It's not really 12 nm, it's basically a refresh of 14 nm (Intel calls their equivalent "14nm+"). We don't know what difference there will be. Could be IPC, could be clock speeds, could be both. I think a clock speed bump is more likely since node improvements typically give you that, although we'd be talking about 200-300 MHz at most, IMO. AMD might make some tweaks to slightly boost IPC but significant IPC improvements probably won't come until the 7 nm version - that'll allow them to add more transistors (and cores). Alternatively, if they can get the infinity fabric to just run quicker (e.g. DDR like the RAM) that'd provide a decent boost in some circumstances too.
-
Competitor rules
Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.
*** AMD "Zen" thread (inc AM4/APU discussion) ***
- Thread starter Boomstick777
- Start date
Soldato
Shame. An IPC improvement would have put them almost in the same realm as intel's latest.
Soldato
Its hard to say since GF 12NM despite being their latest 14NM process actually advertises both density AND performance improvements.
So what I do expect is some fixes here and there which will improve IPC and a few 100mhz increase in maximum clockspeed.
Remember when AMD moved from Bulldozer to Piledriver,they got a decent uplift in IPC and clockspeeds via some smallish fixes and a few 100mhz extra base clockspeeds,but used the same 32NM process.
IIRC,both the die sizes for the FX8150 and FX8350 were virtually the same.
So what I do expect is some fixes here and there which will improve IPC and a few 100mhz increase in maximum clockspeed.
Remember when AMD moved from Bulldozer to Piledriver,they got a decent uplift in IPC and clockspeeds via some smallish fixes and a few 100mhz extra base clockspeeds,but used the same 32NM process.
IIRC,both the die sizes for the FX8150 and FX8350 were virtually the same.
Soldato
Will be interesting to see what they can come up with, maeby their first plan didnt even include 14+nm ("12nm") prosess. Wasnt it stated by someone in AMD, they already had improvements made up to Zen but they just hadnt enough time to implement those? Could not find the link, but if they could add some improvement from technical side + added performance from better node things could get really interesting.
Soldato
Secretly, I hope they have been binning chips on the current process and release a special xt range at 4.5ghz lol.
Soldato
Nice I think that's decided it for me
Caporegime
AMD's PR again, hope they're able to put their money where their mouth is as regards upgrades then.
Soldato
Hardly difficult for them not to make the newer processors backwards compatible, it's not like Intel where they've had to add additional power delivery pins on the 1151 socket to accommodate the 6 core CPU's, which is clearly something they never intended to happen on that socket. AMD started off running, and aren't going to slow down with Ryzen moving forward, they've will be going down in power draw, and current delivery requirements, not the other way.
There's no big changes due for 2 years to the mainboard layout, for PCI-E 4.0/5.0, and then DDR5, but ultimately the number of AM4 platforms out there by then will mean not chosing to support their newly gained market share could damage them seriously, with this kind of marketing.
Caporegime
AMD's and execution and all around marketing etc is generally dreadful. Vega was a sham. Zen launched well, but maintaining that is where AMD's recent history is sketchy. If zen+ came out and it was a SB to IB type of gain overall, then it'd be pretty lackluster for example, regardless of the second iteration of chip.
Hopefully AMD can follow up with something that'll clock higher and perform faster per core. Being able to reach 4.6+ will give decent gains for example etc.
Associate
I had a moment of madness yesterday with Ryzen, tried running cinebench for reference and crashed during test. I had the same settings since i built the system and it was rock solid. Computer restarted and my CPU was locked to 1,5Ghz no matter what I did. So I had to flash BIOS, downloaded the latest one with improvements to Agesa 1.0.06B. Thought It would be better but it turned out that I could not even boot PC in standard settings i.e 2133Mhz, constant boot loop and errors. When I finally did it with some silly speed 2400Mhz got BSOD's with "attempted to write to read only memory". In the end I had to settle for some old BIOS which worked without hassle with rated memory speeds. I don't know what ASUS is doing with their updates but it looks like they making things worst.
Almost no home user is going to benefit from PCIe 4.0 anyway. I highly doubt any GPUs will be around at the time that saturate PCIe 3.0. DDR4 isn't that old so yeah, DDR5 is the only excuse to upgrade socket really.
It's not clear whether AMD s following Intel's tick-tock model or if they will do both die shrink and architecture improvements at the same time with Zen 2. Sandy Bridge to Ivy Bridge was pretty much just a die shrink but if they can replicate Core -> Core 2, or Nehalem -> Sandy Bridge that'd be pretty cool.
Last edited:
Permabanned
I'm finally ready to build my Ryzen rig again, I ended up getting a 1600x rather than an 8 core as I now have an excuse to upgrade when the Ryzen refresh releases next year. As usual I bought all the parts from OCUK except the motherboard as the one I wanted wasn't in stock, I ordered that from the river because I have free next day delivery but when the delivery turned up I opened the door to find that I'd been sent an unpackaged motherboard box that wasn't even sealed, All they'd done was stick a sticker on the motherboard box with my name and address on it. Talk about shoddy, It's a great example of why I prefer to shop with OCUK.
I think I found my problem from last time, I just downloaded the lastest bio's update for my new motherboard and it mentions having a fix for the Intel 600p M.2 drive issue which is the drive I have. I could have done with knowing this a month ago.
I think I found my problem from last time, I just downloaded the lastest bio's update for my new motherboard and it mentions having a fix for the Intel 600p M.2 drive issue which is the drive I have. I could have done with knowing this a month ago.
Last edited:
Just pulled the trigger on a R5 1600! Had this Crosshair sitting in its box for over a month and got some DDR4 at a bargain price over six months ago. It's rated at 3866mhz so should be good for 3200mhz on Ryzen (I hope)
Just waiting to order a Samsung 960 Evo 250GB then I'm all set
Just waiting to order a Samsung 960 Evo 250GB then I'm all set
Caporegime
Got the Crosshair and DDR4 for a very good price and my main game I play is CSGO. So don't need multi-threaded power.... I know Intel is king of singlethread yadda yadda, but for the price I scored all this at, I can't complain - £350 all in
Caporegime
Not sure king is fair comparison. The 7640X is maybe a prince.
Soldato
The next AGESA release is implementing some big changes:
http://www.overclock.net/t/1624603/rog-crosshair-vi-overclocking-thread/28240#post_26391661
https://www.reddit.com/r/Amd/comments/76jzl4/elmorrog_rep_on_agesa_1007/
Also,in the Reddit thread there is noise about potential performance improvements.
Edit!!
I saw this mentioned on AT forums:
http://www8.hp.com/h20195/v2/GetPDF.aspx/c05652570.pdf
Its the spec sheet for the HP ENVY x360 15-bq101na,which appears to be the first Raven Ridge laptop.
http://www.overclock.net/t/1624603/rog-crosshair-vi-overclocking-thread/28240#post_26391661
https://www.reddit.com/r/Amd/comments/76jzl4/elmorrog_rep_on_agesa_1007/
Asus employee said:
Also,in the Reddit thread there is noise about potential performance improvements.
Edit!!
I saw this mentioned on AT forums:
http://www8.hp.com/h20195/v2/GetPDF.aspx/c05652570.pdf
Its the spec sheet for the HP ENVY x360 15-bq101na,which appears to be the first Raven Ridge laptop.
Last edited:
Associate
Well Agesa 1.0.06B on my board would allow me to post under any condition. Stock CPU & Stock Ram, got a constant boot loops.