• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.
  • The forums have been upgraded with support for dark mode. By default it will follow the setting on your system/browser. You may override it by scrolling to the end of the page and clicking the gears icon.

AMD Ryzen XFR Frequencies Revealed

The 1600x looks nice. I may use this as my next VR machine if it lives up to the hype or a 1700x. I just bought a 6800k for a family member, and it's a really good chip, but it fell short on overclock compared to my 5820k. but it will do everything he needs it to do. I still have my 1055t chip that I use for crunching, but I've not used an AMD chip for my gaming rig in a long time. I really do hope these chips out perform the 2011 v3 chips.
 
I am grow tired to see these..speculative/guesstimate
Just. Release. Damn. Chip already !
 
hmm... still not gonna convince me turning to the red camp...


So why thread crap? Are you just being an asshole and felt like your comment should somehow be here?


Anyway. On to the thread topic.....


Im saving for a setup, its been a long time without a new computer for me, and Im buying one as long as the numbers meet the expectation. If not I am going Intel. Whatever is best for the buck in performance.


As far as the XFR and TDP, this is the same company that released overvolted 480's that got too warm to maintain stock clocks. Threre has to be a max voltage put into the chip for the process used, and I'm sure that will be the deciding factor in how high of a clock any chip can achieve, be it 5Ghz or not. They might die test and program each chip for its maximum XFR range and voltage. Either way its going to be interesting to see what the chip is capable of on its own VS manual tweaking.
 
Last edited:
These are fake, read TPU forums.

The renders might be, but they're clearly doing something very similar of you look at the picture tweeted by Raja and there are still three thermal solutions, so no, it's not all fake.
 
I find Ryzen's lineup extremely confusing.

You have the Xs, non Xs, Pros and no Pros,
Different TDPs among the above mentioned,
3 coolers (HS55, HS65, HS81) under 3 categories (A, B, F)
You have the R5 that can either be a 6/12 or a 4/8,
Cherry on top you have the XFR thing
oh, and there's temps, the 1800X is supposed to stay below 60C, while the Pro 1800 (same TDP) can reach 72C

Confusing

The Pro models are likely to be for business applications only, as AMD has done this in the past.

The temperatures are with the supplied coolers, not with third party cooling.
 
i missing 20mb cashe :/



AMD-Ryzen-01.png
 
I find Ryzen's lineup extremely confusing.

You have the Xs, non Xs, Pros and no Pros,
Different TDPs among the above mentioned,
3 coolers (HS55, HS65, HS81) under 3 categories (A, B, F)
You have the R5 that can either be a 6/12 or a 4/8,
Cherry on top you have the XFR thing
oh, and there's temps, the 1800X is supposed to stay below 60C, while the Pro 1800 (same TDP) can reach 72C

Confusing


You mean confusing like an i5 can be a dualcore dualcore with HT and a quadcore? :D

To be honest; Im pretty sure you'll manage WHEN they actually officially release these details other than a clusterf*** of leaks fabrications and segments of info. TO me at least for now it seems more easily understood than intels with their pentium and celeron brand + overlapping i3-i5-i7 segments on laptop market. Lets wait and be NOT confused :)
 
I am so eager to see the single thread performance on these... If it's good then this will be a nice contender for my next upgrade. Holding off on the 7700K for the time being. Might go team red just for the heck of it if the performance is close enough.
 
The renders might be, but they're clearly doing something very similar of you look at the picture tweeted by Raja and there are still three thermal solutions, so no, it's not all fake.
We were talking about images posted earlier, which actually are renders. A member of other forum admitted making them.
 
Well, that's what's XFR. That's like predicting how far exactly NVIDIA Pascal GPU will boost. You can't. It's the same here. CPU has so many parameters involved in it you can't predict the final clocks. They can estimate from their internal tests, but that's about it. If you slam phase change on this CPU it might clock up to 5GHz if parameters and conditions for XFR will allow it. If you'll only have AiO, it might be just 4.3 GHz.
Bottom line is, Turbo boost is guaranteed if basic conditions are in check. XFR just means extended potential which is not guaranteed and heavily depends on conditions.
Sadly enough, the potential and reliability of XFR will probably be like the GPU boost like you say, totally dependent on the conditions and the quality of each chip. Also, this is just going to make open rig benchmarks just even more worthless.
 
