Read more.Quote:
And some benchmark scores for Ryzen 4000-series APUs (Renoir, Zen 2) have shown up.
Printable View
Read more.Quote:
And some benchmark scores for Ryzen 4000-series APUs (Renoir, Zen 2) have shown up.
can I just check that's correct? The slide later in the article says ryzen 3 at 7nm and ryzen 4 at 5nm. Due to their silly naming the ryzen 3 cpus are the 4000 series, so do you mean the 5000 series ryzen 4 will be 5nm, or definitely that the upcoming 4000 due out later this year will actually be 5nm?Quote:
Originally Posted by hexus
If so that upgrade path for a B450 mobo or better will be incredible. 5nm on the AM4 socket!!!
are you all aware intel can simply offload some of its chips to TSMC or even Samsung?
Then why are they still using 14nm+++++ ? :)Quote:
Originally Posted by lumireleon
For the 2 people above Intel would just have to close fabs then... can't see that happening as they have always stated they will never be a fabless company.
They have loads of capacity just on a process that's years behind
Volume manufacturing in Q4 and launching a product is a rather tight timescale
The slide in the article might be outdated, so we might see ryzen 4000 on 5nmQuote:
Originally Posted by ik9000
The DigiTimes piece indicates that the Zen 3 processors will be N5P but we will have to see if its true. As for desktop Ryzen 5000 (Zen 4) CPUs, that is even further away and even less certain. Remember this roadmap slide from late 2019 (and most others) has small print saying "roadmaps subject to change" - though we are more used to things changing in a negative way (delays etc).Quote:
Originally Posted by ik9000
I'd imagine 5000 series will support DDR5 so AM5 will be requiredQuote:
Originally Posted by mtyson
"Simply offload"Quote:
Originally Posted by lumireleon
Taking an existing design and porting it to a different process is a very expensive, time consuming excersise. Where Intel would be "just another customer to their fully committed wafer output" to TSMC.
I think you are unaware of the hundreds of millions of dollars it would cost intel to do that for a single CPU die design.
It's unlikely as other sources have already stated that the design process for 5nm is different to 7nm. I'd say 7nm on AM4 socket for this year and most of next year and then 5nm on a new AM5 socket then end of next year.
Anything sooner is a bonus and would be kicking Intel when it's already down...
What's the latest on that? I seem to remember some talk about it being tested or used in the enterprise space, all be it at rather low speeds, but what's the current thinking on consumer availability?Quote:
Originally Posted by 3dcandy
Don't know about the forecasts (and I haven't paid them any attention), but going by historic switches prices will be very high for the first 'few' years. Vaguely recall that DDR3>>DDR4 took 4+ years.Quote:
Originally Posted by Corky34
So unless AMD want to get bad reputation like Intel did during the RAMBUS-down-your-throat years, I doubt they'd force DDR5 too soon.
Since the IO is a separate chip now, it is actually easy for AMD to release Zen4 with support for both.
Whether they can do Zen4-DDR4 for AM4 with Zen4-DDR5 later on AM5 is a big question. I'd say it's unlikely, but if they don't then whatever DDR4 boards are made for Zen4 will be short-lived.
Server's are less price sensitive so I'd expect them to once again transition far earlier. Mobile is unknown. If DDR5 is far more power efficient then they might transition before desktop.
Found an image from 2015 showing how much desktop lagged behind server:
https://i.imgur.com/EaTtOXc.png
Of course, the memory manufacturers (i.e. Samsung since the other two are far smaller), may try to push towards DDR5 to try and increase margins. The usual thing has been that the new standard gets better density which is usually most important in servers.
Skylake desktop only launched halfway through Q3 15, so it is a bit chicken and egg - no desktop systems will ship with DDR5 until we get components that support it
Zen+ was a process shift. AMD were probably already working on a Zen3+ on 5nm. Given the 5nm process is ahead of plan, they could validate that earlier than initially planned meaning release in 2020 possible rather than 2021 as planned. Add in Zen3 being delayed by Covid), if they'd not started proper bulk manufacturing of Zen3 on 7nm, the COVID Zen3 delay combined with the 5nm process early arrival could mean this happens.
Utterly crazy if it does.
Re:Corky and kompukare - until the last few month I'd have agreed, but I've seen AMD try and shift forward much quicker than before.
DDR5 is in testing stages, but as AMD have shown with the move to pcie gen4 they can get a load of traction with beating Intel to the post....
DDR5 and DDR4 support is another thing they could do so maybe it is another reason they went for an i/o hub....
