AMD Ryzen 2nd Gen Details: Four CPUs, Pre-Order Today, Reviews on the 19th

TaintedSquirrel

[H]F Junkie
Joined
Aug 5, 2013
Messages
12,683
https://www.anandtech.com/show/12642/amd-ryzen-2nd-gen-details-4-skus-reviews-19th

https://hothardware.com/news/amd-announces-2nd-gen-ryzen-2000-processor-family-on-sale-april-19

https://overclock3d.net/reviews/cpu_mainboard/amd_2nd_gen_ryzen_zen_2600x_2700x_preview/1

v9ge0wJ.jpg
 
It's great they include a cooler and all for the price, but I'd wish I could get an OEM version cause I'm going to throw the cooler away anyhow and replace with a water loop or beefier passive style heatsink.
 
It will be interesting to see if the 2700x is worth the extra $30 over the 2700. Can't wait for Kyle's review!
 
It will be interesting to see if the 2700x is worth the extra $30 over the 2700. Can't wait for Kyle's review!

Keep in mind you do get a better cooler as well. That may be enough to sweeten the deal depending on how good it is. I hope there are some indepth reviews comparing the various Wraiths together as well as against some of the popular aftermarket coolers.
 
So, are the X models really worth it over the non-X? I am pretty new to the AMD world but a new rig is in my future in the coming 1-2 weeks hopefully..
 
Wait for the reviews because nothing has been confirmed by reputable sources. Sure we will find out.
 
So, are the X models really worth it over the non-X? I am pretty new to the AMD world but a new rig is in my future in the coming 1-2 weeks hopefully..

X models typically have better binning. They also have 100mhz xfr stepping vs 50mhz on non x models.

Apparently LN2 overclockers have gotten 5.88 Ghz out of the 2700X and memory speeds up to 4500 Mhz on a 2200G on X470 MSI boards.
https://wccftech.com/amd-ryzen-7-2700x-ryzen-5-2600x-5-88-ghz-ln2-oc/

That's pretty good.. all first gens(including the 2200g/2400g) hit a brick wall around that 4.5ghz for the most part.
 
Last edited:
X models typically have better binning. They also have 100mhz xfr stepping vs 50mhz on non x models.



That's pretty good.. all first gens(including the 2200g/2400g) hit a brick wall around that 4.5ghz for the most part.

So for someone who says...
- Gaming Team Fortress 2 / Bf1 / World of Tanks and maybe the new FInal Fantasy game on an RX 570 4G
- Normal photoshop
-Couple VMs under vmware workstation..

Any real benefit?

Do the non X often over clock much?
 
So, are the X models really worth it over the non-X? I am pretty new to the AMD world but a new rig is in my future in the coming 1-2 weeks hopefully..

If you don't overclock, absolutely. If you do, maybe not - however, the 2700x also comes with the wraith prism so you get a good high end air cooler in the box. That coupled with the higher stock clocks and potentially better binning is easily worth the $30 over the 2700, imo.
 
Average OC for 1800X in LN2 is 5046MHz. The record is 5803MHz.


shooting my self for wasting my time responding to you but no.. the verified record for the highest overclock on ryzen is 5.4Ghz using only 2 cores on an 1800x.. the highest overclock using 8 cores is 4.5Ghz.. please do your research before you waste peoples time..
So for someone who says...
- Gaming Team Fortress 2 / Bf1 / World of Tanks and maybe the new FInal Fantasy game on an RX 570 4G
- Normal photoshop
-Couple VMs under vmware workstation..

Any real benefit?

Do the non X often over clock much?

should work just fine for all that.. with non X at least with first gen it's a little more luck based when it comes to overclocking.. some non X overclock great some overclock terribly like my r5 1600 while the 1600x's have a far higher chance of hitting 3.9-4Ghz my non X won't go past 3.8Ghz no matter how much voltage i give it.
 
X models typically have better binning. They also have 100mhz xfr stepping vs 50mhz on non x models.

My 1700 (when not overclocked) hits 3.2ghz when xfr kicks in - I think this was a prerelease thing. All chips do 100mhz xfr.
 
Should have been around during the Athlon and Netburst days and the obscure crap they tried to use to show Netburst as being better.

universal space heating standards, Intel warmed your winters so were more economical per watt.
 
All cores or one core?

All cores and thread activated.

shooting my self for wasting my time responding to you but no.. the verified record for the highest overclock on ryzen is 5.4Ghz using only 2 cores on an 1800x.. the highest overclock using 8 cores is 4.5Ghz.. please do your research before you waste peoples time..

If you say so...

http://hwbot.org/submission/3473875_der8auer_cpu_frequency_ryzen_7_1800x_5802.93_mhz
 
See my edit. All-core records for 1800X.

