AMD Ryzen 7 2700X & Ryzen 5 2600X Review: Zen+ Is Here

" Still, it's well worth noting that the 2700X just managed to edge out the 7820X and that's a big deal."

Not really. Gamers are more likely to buy dual channel Intel chips for gaming, not quad channel HEDT chips. Price matters. And the different markets they are designed for matters. Ryzen could be sold as a work and play chip, but a 7820X is primarily for productivity. You could buy a 16T Ryzen chip because of its low price alone and go home and could get the same performance you desire for less. You don't buy a 7820X for those same reasons. You just don't. Common sense says so.

7820X was out before Ryzen and didn't have its memory compatibility problems. A lot of people are willing to pay extra for stability. Again, common sense.

I think the point was comparing the ipc of a mesh with a mesh instead of a ring bus, as the case with the 8700k. It shows that Ryzen's gaming performance is better than most give it credit for.
 
I don't agree with half of what you say amd taking their zen architecture for gaming/streaming and work related was a smart move.
I don't see alot people on the internet crying/complaining about pc gaming sucking on the ryzen platform. =/
Now for emulation such as dolphin, citra, a little bit of PCSX2, that I can understand because the ipc rate still sucks.
This is the main reason I steered into the kabylake platform and I am glad I did, but all in all amd has done some really good things lately.
I do not disagree with you; however, AMD, as I see it, did just enough of a job with gaming performance to keep their gaming customers satisfied. Benchmarks indicate that their lead over Intel in MT apps is significant, and this is where they need to be for the enterprise market.

IF AMD walks their talk, gaming will be the realm of Zen 2.
 
"Overclocked, we see good gains from the second-gen Ryzen parts but again I'll have to revisit these tests once I've had time to overclock the Skylake-X CPUs."

Some of this was already done with your 8 core Royal Rumble. The gaming results can't be compared as you have used new games of course.
New drivers can change things, but I included the 1700 ad 1800x results @ 4ghz to show little actually has.

CINEBENCH MT:
Before: 1700@4ghz = 1765 / 7820@4ghz = 1733 / 7820@4.5 = 1935
Now : 1800x@4ghz = 1755 / 2700x stock = 1771 / 2700x@4.2 = 1879

7 ZIP Compress (x1000):
Before: 1700@4ghz = 37.6 / 7820@4ghz = 47.4 / 7820@4.5 = 52.5
Now : 1800x@4ghz = 39.8 / 2700x stock = 40.5 / 2700x@4.2 = 41.0

7 ZIP Decompress (x1000):
Before: 1700@4ghz = 53.3 / 7820@4ghz = 48.6 / 7820@4.5 = 53.5
Now : 1800x@4ghz = 53.6 / 2700x stock = 52.4 / 2700x@4.2 = 56.1

Blender (lower = better):
Before: 1700@4ghz = 22.8 / 7820@4ghz = 20.7 / 7820@4.5 = 18.5
Now : 1800x@4ghz = 22.5 / 2700x stock = 23.2 / 2700x@4.2 = 21.6

Corona (lower = better):
Before: 1700@4ghz = 123 / 7820@4ghz = 117 / 7820@4.5 = 106
Now : 1800x@4ghz = 122 / 2700x stock = 120 / 2700x@4.2 = 108
 
Maybe I'm the only one who busts out laughing about this fascination with games , you're an adult for Christ sake , get a life !!!!!!!!


I played games at one time but I grew up after reaching ten years old.
 
"Overclocked, we see good gains from the second-gen Ryzen parts but again I'll have to revisit these tests once I've had time to overclock the Skylake-X CPUs."

Some of this was already done with your 8 core Royal Rumble. The gaming results can't be compared as you have used new games of course.
New drivers can change things, but I included the 1700 ad 1800x results @ 4ghz to show little actually has.

