AMD launches Zen+ 12nm Ryzen and X470 motherboards

anyone have any guesses how the zen + desktop apu will preform? or what gfx card it might be on par with?
 
I’m totally in for whatever the cheapest Ryzen 2 CPU is. I have a mining rig running 2x 1070s on a Core 2 Q6600. The PC is a piece of crap and needs to get replaced, I want a modern system and then it can double as a media PC.
 
I've been having the upgrade itch, but seeing RAM and GPU just makes me sad. Will the recent MSRP drop on Ryzen hold for a while?
 
Get the 2200 APU then you get a Vega 8 to mine too. February 12th.
feb 12th is when raven ridge releases? or just the 2200?

hopefully it's all of them because my wallets ready to spend money to build a new htpc.
 
Interested in the performance of the memory controller. My R5 1600 is sketchy with the Corsair Vengance hynix chip stuff.
 
Interested in the performance of the memory controller. My R5 1600 is sketchy with the Corsair Vengance hynix chip stuff.

2000 series mobile increased the IMC to 2993MHz. I think we can expect similar 12% improvement for 2000 series desktop.
 
The great news for enthusiasts that have already bought into AMD’s current generation platform is existing motherboards will support this processor update, as long as you have the associated BIOS. Motherboards are already being updated today for the channel (to support the Ryzen APU launch) so there should be little concern with compatibility come April.

Has there been any word on how long the new chipset (T4) is expected to have support? I'm considering making the jump to AMD for my next build and just wonder about "future proofing" with T4 vs A4.
 
Hopefully this time around they actually overclock better :D

I am hoping for better memory support + a guarantee that the SMP/binning bug is 100% fixed. I don't expect a decoupling between the fabric and the memory frequency until Zen2.
 
Benchmarks from new processor!

8bc4d35b64bdc7c4759ebb75df3d9122_1520006983_3284.jpg

8bc4d35b64bdc7c4759ebb75df3d9122_1520009482_6272.jpg

8bc4d35b64bdc7c4759ebb75df3d9122_1520009788_7563.jpg

8bc4d35b64bdc7c4759ebb75df3d9122_1520009820_9329.jpg

8bc4d35b64bdc7c4759ebb75df3d9122_1520009863_3887.jpg


Source: http://www.hwbattle.com/bbs/board.php?bo_table=hottopic&wr_id=8803
 
I'm guessing 2800x? That's 100mhz base and 250mhz boost/xfr, not quite what I was hoping for. I was hoping for a little more, like 200mhz base and 300mhz boost/xfr but it's not far off. I wasn't expecting any ipc gain, so that's welcome and coupled with the clock bumps the net is a little over the 10% I was expecting. We'll see if that's where it actually ends up when it releases.
 
looks like they were testing with ddr4 -2667 as well at least in the first two screenshots, not sure if that carried over but that seems like a pretty decent bump from a minor update
 
I don't know what you expect in a year.

Look at Intel:

2008 - Nehalem

2010 - Westmere

2011 - Sandy Bridge

2012 - Ivy Bridge

2013 - Haswell

2015 - Broadwell/Skylake

2017 - Kake Lake/Coffee Lake

________________________

Kaby Lake/Coffee Lake isn't a even a real improvement over Skylake.

We have really been stuck on Skylake since 2015.

Meanwhile, Cannon Lake, which is suppose to succeed Skylake, is stuck in development hell somewhere.


Sounds freaking delicious
 
Ten percent faster from the clock bump, just like I was expecting.

Not groundbreaking, but that bump should make it a bit more competitive with Coffee in games that don't scale beyond 4 threads.
 
yeah the more aggressive turbo should help in those games as well (if you are not Overclocking yourself) and if it can hit 4.35ghz that seems pretty impressive to me
 
not bad for a "plus" refresh. If they keep prices inline with where they are now I think these will be very competitive.

Happy with my 2400g build right now, hoping AMD can continue to refine and improve.
 
