AMD 2nd Gen Ryzen 2 2700X Zen+ CPU Review @ [H]

In my opinion you played it right when you had some doubts. It is better to err on the cautious side than mislead a whole bunch of your viewers. It is easy for us scrubs to throw stones....after all we have no accountability. Thanks Kyle for the honest approach. It worked out right in the end. And isn't that what testing and the bleeding edge is about? It is not the absolute assurance of error free installations...it is finding what is actually the case.
 
Ihaveworms said:
That poor 8700k died for this review and for our entertainment. I am a little surprised seeing the 1800x beating the 2700x at the same clockrate in some benchmarks.
Click to expand...
For decades Intel enjoyed a monopoly, they started ignoring this part. I am glad to see the competition soon begins
 
Man, looks good for Ryzen 2.
The only benchmark I ever care about is video transcoding with HandBrake.
Video codecs utilize all cores and provide a real productivity oriented benchmark.

I really wish audio codes would be reworked to utilize more than 1 core.
 
Kyle,
Would a review of different ram with the 2700x be possible?
I myself have the same Asus ROG Hero and have debated on faster ram or tighter latency.
Im thinking better latency would increase performance than higher clocked ram.

just my $0.02
 
griff30 said:
Kyle,
Would a review of different ram with the 2700x be possible?
I myself have the same Asus ROG Hero and have debated on faster ram or tighter latency.
Im thinking better latency would increase performance than higher clocked ram.

just my $0.02
Click to expand...
I will have full followup coming with 3400 and 16 cas timings at stock and PB2.
 
Hey, I got 3000MHz to work with this RAM kit finally. Needed a *slightly* higher voltage. 1.36v lol. For 32GB double sided RAM, that's not half bad.
 
I put together my new 2700X build last night. Used 14CAS 3200 on the Strix X470 board, everything is working great so far. Just running stock for now on the stock cooler, haven't decided what to cool it with yet, will be looking out for that next.
 
Ritorix said:
I put together my new 2700X build last night. Used 14CAS 3200 on the Strix X470 board, everything is working great so far. Just running stock for now on the stock cooler, haven't decided what to cool it with yet, will be looking out for that next.
Click to expand...

Same but on a MSI M7 board. I actually love the stock cooler. I can't hear it and it has no problems keeping it cool. Plus the air flow cools the VRMs.

Things feel more snappy than my i7 5960x but it's probably just in my head.
 
travm said:
Having used way too many FX cpu's I can say, they often ran better when you let them control their own clocks, especially if you didnt have a mobo that had the ability to reliably supply 200w of power to the cpu. AMD has had dynamic clocking and power under control for some time. Bulldozer cpus were really a tech marvel, unfortunately they caught fire and burned on launch. I expect overclocking will die in the next 2-3 years as AI takes over and just pushes the chips as far as they can go all the time. Its been happening for 5+ years now.
Click to expand...

How well does dynamic clocking take advantage of the type of temps that would only be seen in high end custom loop or sub ambient loops?
 
Zarathustra[H] said:
How well does dynamic clocking take advantage of the type of temps that would only be seen in high end custom loop or sub ambient loops?
Click to expand...

Well for one, I'm pretty sure Kyle is seeing higher stock clocks on his cooling setup than I'm seeing (I am on a Noctua U12). His scores are marginally higher than mine across the board at stock, but were looking about the same OC'd to 4.2. This suggests Precision Boost & XFR2 are giving him a bit more headroom.
 
Zarathustra[H] said:
How well does dynamic clocking take advantage of the type of temps that would only be seen in high end custom loop or sub ambient loops?
Click to expand...
On the fx it didn't at all. Cooling was only helpful for all core overclocks. I however never bought a motherboard with good enough power delivery to be able to overclock much without the motherboard throttling. In one case overclocking at all caused throttling, and the vrms and Northbridge would get extremely hot. Best performance letting the cpu boost itself.
PB2 seems to be the evolution of this.
 
DuronBurgerMan said:
Well for one, I'm pretty sure Kyle is seeing higher stock clocks on his cooling setup than I'm seeing (I am on a Noctua U12). His scores are marginally higher than mine across the board at stock, but were looking about the same OC'd to 4.2. This suggests Precision Boost & XFR2 are giving him a bit more headroom.
Click to expand...
Will have Wraith Prism and Water numbers this week.
 
OK, finally got the data done just now for the Precision Boost 2 followup. A total of 96 benchmark runs will be represented on the graphs....which was likely 400+ runs including clocks, temps, and vCore runs, failures and verifications. That said, we have a hand on PB2, just need to get it all written and graphed up.
 
