New Ryzen 2 (Pinnacle Ridge) gets only 200 MHz boost according to a leak

My experience has been that IdiotInCharge and I might disagree with certain things, but at least I find him consistent and I respect his view because we have different needs for our computers. I believe that if AMD had a solution that met his needs, he'd buy one. He's not really in the same camp as juanrga who will blindly defend Intel at all costs.
 
juanrga said:
Funny comment, when the $999 Threadripper is slower than the $499 1800X in the graph he gave.
Click to expand...

Irrelevant. Anybody who buys a 16 core CPU for strictly gaming is a lunatic.

That and the 1950X shows decent enough showings in most games. It's just irritating that you have to play with the modes to get better performance in different games.
 
Gideon said:
You and a couple of others never have a point other then to say nope Intel is better. The mode operates like AMD intended it to, until you get that engineering degree your out of luck on telling them how it should work. Plus it's not freaking hard to manually set a frequency if someone wanted to, review sites reviewed it in stock form.
Click to expand...

Mi comment #387 was about AMD vs AMD: TR 1950X vs R7 1800X. You can believe there is no way to improve the gaming mode in ThreadRipper, but you will continue being wrong.
 
DuronBurgerMan said:
Irrelevant. Anybody who buys a 16 core CPU for strictly gaming is a lunatic.

That and the 1950X shows decent enough showings in most games. It's just irritating that you have to play with the modes to get better performance in different games.
Click to expand...

I would but then again, I love having what every I want and can afford. That said, I have a ton more games than 15 years ago, far better hardware now that what I would have had 15 years ago, compared to what was available then and yet, I hardly ever game anymore. (I even bought BF1 for $9.99 just because it was cheap. :D :)
 
juanrga said:
Mi comment #387 was about AMD vs AMD: TR 1950X vs R7 1800X. You can believe there is no way to improve the gaming mode in ThreadRipper, but you will continue being wrong.
Click to expand...

Which has all of nothing to do with anything thing he said. It would be like him saying, look, the sky is blue and you respond with but my car is faster than yours. :rolleyes:
 
demondrops said:
that im not sure of as most benchmarks i did look at before showed nothing significant gains in performance for gaming with quad channel.but then again i could probs run my ram at stock dual channel without significant loss in performance where amd would require what 3000 mhz with ok timings to work properly?
Click to expand...




It depends on the title, but RTS games with lots of constant rendering would be very much affected by added memory channels, GTA also has a decent effect on performance. I wouldn't say it is a necessity for anyone to really consider quad channel at this point but in some instances there is notable differences.
 
but hey this is for threadripper? so it could be different for intel rigth? just like as i understood it amd is more sensitive to ram. i guess pointing out have intel got quad channel x99 falls out as an argument compare to amd enthusiast platform, or productivity platform. guess no one who plays game only need past 8 core now. at this point enthusiast platform do seem to be just casual productivity platform as its not really that great for gaming. well it is great for gaming but LOL... but faster mainstream for most part it seems.
 
Last edited:
OrangeKhrush said:


It depends on the title, but RTS games with lots of constant rendering would be very much affected by added memory channels, GTA also has a decent effect on performance. I wouldn't say it is a necessity for anyone to really consider quad channel at this point but in some instances there is notable differences.
Click to expand...


I wouldn't mind seeing a thorough comparison on an Intel quad-channel setup too- Ryzen just has too much memory goofiness in its first round.

[my guess is that the Intel system wouldn't care; also need to be careful as the quad-channel setups tend to run much slower individual memory modules, so ideally you'd want to run quite a number of comparisons, per arch]
 
IdiotInCharge said:
I wouldn't mind seeing a thorough comparison on an Intel quad-channel setup too- Ryzen just has too much memory goofiness in its first round.

[my guess is that the Intel system wouldn't care; also need to be careful as the quad-channel setups tend to run much slower individual memory modules, so ideally you'd want to run quite a number of comparisons, per arch]
Click to expand...

I don't know about the extreme high end (3600mhz+), but I'm running an overclocked 7820X at XMP 3200Mhz/Cas14 with "Ryzen" memory G.Skill FlareX in quad channel (4x8GB) and I haven't had a single issue.
 
juanrga said:
Wasn't it only 999% better? :p Also I am giving gaming averages, just to avoid dependences on specific games and outliers. I am not the one mentioning Civ 6 and ignoring any other game.







So, not only now you cherry pick one game and ignore the rest of games, but you also cherry pick the review!!!

Here a collection of reviews that show a different picture for Civ 6, with Intel ahead of AMD







Also https://www.pcgamesn.com/intel-core-i7-8700k-review-benchmarks

Other reviews don't check FPS, but check the AI. Those are trustedreviews, gamernexus,... and also show Intel ahead of AMD

8700k-civ-vi-turn-time.png


I will stop here.
Click to expand...

Ran the AI benchmark on my machine and got 13.56 at 1440p with the current build of CIV VI with my chip locked at 3.8 and 3200 MHz memory. So I dont think that chart your using really represents how things are but also why canned benchmarks dont tell you jack and without the current build of the game listed , it makes it impossible to compare.
 