CINEBENCH MT:
Before: 1700@4ghz = 1765 / 7820@4ghz = 1733 / 7820@4.5 = 1935
Now : 1800x@4ghz = 1755 / 2700x stock = 1771 / 2700x@4.2 = 1879

7 ZIP Compress (x1000):
Before: 1700@4ghz = 37.6 / 7820@4ghz = 47.4 / 7820@4.5 = 52.5
Now : 1800x@4ghz = 39.8 / 2700x stock = 40.5 / 2700x@4.2 = 41.0

7 ZIP Decompress (x1000):
Before: 1700@4ghz = 53.3 / 7820@4ghz = 48.6 / 7820@4.5 = 53.5
Now : 1800x@4ghz = 53.6 / 2700x stock = 52.4 / 2700x@4.2 = 56.1

Blender (lower = better):
Before: 1700@4ghz = 22.8 / 7820@4ghz = 20.7 / 7820@4.5 = 18.5
Now : 1800x@4ghz = 22.5 / 2700x stock = 23.2 / 2700x@4.2 = 21.6

Corona (lower = better):
Before: 1700@4ghz = 123 / 7820@4ghz = 117 / 7820@4.5 = 106
Now : 1800x@4ghz = 122 / 2700x stock = 120 / 2700x@4.2 = 108

Data pre-Spectre/meltdown isn't comparable to data now.
 
Maybe I'm the only one who busts out laughing about this fascination with games , you're an adult for Christ sake , get a life !!!!!!!!


I played games at one time but I grew up after reaching ten years old.

Your personal sample size of one is exactly that. I'm sure you'll find some other who likes to make the same kind of unjustified insults. But here is the reality.

Please explain to me why there are so many elderly at the casinos gaming? Please explain to me why there are professional poker leagues and professional player making a living of a game as adults? Please explain to me why there are all these adults participating in the e-sports in leagues and tournaments as such, see list:
https://en.wikipedia.org/wiki/List_of_eSports_leagues_and_tournaments

Go ahead and laugh, but I've been building gaming machines for at least 15 years now, and found computer games to be really good mental exercise and fun for at least 30 years now.

I still can't appreciate modern abstract art, gangsta rap, or whatever that style of music that has lots of screaming and howling, but it does not give me the right to insult those who do.
 
The 2600X's multi core CB score is at 4.2GHz not 4.1. At 4.1 you should be seeing closer to 1400 points. You can cross reference the 2700X at 4.2 and 1600 at 4 to prove this.
 
Great article! Seems XFR2 nullifies the need for overclocking on the 2600X. When I initially got my 2600X, I wanted to test XFR2 on the Wraith Cooler vs. my H100i V2 AIO. Tried Cinebench with the included Wraith Spire I ended up with 173 Single Core and 1375 Multicore. Switched to H100i V2 with no overclock and it went from 173 to 178 Single core and went from 1375 to 1419 Multicore. XFR2 in action. I also wanted to know how many cores were boosted when gaming to see if the BOOST claims were bogus. Turns out it boosts ALL CORES the same. Play PUBG for 30 or 40 minutes and the boost on all cores using my AIO and watching while I play it bounces between 4.1 GHz and 4.225 GHz with an average of 4.15 GHz monitored with MSI Afterburner. That was about 150 MHz better than when I tried this with the Wraith. Can perform even a bit higher if I use a more aggressive cooling profile on the H100i. This is tested on an X370 MSI board with DDR4 3200 CL16 RAM.
 
It wasn't that long ago AMD didn't have anything to offer faster than an i5. You should be happy that is no longer the case.

I agree there, just underwhelming compared to current Ryzen offering. Seems like a waste of time for them, why bother?

Huh? It is flat out 10% - 13% faster out of the box than a 1600X (Depending on cooler). I know because I'v had both. When gaming I get on average 4.15 GHz all cores on my 2600X without overclocking. My 1600X out of the box was 3.7 GHz all cores. Highest I could get out of the 1600X was 3.9 GHz (the case for many users). So the out of box performance of the 2600X is almost 7% compared against an overclocked 1600X. That to me is worthwhile and a decent refresh. Add the fact that they included a cooler for less than what the 1600X was sold for? Icing. And then there is the memory controller. I went from 2933 CL16 stable on my RAM to instant 3200 CL14. Nice. Getting the most of the RAM too. If they do another release and it is 10% - 15% faster again and another IPC boost? Intel will have their hands full.
 
Back