Looks like anandtech updated all its gaming charts again. 8700K back on top. Maybe next year, Zen2 :(

That said, still great value for money.
 
mda said:
Looks like anandtech updated all its gaming charts again. 8700K back on top. Maybe next year, Zen2 :(

That said, still great value for money.
Click to expand...

Yeah no regrets here. Difference is pretty negligble and factor in the 2700x comes with an actually useful heatsink it's really good value.
 
mda said:
Looks like anandtech updated all its gaming charts again. 8700K back on top. Maybe next year, Zen2 :(

That said, still great value for money.
Click to expand...

I should have saved the old charts. I suspect their Intel gaming results were the ones that were off, given that their Ryzen productivity results were in line with everyone else. But I'd like to know for sure.
 
DuronBurgerMan said:
https://www.anandtech.com/show/12678/a-timely-discovery-examining-amd-2nd-gen-ryzen-results/5

Well that's funny. The HPET problem was largely on the Intel side. AMD results within margin of error compared to initial pass, but Intel results wildly improved. That's about what I suspected, a problem on the Intel side of their testing.
Click to expand...

I don't fully understand, is Intel actually faster or is inaccuracy in the timer when HPET isn't used causing them to appear to be faster with it disabled?
 
Just slapped my new 2700x into my ASUS PRIME B350-Plus .. working great! Make sure you update your BIOS though sports fans before you make the swap!

** and ** .. a bit of an update ** I am now running my Team Group (2 x 16GB) ram at 2933 .. surfing while re-encoding an MKV movie to MP4 without any "wierdness". My setup wouldn't even boot with the 4008 BIOS & 1700 using the D.O.C.P. memory setting . This memory ran above specs on previous 7700K Z170 setup.

We'll give it a couple days of gaming and what not and see if it's for realz ..
 
Last edited:
DuronBurgerMan said:
https://www.anandtech.com/show/12678/a-timely-discovery-examining-amd-2nd-gen-ryzen-results/5

Well that's funny. The HPET problem was largely on the Intel side. AMD results within margin of error compared to initial pass, but Intel results wildly improved. That's about what I suspected, a problem on the Intel side of their testing.
Click to expand...

Correction, the problem is on the intel side, not the intel side of their testing. Intel's hardware and platform is a mess, it's their fault and their problem that their hardware is full of glitches and bugs and can't cope with everyday computing that is expected of it.
 
i not looked at the 12 pages yet , i assume MCE was disabled on the 8700k stock tests (as some high end mobos have it enabled by default which is incorrect, as it should only ask to be enabled when XMP is used or the person manually turns on MCE)
 
Last edited:
bobzdar said:
I don't fully understand, is Intel actually faster or is inaccuracy in the timer when HPET isn't used causing them to appear to be faster with it disabled?
Click to expand...

In games, the 8700k is generally faster. The HPET issue was causing the 8700k to appear slower than it was in Anandtech's review. Anandtech has updated the charts, which are now in line with the other reviewers, showing the 2700X at somewhere around 90-94% of the 8700k's gaming performance.

The 2700X remains generally faster in multi-threaded applications, as expected.
 
DuronBurgerMan said:
In games, the 8700k is generally faster. The HPET issue was causing the 8700k to appear slower than it was in Anandtech's review. Anandtech has updated the charts, which are now in line with the other reviewers, showing the 2700X at somewhere around 90-94% of the 8700k's gaming performance.

The 2700X remains generally faster in multi-threaded applications, as expected.
Click to expand...
Are there many gaming reviews at 3400-3600? The gaming performance seems to jump from limited data I've seen.
 
N4CR said:
Are there many gaming reviews at 3400-3600? The gaming performance seems to jump from limited data I've seen.
Click to expand...

Most are running at least 3200. I've seen some running faster. The 8700k is still a better gaming-only CPU. For mixed use, the 2700X is at least as good - if not better, depending on use case.

I regard this release as a fundamental tie. Both chips are good, your choice should boil down to what you're doing with it.
 
DuronBurgerMan said:
Most are running at least 3200. I've seen some running faster. The 8700k is still a better gaming-only CPU. For mixed use, the 2700X is at least as good - if not better, depending on use case.

I regard this release as a fundamental tie. Both chips are good, your choice should boil down to what you're doing with it.
Click to expand...

Well said and agreed. Hilarious turn of events really, the usual predicting suspect appears to have had to eat more than their share of humble pie this time around. Many [H] users seem to be dual use or slightly more gaming orientated. When it's 5-10%, I'd chose the Ryzen because you'd notice that multi thread difference much more than a few % gaming at low res (of course if you're one of the billions of pro-CS gamer champs running 10KHz/140p then intel all the way)..
I didn't expect Zen+ to be so competitive either, few % less in gaming than we see was what I'd expected, the reason is I think most of us overlooked the IMC/memory speed optimisation and how important that has been for performance. No wonder Intel hired Jim! Pretty hilarious really.