Waiting on Launch and getting hands on it for myself to tool with it, then manually oc it the oldschool way
 
Last edited:
TDP doesn't nessessarily reflect actual power consumption, it's the reference for required cooling capacity. They may put whatever number they want, as long as your system doesn't get unstable/fried with the cooling system meets that number. They may just overrated the 6-core R5 like a bunch of intel celerons/pentiums or underrated the 8-core r7, as turbo boost by definition is an overclock and comes with power consumption increase.
My router has similar situation. It comes with a 30w charger but I seriously doubt it can ever reach 15w at peak.
Also up for consideration is binning. It is posssible that the quad cores will be failed bins, which would result in higher power consumption per core.
 
I always wondered when this type of "tuning" would make it to a CPU.

Some newer direct injection vehicles use a similar technique to gain as much power and fuel economy based on the Octane ratio of the fuel. For example some DIT engines will advance timing until knock is heard. Once knock is heard it either pulls timing or stops advancing timing. After a while the vehicle learns the optimum timing. If you get a bad tank of gas the car will recognize this and begin to pull timing until knock is no longer detected.

It would be sweet if the new AMD cpu is constantly trying to up the frequency until it finds an optimum speed and then adjusts when temps get too hot(dirty heat sink)
 
I always wondered when this type of "tuning" would make it to a CPU.

Some newer direct injection vehicles use a similar technique to gain as much power and fuel economy based on the Octane ratio of the fuel. For example some DIT engines will advance timing until knock is heard. Once knock is heard it either pulls timing or stops advancing timing. After a while the vehicle learns the optimum timing. If you get a bad tank of gas the car will recognize this and begin to pull timing until knock is no longer detected.

It would be sweet if the new AMD cpu is constantly trying to up the frequency until it finds an optimum speed and then adjusts when temps get too hot(dirty heat sink)


This is what most open center ECU's do, advance timing to a maximum preset point or until knock occurs. Direct injection works a bit different, it uses the injection point and multi-pulse injection to start the "explosion" at a preset time, then determines when to inject the majority of the fuel load to achieve maximum power and may inject more fuel after that to control emissions.

I think it would be cool too, even with dirty heat-sink it may allow short burst high speed for single threaded applications.
 
I like what im seeing so far and this XFR feature is a cool and smart idea, this also helps with aftermarket cooler sales as well id say as if you want higher XFR turbo speed just get a better cooler? Also would this mean a lower model could clock as high as the top models just by having a better cooler on them? (removing all manual OC of course)
 
I like what im seeing so far and this XFR feature is a cool and smart idea, this also helps with aftermarket cooler sales as well id say as if you want higher XFR turbo speed just get a better cooler? Also would this mean a lower model could clock as high as the top models just by having a better cooler on them? (removing all manual OC of course)

I agree, neat way of trying to maximize the chips potential.
 
I always wondered when this type of "tuning" would make it to a CPU.
Unfortunately this seems to be the norm for GPUs, and probably all future CPUs as well, which makes it very hard for end users to determine the actual performance they can expect. It wouldn't surprise me if Intel will do something similar with Cannon Lake to achieve it's "15% performance increase".

Some newer direct injection vehicles use a similar technique to gain as much power and fuel economy based on the Octane ratio of the fuel. For example some DIT engines will advance timing until knock is heard. Once knock is heard it either pulls timing or stops advancing timing. After a while the vehicle learns the optimum timing. If you get a bad tank of gas the car will recognize this and begin to pull timing until knock is no longer detected.
I would say that turbo-charging of combustion engines is a better analogy; Let's say you have an engine rated for 150 hp, but with a large turbo it's rated 200 hp, but that's under ideal conditions, in the real world it introduces a huge variance in actual performance, which in turn makes it harder to drive comfortably. This variance is greater with smaller engine blocks and larger turbos, which makes it impossible to compare one "200 hp" engine to another based on rated performance.
</off topic>
 
How can a Ryzen 7 (8 cores) and a Ryzen 5 (4 cores) - with the same clock speed - have the same 65W TDP? My math doesn't accept that :p I'm building a HTPC and was hoping for something way lower than that, as you can get a i5 7500 using only 42W (65W TDP includes the iGPU).
Since your building a HTPC you could probably save even more power if you wait a bit longer for Raven Ridge with Vega on board. Even HBM2 will be used on that APU, which is very interesting to say the least. Depending on power consumption it could revolutionize the HTPC ITX market and possibly even decrease the overall footprint further.
 
hmm... still not gonna convince me turning to the red camp...

Thats ok as ill be going to camp green, as camp red we still have to wait till may. :p
 
So the question is to get a XFR based Ryzen or not?
The Canadian $$$$ amount for all 8-Core Ryzen CPU's are Expensive. Really Expensive. lol, I hate that in the USA, its a lot cheaper for stuff.
 
Back
Top