Intel has a lot of intentional under binning headroom while still being able to have higher relative non turbo-boosted base clocks in all of its past/latest CPU SKUs at 14nm. So maybe there is some intentional under binning to give the impression of more Overclocking ability on the part of Intel’s consumer CPU/SOC product lines.

March 4, 2017 | 11:00 PM - Posted by Anonymous (not verified)

P.S. officially supported means explicitly covered under warranty. Any unofficial usage will void the part’s warranty.

Also Note that AMD’s parts have never been artificially clock locked and AMD has never offered any “K” like branding in the past to ever imply that overclocking is “officially” sanctioned. Even the Ryzen XFR feature is not really the same as user/manual overclocking that can damage a CPU and void any warranty. XFR is more of an extended boost under strict control of the firmware so as to not damage the CPU part’s warranty viability.

March 6, 2017 | 04:33 AM - Posted by Pholostan

It's the very first version of it. The engineers stressed this at the presentation. They want to improve it for the next version.

March 4, 2017 | 12:09 PM - Posted by StephanS

XFR seem like a super minor feature right now.
It might have worked better if it was tied to voltage boost.

8core, 3ghz, 65w, 1.2v
why cant it do , 4 core, 4ghz, 1.35v

like a 65w gaming mode : 4 core 4 ghz
and a 65w productivity mode : 8 core, 3ghz

At $330, I'n still curious to see how a 1700 perform with 4 core disabled for gaming. Seem like no one had the idea to experiment with this. ?

March 4, 2017 | 03:36 PM - Posted by Anonymous (not verified)

Hey, AMD has a great Server/workstation micro-arch with Zen so that's what will really save AMD as a business. Zen has some good power metrics and even if this first Zen Micro-Arch lacks the Total FP performance metrics of Intel's SKUs in the HPC market's need for good FP metrics, AMD has its professional GPU Accelerator/AI SKUs to pair with Zen/Naples and other Zen server/HPC/WS CPU only SKUs to do even more FP/AI number crunching than any CPU can do. Those Xeon/Kignts whatever SKUs run very hot at very high clock rates just to be able get near any GPU SKUs in DP FP/T-flops metrics and still the new GPU SKUs from both AMD and Nvidia can to do more DP FP/SP FP and 16 bit/AI FP T-Flops at much lower power saving clock speeds than any CPU can.

AMD will also be introducing a new class of APUs on an Interposer designs for the workstation/Server/HPC, and even exascale markets, that will eat any CPU only SKUs lunch for RAW FP performance!

March 4, 2017 | 07:17 PM - Posted by CNote

I'd like to see them do some dual socket capable Ryzen for the workstation/server market. Would be vastly cheaper to get 16 cores/32 threads or more.

March 5, 2017 | 01:09 AM - Posted by Anonymous (not verified)

Yes they will but under the Opteron/whatever Pro/Server or workstation Motherboard/SKU branding and not the Ryzen consumer branding. There are plenty of Opteron/server features in the Ryzen parts that are turned off/disabled but any parts that have these features enabled for dual socket/NUMA usage will come with a different branding. Zen is the micro-architecture used on both the Ryzen(consumer) and Opteron(or New server brand name if AMD chooses) Server SKUs.

I'm sure that there are plenty of features baked into the Ryzen core consumer dies that can be very easily not disabled on some Ryzen refresh SKUs to make them more competitive with Intel's consumer E parts. The Server parts tweaking and finial RTM certification is very likely still going to be ongoing on up to the server parts RTM dates but the server customers will get those RTM samples much sooner that any consumer market gets the final product to test.

They can not play around with the server/workstation/HPC customers like they do with consumer market products. Those server customers have the PHDs or access to the PHDs as consultants to really get any issues fixed before any contracts are signed. Look for 16 or even maybe some 12 core Opteron(Or new pro brand name) low cost Server/Workstation variants from AMD in addition to the Zen/Naples 32 core flagship variants.

Look to the professional server/workstation/HPC market review websites to get the proper information on the AMD Infinity fabric IP questions answered also.

March 4, 2017 | 06:32 PM - Posted by Anonymous Nvidia User (not verified)

Yeah with XFR automatically boosting CPU to max does that mean their TDP is false advertising and possibly actionable with class action suit.

Did any reviewer give any hard numbers on TDP at normal maximum?

Let's forget about the fact that AMD sent an email to Gamers Nexus telling them to GPU bottleneck game benches. I don't think they are the only site to get one.

Heck after waiting this long, they still couldn't get everything right. Maybe the 4 core Ryzens to come were designed with gaming in mind and may be more competitive.

No matter what happens, all Intel needs to do is slash prices and lose a little profit until it's new chips are released.

March 4, 2017 | 07:23 PM - Posted by CNote

Makes sense the 4 core models will be better for gaming. These 8 core monsters are overkill for regular gamers unless they also 3d render at the same time. These might even be overkill for the Twitch/Youtube crowd but the 4/6 core models will hopefully be able to clock higher and actually use the XFR. To me the XFR sounds more useful for a mobile chip with some kind of fancy cooling, maybe liquid or special vapor chamber, or the way lower end ones. I was never fooled by the X designation meaning they were the only ones to get XFR, clearly it means they are the 95 watt versions.

March 5, 2017 | 04:59 AM - Posted by Anonymous (not verified)

Speaking of overclocking apparently Ryzen responds very well to faster RAM.

Post new comment

The content of this field is kept private and will not be shown publicly.
  • Lines and paragraphs break automatically.
  • Allowed HTML tags: <a> <em> <strong> <cite> <code> <ul> <ol> <li> <dl> <dt> <dd> <blockquote><p><br>
  • Web page addresses and e-mail addresses turn into links automatically.

More information about formatting options

By submitting this form, you accept the Mollom privacy policy.