What I'm really excited for is Zen2 and I will very likely jump on that, I think the gaming difference will be down to a few % by then with the clock bumps and further tweaks. Hopefully Navi is out by then too.. be nice to run an all-AMD rig again, been over a decade!
 
Bought a R5-2600X from a forum member, along with the Asus X470 Prime. I cannot wait to play with them!
Way to go AMD, been a while since my last AMD proc, the Barton Core 3200.
 
Are there any real world gaming benchmarks out? At 4k everything is gpu limited, and even at 1440p some titles are...so is there any reason to go Intel over AMD. My 3570k can't keep up with my 1080ti so I was waiting for this before deciding. Seems the additional cores would prove more useful for streaming and doing other crap while gaming.
 
bos said:
Seems the additional cores would prove more useful for streaming and doing other crap while gaming.
Click to expand...

This is generally the case; streaming is system intensive, 'other stuff' is obviously open to interpretation.

But bear in mind that if you keep CPUs as long as you've kept the 3570k, games will evolve and single-core performance will likely increase in demand.
 
IdiotInCharge said:
This is generally the case; streaming is system intensive, 'other stuff' is obviously open to interpretation.

But bear in mind that if you keep CPUs as long as you've kept the 3570k, games will evolve and single-core performance will likely increase in demand.
Click to expand...

You could also argue games are likely to become more thread aware, so the extra cores may end up out weighing the ST performance. There's also the fact that this honestly doesn't matter much for anyone playing above 1080p, especially 4k users. The 8700k is definitely the better gaming cpu if every single frame matters to you (144hz monitor or something), but beyond that, the new Ryzen is close enough that most people won't notice any difference tbh.
 
IdiotInCharge said:
This is generally the case; streaming is system intensive, 'other stuff' is obviously open to interpretation.

But bear in mind that if you keep CPUs as long as you've kept the 3570k, games will evolve and single-core performance will likely increase in demand.
Click to expand...

Yeah maybe, but it seems like most are getting to a point they're going to start using up to 12 threads if available. The 8700k could temp me with the right price drop, but it'd need to get in the 250-275 range. I'm half tempted by the 2600x as well as an alternative.
 
bpizzle1 said:
the new Ryzen is close enough that most people won't notice any difference tbh.
Click to expand...

I mostly don't disagree. But that 1080Ti...

Also, the challenge (as I'll respond to below) is that we don't know what games have in store. But generally speaking, what games do- real-time interactive simulations- is hard to parallelize, and expanding features or whatever comes will likely come with single-core stresses as much as ability to use extra threads.

bos said:
Yeah maybe, but it seems like most are getting to a point they're going to start using up to 12 threads if available. The 8700k could temp me with the right price drop, but it'd need to get in the 250-275 range. I'm half tempted by the 2600x as well as an alternative.
Click to expand...

They are! Though the performance of the 8600k is frankly surprising and indicative of the uneven nature of game development. Some games love threads, some just need enough threads and love high single-core throughput. Do you know what the games you will want to play will err toward?


My bet is that the 8700k has the best mix of both. The extra two cores are themselves pretty extraneous as performance over the 6700k/7700k at same speeds is generally nil outside of significantly multi-threaded corner cases, and represent future-proofing, which is what makes the 2600k such a great gaming budget buy now that RAM compatibility has stabilized and boost speeds are competitively useful, and it will probably present an increasingly better overall experience 2+ years into the future than the 8600k (though will likely still be slower for dedicated gaming). But above that, for gaming, 8700k all the way. For anything else, 2700k until Intel's next release.
 
DuronBurgerMan said:
The 8700k is still a better gaming-only CPU.
Click to expand...

Is that really true if you're not using a 144 or 120Hz monitor? I mean, I'm not a hardcore gamer, and I use a 60Hz monitor. At that point, whether a game gets 115fps or 130fps is irrelevant.
 
I have a 1080ti and with my 5960x I'd sometimes get hitching. Haven't seen it with the 2700x and I did absolutely nothing software wise... I just plopped in Ryzen/mobo/ram and hooked it up. So I can't complain.
 
Dayaks said:
I have a 1080ti and with my 5960x I'd sometimes get hitching. Haven't seen it with the 2700x and I did absolutely nothing software wise... I just plopped in Ryzen/mobo/ram and hooked it up. So I can't complain.
Click to expand...
What did you have the 5960x OC to?
 
You must log in or register to reply here.
Back
Top