• 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.

Ryzen "2" ?

DragonQ said:
How did you end up with a bad flash in the first place? Never had that happen.
Click to expand...

Wellll, in my case admittedly, it was back in the days we had to use floppies to flash the BIOS and I used a very old, much-overwritten disc that just happened to be the closest to hand at the time.

That sinking feeling as the update bar stopped and the minutes ground on, with the dawning suspicion... and then inevitable realisation... that it had ballsed up. :(

It still makes me nervous about BIOS updates, though :)
 
AndreiD said:
I'd say only update the BIOS if you're experiencing issues. If you're fine and stable then there's not really much of a point in upgrading, maybe wait it out for a few months until a few more versions come out and update those.
Click to expand...

I agree 100%, and very strongly. Bios flashes can cause all sorts of issues, so only flash if you have a specific reason to, not because its "newer".

As an example on my board all the newer bios's make adjustments to the LLC voltage levels, so if you have a specific tuned system thats stable and then flash, bam it may be unstable as the voltages are out of whack, on top of that asrock also made adjustments to the voltages applied to each p-state on the cpu, the changelog simply stated "enhanced LLC state".

Also on asrock boards, default voltages in XMP mode for vccsa and vccio got bumped significantly causing people with previously stable ram to start having issues, the changelog stated "enhanced memory compatibility".

On an old board I had for my core2duo I remember when I flashed the bios and c1e completely broke, there was no documented change to c-states but newer cpu's were added to the compatibility list. Regardless c1e broke.

Always be very wary of upgrading your bios.
 
I've been doing the 'flash every BIOS version' thing since I bought 1st gen Ryzen early on and my mobo had lots of issues with early versions, but if you find a stable version, I'd recommend sticking to it. I remember an early version working out pretty well for me and kinda regretting upgrading since I can't do 3200Mhz on my RAM reliably stable anymore.
 
You can kinda do it if you have an ASUS board with USB Flashback, but not really worth the effort since 2933Mhz C14 is good enough for me and the older BIOS versions all came with their issues.
Plus I'm pretty sure some of the newer Agesa versions include a few security updates.
 
Oops, should have added both sources:
http://www.smarteranalyst.com/brief/advanced-micro-devices-amd-receives-a-buy-from-jefferies/amp/
and
https://seekingalpha.com/article/41...-amd-q2-2018-results-earnings-call-transcript

“We raise our 2019 EPS est and our price target following our meeting with management at our annual Chicago Summit. We have higher conviction that in 2H19, AMD will ship a server MPU with higher transistor density than INTC for the first time in recent history, if not ever. We see this as a foundational shift in competitive dynamics."

And from the earnings call transcript:
"So, Hans, on the timing of the 7-nanometer Ryzen, I would just keep it as it's after the 7-nanometer EPYC. So we'll launch 7-nanometer EPYC first."


They're launching with 7nm EPYC first, which is H2 2019, followed by consumer Ryzens probably not too far off. Which makes sense, data center is a much bigger & higher margin market than the consumer one and having a density advantage over Intel would be a pretty big advantage for AMD, they will most likely eat into Intel's data center marketshare.

Of course, part of this is not officially confirmed, so take it with a grain of salt.
 
AndreiD said:
Oops, should have added both sources:
http://www.smarteranalyst.com/brief/advanced-micro-devices-amd-receives-a-buy-from-jefferies/amp/
and
https://seekingalpha.com/article/41...-amd-q2-2018-results-earnings-call-transcript

“We raise our 2019 EPS est and our price target following our meeting with management at our annual Chicago Summit. We have higher conviction that in 2H19, AMD will ship a server MPU with higher transistor density than INTC for the first time in recent history, if not ever. We see this as a foundational shift in competitive dynamics."

And from the earnings call transcript:
"So, Hans, on the timing of the 7-nanometer Ryzen, I would just keep it as it's after the 7-nanometer EPYC. So we'll launch 7-nanometer EPYC first."


They're launching with 7nm EPYC first, which is H2 2019, followed by consumer Ryzens probably not too far off. Which makes sense, data center is a much bigger & higher margin market than the consumer one and having a density advantage over Intel would be a pretty big advantage for AMD, they will most likely eat into Intel's data center marketshare.

Of course, part of this is not officially confirmed, so take it with a grain of salt.
Click to expand...

Er :confused: New EPYC is due in early 2019, January or February 2019.
That density I do not know what they are talking about. Do they expect first EPYC to be with lower density and then followed by something denser or they simply state a time period when AMD will fully cover, so there are financial results covering the entire period...
 
How do you know new epic is due in early 2019? TSMC 7nm HPC hasn't even taped out. This is a lot more info on when Zen 2 should launch than anything we've seen thus far.

And they mean that Epyc Rome will use dies with higher density than Intel, which should pose a pretty significant advantage for AMD in a market that's perf/W conscious.
 
 
4K8KW10 said:
Er :confused: New EPYC is due in early 2019, January or February 2019.
That density I do not know what they are talking about. Do they expect first EPYC to be with lower density and then followed by something denser or they simply state a time period when AMD will fully cover, so there are financial results covering the entire period...
Click to expand...
Density means how many cores can be put to same size silicon area.
Modular design already allows AMD to make high core count CPUs from easier to manufacture smaller dies than Intel's huge monolithic chips.
And TSMC's 7nm node will decrease die size lot compared to current GloFo chips, or then allows more cores in same die size.
So AMD can gain really big production cost advantage over Intel.
(who is also running out of 14nm capacity)

Also datacenter/server CPUs don't need highest clock speeds (needed for consumer CPUs) so those can be made more easily during maturing phase of node.
 
If it is gonna be 2H 2019 that's pretty disappointing. It might be competing with 10nm Icelake after all....haha no that'll never happen. 10nm delayed until 2020 am I right guise?
 
Don't think we'll see 10nm Icelake in 2019, or if we do it's going to be only mobile parts towards the end of the year. AMD is going to have at least 1 quarter where they will be on a considerably higher density node than Intel, which is historic.
Intel will probably refresh their consumer & data center CPUs on 14nm next year, and 10nm for 2020 on those lines, so AMD can be very competitive depending on 7nm HPC performance.
 
You must log in or register to reply here.
Back
Top Bottom