Sites are claiming that 2700X got 5.88GHz on all cores. Not true.

https://valid.x86.fr/vmjw9n

So 1800X gets 5803MHz one-core @0.787V
So 2700X gets 5884MHz one-core @1.76V

No one cares anyway what someone got on LN2 or liquid Helium. People will care what they get on water or high end air cooling also if you have a issue with another site then discuss it on their site unless it's a [H] review then their is no point complaining about it here.
 
No one cares anyway what someone got on LN2 or liquid Helium. People will care what they get on water or high end air cooling also if you have a issue with another site then discuss it on their site unless it's a [H] review then their is no point complaining about it here.

Do you know there is a functional relationship between LN2 cooling and cooling in air/watter? The structure of matter and the laws of thermo are the same. I am not comparing chips under LN2 because a direct interest in extreme cooling, but because from LN2 data we can extrapolate how the new Ryzen will overclock on air/watter.

If the 2700X had got a ~1GHz higher than 1800X, which is basically what sirmonkey1985 said when he pretended that "the highest overclock using 8 cores is 4.5Ghz" for 1000-series, then we could almost confirm that 2700X would be an excellent OC on air/watter. But as demonstrated, that 2700X on LN2 is performing worse than 1800X on LN2. Both get in the 5.8GHz range but that 2700X required ~2x higher voltage.

So don't expect miracles for the 2700X on air/watter. I still maintain my older prediction of 4.3--4.4GHz, but it starts looking as optimistic expectation.

Are you aweare that gigaxtreme1 introduced the WCCFTECH article about LN2 OC in this thread? That article says "The first result we have is for the AMD Ryzen 2700X which was overclocked to 5.884 GHz across all of the 8 cores and 16 threads". So it seems relevant to mention here, in this thread, that the 2700X OC was only for one-core.
 
Last edited:
So gigaxtreme1 mentioning WCCFTECH article about LN2 OC and sirmonkey1985 replying with a "That's pretty good.." and then he pretending that "the highest overclock using 8 cores is 4.5Ghz" for 1000-series and no one of you complain...

No need. You had it covered.

But just mentioning 1000-series already got 5.8GHz on LN2, mentioning those sites are wrong, that the OC was in a single core, not all-core (from the link given by gigaxtreme1 "The first result we have is for the AMD Ryzen 2700X which was overclocked to 5.884 GHz across all of the 8 cores and 16 threads" and suddenly "no one cares". Interesting.

Just pointing out a fact, mang. You of all people should appreciate accuracy, right?

Do you know there is a functional relationship between LN2 cooling and cooling in air/watter? The matter and the laws of thermo are the same. I am not comparing chips under LN2 because you are directly interested in LN2 cooling, but because from LN2 data we can extrapolate how the new Ryzen will overclock on air/watter.

Yes and no. The problem with LN2 comparisons is that we don't have a whole lot of them, and we have a lot of individual variance between them. There is a relationship, but the margin of error is much higher due to the smaller statistical size. I don't put a whole lot of stock in LN2 results, unless the difference is very large. I recall reading somewhere that somebody got 7.4GHz out of an 8700k LN2 overclock. Can't remember where offhand, but the difference between what the 8700k got on LN2 and the highest Ryzen LN2 OCs... we can see the relationship there. The 8700k will OC more. When we start picking nits between 5.4 and 5.8 though, it starts to become kind of lol. Better to just see the aggregate air/water results than try to extrapolate from the LN2 results in that case.
 
The Ryzen 1000 family seems to have issues with CPUID detecting real voltage.

uul2vi.png


7ghcct.png


zpjxdn.png


To achieve 5ghz + on any core count you need in excess of 1.7v

sdvjpz.png


I don't believe this is a legitimate voltage ready for De8aur's run, CPUID didn't properly report voltages and this is confirmed going through the full list of HWBOT's top 10 Ryzen 1800X Overclock scores 8 of the 10 reported 1.7v + and two got glitched readings.

vmjw9n.png


The difference with the 2700X is not that much, slightly lower voltages, I will think the difference in performance will be due to higher sustained clocks and some minor tweaks but Pinnacle Ridge is Ivy to Sandy, not worth upgrading to if you are presently on a higher end Summit ridge. Your Haswell upgrade is the next architecture. If you have the cash and don't mind exchanging parts then cool but I don't see people who are on Ryzen 7's now running to go buy these.
 
Do we have any insight into the next gen threadripper.

I need a workstation, I’m really struggling to pull the trigger. Money is tight but I need a heap of cores. Is it just going to be the same 200mhz bump do you reckon?

Not even convinced that 16 is enough but there’s such a large jump to epyc or the Xeon golds.

Stupid non-distributable, non cloudable workloads.
 
Back
Top