Who knows... I'm just hoping they may push it forward
Was just about to post something along the lines of this myself. Even for the title story, it's really not a case of AMD just making a last-minute decision to click go on a completely new node. Usually I'd just dismiss this out of hand as a misunderstanding, but AMD have been somewhat surprisingly lately so I guess there's an outside chance. If so, they would have had to be planning this for quite a long time, and that possibly means simultaneously designing Zen3 on both 5nm and 7nm - sounds mighty resource and cost intensive to me! I think it was one of Anandtech's writers suggested perhaps they might have one for Ryzen, another for EPYC if it does actually happen, but that would seem to negate some the advantage of using chiplets across both markets.Quote:
Originally Posted by badass
Lisa Su was talking about mobile,so I think this more to do with the collaboration they have with Samsung,or a mobile part. Roadmaps,even published this year show Zen3 is on 7NM,and not even 7NM EUV.
Yep, I could be wrong but I think this could be the latest 'official' information regarding the nodes used: https://www.anandtech.com/show/15589...-not-specified
Is Zen 3 on 5nm+? At first you think there is no way this is true simply because we have all had 7nm cemented in your mines for the last 4 years! But there are things that we know that could back it up things like:
- 7nm is hugely over subscribed with a long lead time for production. We have RDNA 2, Ampere, PS5, XSX, current Zen 2 chips, plus a load of other stuff all fighting aggressively for capacity. If TSMC thinks 5nm is ready and it has capacity it would make sense to move some customers over to this node so TSMC can maximise production.
- So then the question is, does 5nm have spare capacity? We also know the answer to this and it's a massive YES after Trump's demanded that TSMC stopped working with Huawei which until this happened was TSMC's second largest customer so we know that TSMC must have a huge gaping hole in 5nm orders with Huawei gone and so AMD could easily and happily step in here.
- So then it's a question of, is it possible to got from 7nm to 5nm and how long would it take? This requires a more technical breakdown than I can offer but we do know that many products get a 'die shrink' from one node to another so it certainly possible, as for how long? Well you would think it would take a good 6 months. So have AMD done anything recently to indicate to us that they might actually be stalling for a little more time? Well the answer is another big YES once again, they are bringing an out of the blue but half-hearted Zen 2 refresh on a few key processors to seemingly 'tie us over for a while'. This refresh would look very odd if we were going to get Zen 3 in September but not so odd if it now wasn't coming until early 2021 as they need time to move Zen 3 from 7nm to 5nm.
- Finally, we know from TSMC orders that AMD has ordered a lot of 5nm wafers for production this year and so clearly been working with 5nm for quite a while and full expected to be using 5nm this year, if it's not Zen 3 then what is it?
As for the Warhol and Raphael leaked slide, this could easily just be an old slide that hadn't been updated since Huawei got kicked out and TSMC needed a big 5nm order to replace them.
- On a separate note, it's really important to emphasise that we are only talking about Vermeer being 5nm+ and NOT Milan. This helps AMD make enough chiplets for both the server market and desktop market plus it also means none of AMD's slides or recent presentations were wrong or inaccurate as Zen 3 'Milan' is still 7nm and crucially that Milan won't be delayed which would get shareholders annoyed simply because they don't like last minute changes even when they are possibly for the best.
- One additional point which doesn't seem to be getting any coverage is that it looks like Zen 4 has been delayed pretty badly, Zen 4 was going to be the first 5nm CPU and required AM5. But if Zen 4 isn't coming until well into 2022 or later then AMD needed to change it's 5nm plan anyway. There appears to be mounting evidence that we are going to get a Zen 3+ before Zen 4. Why do I believe this?
- Firstly, the Gigabyte leaked motherboard support slide shows "future CPUs being Vermeer & ...." which is clearly indicates there is another CPU line coming that will be compatible with x570 mobos.
- Secondly, when AMD back-tracked on ditching 400 series compatibility with Zen 3 they issued a statement that said "This is the final pathway AMD can enable for 400 Series motherboards to add new CPU support. CPU releases beyond Zen 3 will require a newer motherboard." Which again, unless there is another AM4 line of CPUs coming after Vermeer then AMD didn't need to make this statement as we would then be on AM5 and even the most die-hard AMD fans can't complain about no comparability here!
TSM are also getting a nice financial leg up from Trump, to fund their promptly agreed, stateside fab.
You do make some good points but this is the crux of the matter really. TSMC can't just move customers to a new node as we've mentioned earlier in the thread (and you have mentioned yourself). It would have to be a decision made by both AMD and TSMC, and would not be without risk and cost. With the increased complexity and design rules of current nodes it's not quite the same of the half-node type die shrinks we used to see quite frequently on mid-cycle product refreshes. Moving from 7nm to 5nm at TSMC would require significant reworking of the processor and would need to bring some significant advantages to be worthwhile. That may well be the case, but it does seem like a long shot?Quote:
Originally Posted by Oberoth
Really? Granted I've not been looking for it, but I've not read anything since the very early, preliminary plans to build a relatively small (by TSMC standards) 5nm fab due to be finished some time around 2024. Not sure why that would impact this story given TSMC are already a very wealthy company?Quote:
Originally Posted by msroadkill612
subsidization
What the rich people use to get richer and have the tax payer fund it that way too.