Dan_D said:
I figured this would be an incremental upgrade at best. I'm more interested in X470 and why it's getting a new chipset. This isn't normal for AMD.
Click to expand...
Its a upgraded chipset yes but its not a REQUIRED chipset. That is the difference.
 
jenosmaverick said:
Just saw this on reddit, the Zen+ 2700x at 4.1 and will probably reach 4.2 with xfr.

https://www.computerbase.de/2018-03/amd-ryzen-7-2700x/
Click to expand...

hopefully there's some overclock headroom there.

i still think the R5 2600 clocks are wrong though.. if they're planning on releasing an 2800x the 2600x will have matching clocks which means they'd be at least 200mhz higher than the 2700x and the 2600 would make much more sense to have matching clocks with the 2700x. even though AMD has the 1600 rated at 65w with the lower 3.2/3.6 clock rate the 3 R5 1600 systems i've built have run at 3.4Ghz/3.8Ghz(same as the 1700x) out of the box with 3 different manufacture x370 motherboards, my assumption is that the paper rated clocks are for B350/OEM shit.

soooo if the 2700x numbers are in fact correct and AMD does plan on releasing a 2800x the base clocks would be 3.9Ghz with a boost of 4.3Ghz and XFR clocks @ 4Ghz/4.4Ghz which would mean you have roughly 600-700mhz of overclock headroom on a 1700x.. but thats just personal speculation dependent on how correct computerbase's information is and whether or not AMD has decided to drop the 2800 series from their processor lineup. either way it sounds promising.
 
I'm thinking 4.1-4.2ghz overclocks is probably what we'll see out of these chips with the golden sample/unicorn chips doing 4.3ghz with crazy voltage. 200mhz increase isn't bad, but it's also not an incredible jump. It will help close the gap a bit though.
 
Sooo... about the same as a 1900x. Not very impressive. I don't think they are making a 2800x either. I always thought it was goofy that there was an R7 1700, 1700x and 1800x. They often clocked very close anyhow and, unlike Intel, overclock able motherboards are cheap.

Still, too early to make assumptions of max overclocks, especially with b450/x470, but I have a feeling that I will not have any buyers remorse with my 2400g.
 
Nightfire said:
Sooo... about the same as a 1900x. Not very impressive. I don't think they are making a 2800x either. I always thought it was goofy that there was an R7 1700, 1700x and 1800x. They often clocked very close anyhow and, unlike Intel, overclock able motherboards are cheap.

Still, too early to make assumptions of max overclocks, especially with b450/x470, but I have a feeling that I will not have any buyers remorse with my 2400g.
Click to expand...

i don't think x470's going to make much of a difference as far as overclocking.. it'll be more of a feature upgrade then anything else. i agree on the 2800x.. the 1800x always seemed like a halo product to show what ryzen was capable of reaching but not a mass market processor. they may do the same thing with zen+ since there's not much of a profit loss on their end releasing it if the yields are good.
 
Moderately disappointed, TBH. I mean, I guess we'll see the real benchmarks soon, and we can see how the things overclock and how the latency and memory controller improvements factor in - and maybe that'll make it a little more impressive, but 250-350 MHz higher boost clocks depending on model, and 100MHz higher base clock for the 2700X vs 1800X just isn't particularly interesting. Unless AMD is holding back a heavily-binned 2800X model for release later in the year or something.

If the OC headroom is in the 4.4GHz range, though, that's a different kettle of fish. That could be interesting.
 
DuronBurgerMan said:
If the OC headroom is in the 4.4GHz range, though, that's a different kettle of fish. That could be interesting.
Click to expand...

Very much so. That closes the gap significantly for gaming, and probably (along with the other enhancements over OG Ryzen) will put the majority of multitasking workloads firmly in AMD's wheelhouse.

The kicker is whether or not that OC headroom is broadly realized :).
 
  • Like
Reactions: N4CR
like this
Don't know. I would diffuse the slightly lower bin parts first and save the golden samples for an encore.
They have to conserve higher bin samples for TR+ I would think.
 
CPC Hardware will be publishing a review of the 2700X on a A320 mobo on Monday. This comment says the tldr is 20W for 200 Mhz over 1000 series (from someone that has the magazine already somehow). Explanation from CPC Editor in Chief on why test on a A320 mobo.
 
DuronBurgerMan said:
Unless AMD is holding back a heavily-binned 2800X model for release later in the year or something.
Click to expand...

Yields will only continue to improve as this process matures meaning that your highest binned chips are most valuable the second they're made. It's also when demand will be the highest. If they get, some time down the line but far enough preceding Ryzen2, enough good chips to create a new differentiated product, then it's time to create a new SKU. Until then having fewer part numbers lowers costs/inventory. There's assuredly a cost:benefit ratio here. :)
 
IdiotInCharge said:
Well, that's better than a fracking ceiling no matter how many watts you throw at it. Call it progress!
Click to expand...

Unless AMD has already clocked it max in the factory. Sadly, for some reason they aren't testing it on a overclockable mobo, and tbh, I have no idea what their explanation means regarding why.
 
Chimpee said:
I expect better overall Overclocking if yields are better.
Click to expand...

I would too, if the precedent set by OG Ryzen had not been wall

And this isn't an architecture change, but a process upgrade with perhaps some minor tuning- so I give it 50/50?
 
Like you guys said, I'd expect a higher %age (assuming no new SKU's) of the later 2700x's to be better overclockers as the process improves. Right off the bat? All bets are off until we have a decent idea what people are finding from the silicon lottery.
 
I look forward to seeing what these chips can do but I will wait for the 7nm version coming next year.
 
As a point of caution -- I have *no* idea how well the 12nm process will actually overclock, just that we'll generally see a downward trend in voltage needed to hit a frequency, or an upward frequency at a given voltage. But these could be really marginal, or end up as significantly cooler chips with miniscule OC headroom. We don't have that kind of info *yet*.
 
SighTurtle said:
Unless AMD has already clocked it max in the factory. Sadly, for some reason they aren't testing it on a overclockable mobo, and tbh, I have no idea what their explanation means regarding why.
Click to expand...

"Pour éviter les cheats à la con omniprésents dans les BIOS des X370/X470" = To avoid the omnipresent cheats in the X370/X470's BIOS.
 
Last edited:
You must log in or register to reply here.
Back
Top