I'm guessing 2800x? That's 100mhz base and 250mhz boost/xfr, not quite what I was hoping for. I was hoping for a little more, like 200mhz base and 300mhz boost/xfr but it's not far off. I wasn't expecting any ipc gain, so that's welcome and coupled with the clock bumps the net is a little over the 10% I was expecting. We'll see if that's where it actually ends up when it releases.


4.35Ghz boost that's 350Mhz not 250Mhz
 
I was expecting 200--400MHz extra and zero IPC bump.

pretty sure if you get the former it will automatically ensure the latter, unless they are real putzes and make sure to reduce the design to hold IPC back (really not recall of any situation this was the case save for P4 or Bulldozer, but, those are a different story)

either way, any "rumor" is worth as much as a load of sand, past history has shown often times that we might see a glimmer of "truth" and at the last moment they release something even more amazing (with features not even mentioned in the so called rumours that would have been great to know)
or conversely folks make all kinds of review site posts for clickbait and when the product finally releases it ends up "seeming" a let down because folks were expecting much more.

rumors are like opinions and Aholes afterall, everyone has one ^.^
 
Since the benchmarks from HWBattle was conducted using Radeon RX Vega 64, do the Physics Score(s) change if Geforce GTX 1080/1080Ti were used instead?
 
Since the benchmarks from HWBattle was conducted using Radeon RX Vega 64, do the Physics Score(s) change if Geforce GTX 1080/1080Ti were used instead?

I wouldnt think it would make much of a difference, supposed to be all on the cpu
 
pretty sure if you get the former it will automatically ensure the latter, unless they are real putzes and make sure to reduce the design to hold IPC back (really not recall of any situation this was the case save for P4 or Bulldozer, but, those are a different story)

either way, any "rumor" is worth as much as a load of sand, past history has shown often times that we might see a glimmer of "truth" and at the last moment they release something even more amazing (with features not even mentioned in the so called rumours that would have been great to know)
or conversely folks make all kinds of review site posts for clickbait and when the product finally releases it ends up "seeming" a let down because folks were expecting much more.

rumors are like opinions and Aholes afterall, everyone has one ^.^

I was expecting 200--400MHz because that was my guess for a relabeled 14nm+. I was expecting zero IPC gain per core because the core is the same. Any IPC chip gain would come from L3/IF/IMC clocked higher.

juanrga is hoping the new chips will be a complete failure... But he's not an idiot, and his expectations are different.

My expectations were based in tech estimations, as they always are.
 
Last edited:
Any word on price? The 2800X is not going to launch at $500 again, right?

I am guessing that AMD would drop an SKU so that the 8-core Pinnacle Ridge would only have 2 SKUs.

Ryzen 7 2700X (or whatever the top SKU is called) will probably launch at the same price as Core i7-8700K's or slightly lower.
 
Well, that looks like it's almost retirement time for my 2500k :)
Those leaks are more than I think we could of hoped for and people complaining at that price point (vs competition) are just been picky.

Yes it might not be an upgrade VS last gen Intel but for 2500K ... (and we're a bunch !!) it will be.
 
pretty sure if you get the former it will automatically ensure the latter, unless they are real putzes and make sure to reduce the design to hold IPC back (really not recall of any situation this was the case save for P4 or Bulldozer, but, those are a different story)

either way, any "rumor" is worth as much as a load of sand, past history has shown often times that we might see a glimmer of "truth" and at the last moment they release something even more amazing (with features not even mentioned in the so called rumours that would have been great to know)
or conversely folks make all kinds of review site posts for clickbait and when the product finally releases it ends up "seeming" a let down because folks were expecting much more.

rumors are like opinions and Aholes afterall, everyone has one ^.^
IPC is instructions per cycle, it doesn't change without an architectural change. Increasing clock increases performance, but not IPC.
 
IPC is instructions per cycle, it doesn't change without an architectural change. Increasing clock increases performance, but not IPC.
Output IPC can change with something like reorganised caches, different uncore clocks, things like that. BUT the change in IPC will be very minor.
 
Back
Top