Much used in my country and it is one of the main reasons i am embarrassed to be a Dane.
They will also use other names for it, but it is what it is and you the little guy will be the one paying for it, and little Danes pay for a lot of stuff, and they pay a huge price for it too.
And the very worst part is when companies from another country come her and grab money too, some do that a lot and dont pay a dime in tax here.
That 'load of Capacity' that Intel has at 14nm......++ etc,Quote:
Originally Posted by 3dcandy
is becoming more and more pointless as their TDP's start
to skyrocket for ever shrinking performance gains.
Indeed but that's what they've got to work with...Quote:
Originally Posted by albert89
They will be on 10nm sometime soon(ish) which actually should give them a boost. To be fair gamers don't care so much about the extra TDP's do they which is really the only place I see Intel winning overall when you take off price/performance/lower tdp from AMD
Intel could outsource non-CPU chips eg. MB chipsets.
Why would they want to do that?
It's not to say Intel haven't outsourced some things e.g. when they were making some mobile phone SoCs, but moving designs to another fab costs a lot of money - there would need to be a good reason to do it. Capacity is one possible reason and I believe there were some rumours about this, but I believe they ended up going back to 22nm for some of their entry level parts instead. Not certain on that though.
Intel have spare capacity? I'd heard they were in meltdown.Quote:
Originally Posted by albert89
sorry couldn't resist. I'll get my coat...
I'm still not seeing any evidence that 10nm will be any use for desktop, only for laptop parts.Quote:
Originally Posted by 3dcandy
The prediction at SemiAccurate ages ago that Intel would only leap forward on server & desktop at 7nm is looking spot on.
Nope that is true - but then again they are years behind on 7nm as well...Quote:
Originally Posted by DanceswithUnix
This has all the beginnings of a perfect bad storm for Intel
We can expect Intel's 7nm to be more like TSMC's 5nm, so there is a chance at that point for Intel to catch up. Still a while off, but soon enough to make trying to fix 10nm for performance parts not worthwhile as it would take longer and just divert resources.Quote:
Originally Posted by 3dcandy
I remember this about Intel outsourcing production to TSMC:Quote:
Originally Posted by watercooled
https://www.techpowerup.com/249039/i...cation-to-tsmc
However,did it ever happen??
I've no idea to be honest, but if I had to guess I'd say probably not. I reckon CPUs would have been the last thing they would have considered moving if anything - while porting a chipset still seems quite anti-Intel, it wouldn't be nearly as complex or costly, and would avoid having potentially differently-performing products.
Edit: Wow 2018! I didn't realise it was that long ago!
These issues have been with Intel for years haven't they. As cock ups go it's as monumental as Nokia dropping the ball and getting soldQuote:
Originally Posted by watercooled
;)
Only the mobile phone bit - Nokia is still a major telecoms company, and one of the few major names in 5G networks.
A lot of the big telecoms companies ended up selling or ending their mobile phone businesses in one way or another - Nokia, Ericsson, Motorola, Alcatel, etc.
Well there's still Huawei but the US government seem hell-bent on removing that competition from the market. But they've made some of the most competent phones (including their own SoCs) on the market over the past few years, especially in the more budget-oriented segment where some of the other big names seem to put some effort into making that sector unnecessarily poor to up-sell people.
that was either an act of major negligence almost to the point of almost criminal ineptitude by the CEO, or else he did it deliberately as some kind of corporate sabotage but again wtf? Either he knew what he was doing and is therefore an idiot, or he didn't and therefore an idiot. IIRC MS hired him though after the takeover so they clearly didn't think he was an idiot or a liability. Draw your own conclusions. Great article on it here: https://communities-dominate.blogs.c...-pictures.htmlQuote:
Originally Posted by 3dcandy
I do wonder if they might get back into the mobile business again, perhaps with an investment of some sort into HMD?
Had Windows Phone been an emerging and successful platform then he could have been onto a huge winner. But Nokia backed Microsoft at the point that Microsoft went completely nuts and seemed to be busy trashing their own phone OS and hacking off their customers.Quote:
Originally Posted by ik9000
If Elop had jumped from their burning platform to Android, I expect it would have worked.
At the time didn't they basically say something along the lines of "we don't want to have compete with all those cheap Android also-rans". Which is sort of saying "we were to late to move into the market and are now not able to compete".Quote:
Originally Posted by DanceswithUnix
Very much like Apple in 90s. From the Mac II until the mid 90s they had better hardware and operating system than PCs but never wanted to licence. When they finally allowed PPC third-party clones it was too late and Power Computing just ended up taking shares from Apple rather appealing to PC buyers.
Nokia crawling to Windows Mobile because they though they couldn't compete with generic Android was similar.
Very much like now AMD will push out zen4 on 7nm ..(its too late to change) BUT they will hold in reserve a refresh to 5NM .. they will wait to see what Intel brings to the table .. two pair or a straight flush. If its needed then the zen3 "XT" Cpus will appear, for a mid life refresh, thatw ill be foillowed by ZEN4 which will have the much vaunted additions
Ryzen 4000 is Zen3, not Zen4. And the current Ryzen 3000 is Zen2.Quote:
Originally Posted by persimmon
While I think you are right, I personally can't see why generic Windows Phone was any better than generic Android. If they wanted a small ecosystem they should have stayed on their burning platform, at least they had control over that. I have seen people who adopt Microsoft platforms crucified over an over in the decades I have been in IT as Microsoft adopted and then abandoned the likes of MIPS and SuperH leaving "partners" hanging out to dry.Quote:
Originally Posted by kompukare
exactly. Which genius though that was a good idea?Quote:
Originally Posted by watercooled
Ryzen 2000 was still based on the Zen microarchitecture, so the model name advanced by one but the core name didn't.
I know. What I'm saying is all they needed to do was call it the 1000+ and everything would have been ok, and future generations would nicely make sense.Quote:
Originally Posted by watercooled
Mate, you can't use "Ryzen 3" to mean Zen 3. Those names are referring to totally different aspects of the CPUs, and phrasing it as you have in your post is completely incorrect and confusing as hell.Quote:
Originally Posted by ik9000
Zen = Ryzen 1000 Series (unsure about APUs)
Zen + = Ryzen 2000 Series (unsure about APUs)
Zen 2 = Ryzen 3000 Series (except APUs)
ZEN 3 = Ryzen 4000 Series (except APUs)
Ryzen 3 = Lowest tier Ryzen CPUs
Ryzen 5 = Lower-Mid tier Ryzen CPUs
Ryzen 7 = Upper-mid tier Ryzen CPUs
Ryzen 9 = Top tier Ryzen CPUs
I think the reason the article contradicts the image posted, is because the image posted was showing the planned timeline, where the article is saying that 5nm may be coming so much sooner, that Zen 3 might/will actually be coming on 5nm instead of 7nm.
Where in my post did I say ryzen 3 means zen 3? Maybe try reading what you're quoting.Quote:
Originally Posted by Noxious89123
To be fair it's confusing as hell, not your post, well sort of your post but the naming of them, it's the whole APU's using different series' than the desktop parts that throw me, i get that the series' are running one step ahead of the architecture naming but then when you look at APU's they don't, or is it they run a step behind, or in front...oh i give up. :)Quote:
Originally Posted by ik9000
I'm not making the 'someone else does it so it's good' argument, but it's not unusual for this to happen. Multiple core generations have appeared in a given CPU generation for Intel, for example the i7 6700k came out before the i7 6800k. Both use the same microarchitecture right? Wrong! So the 6800 must use the newer core? Wrong again! 6700k came out first using Skylake, then the 6800k came out the following year using the older Broadwell microarchitecture.
And now we have several generations and CPU codenames all using Skylake cores, the first of which was called Skylake (when Intel used the same name for core and CPU), followed by Coffee Lake, Coffee Lake Refresh, and Comet Lake.
Intel's naming has also confused the hell out of me for decades, trying to decipher their product stack is like trying to read the Rosetta stone, I'm sure it all makes sense to someone.
It is the case that you have to keep up to date with the naming to know what you're getting as there's no real logic to it.
Agreed. At least Nvidia sorted out their confusing names of old with their revamp to do away with all those GT, GTS, GTX variants and criss-crossing tier numbers that weren't in sync GTX465<GTX460<GTX470 etc. Now it's simply the GTX and RTX. No wait, the GTX-Ti, GTX-Super, GTX-Ti Super, RTX Titan and... oh ffs who comes up with this crap? "I'd like the RTX shizzle with cheese whizz and sprinkles of platinum oxide if you please good sir! What's that? You're out of stock but have the RTS SuperTitanicMobyDick? Is that really as good?"Quote:
Originally Posted by watercooled
We still have GT, just on the 1030 thoughQuote:
Originally Posted by ik9000
Well this puts paid to some rumours, including this one: https://www.techpowerup.com/268650/a...launch-in-2020