I came here wondering the same thing, not a daft question at all! Glad you asked it
-
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 2 (Ryzen 3000) - *** NO COMPETITOR HINTING ***
- Thread starter Jamin280672
- Start date
I came here wondering the same thing, not a daft question at all! Glad you asked it
This is the main thing that's making me wary. Not that I won't get 4.0 (not sure if we'll ever see many 4.0 cards), but that it'll be buggy as it's the first time out for a new tech, in the way PCI cards hanging off early pcie bridge chips were often flaky.
Associate
Wilful ignorance is painful to watch, especially when you display it obviously. The 10c Comet Lake CPU is just a 9900K with another two cores. That means higher power draw or lower clock speeds. Higher temps. Larger VRM requirement. Yet you seem to think it's going to be enough to compete with the 12c Ryzen 3xxx? Sorry but it doesn't look like it. Intel are 3yrs behind at minimum at the moment. The 8xxx series were rushed to release 6 months early due to the first Ryzen 1xxx release worries. Currently it's Intel responding to AMD's lead, not AMD responding to Intel's lead. Which is why I explained that Intel won't be able to have proper competitive CPU's until 21/22 when they finally get to their 7nm process. Even if it's better than TSMC's 5nm process Intel are in for a hard slog to keep up.
I'm not being brand loyal in this argument either. It's just fact that Intel are on the back foot and are responding to AMD's lead. That they still command stupid prices for their tech and have a product shortage doesn't alter that they are losing market share and mindshare at the moment hand over fist to AMD. Which is a great thing competition wise.
ignorance there is no benchmarks yet intel are behind now somehow. it is painful to watch. just speculation at the moment so what i said is true. once we have real data then we can see if amd are forging ahead or they just marketing hype and hoping it works.
also if amd does make a faster cpu range for considerable time dont think retailers are going to give them to you. they will be priced probably similar to intel.
also if amd does make a faster cpu range for considerable time dont think retailers are going to give them to you. they will be priced probably similar to intel.
Associate
More willful ignorance. I wasn't claiming any of that ****. You post more strawman arguments in response. You are literally fake news. It's quite telling how you never actually respond to any points presented, just put up more strawmen and burn them instead. FFS mate.
I don't care what retailers do. People will find the best deal as per usual market mechanics. You are literally positing that the cost of the product will go up if it's better, I mean you could be a first year business student and work that out. The speculation on Comet Lake is based off the last 10yrs of Intel being unable to innovate. It's clearly a 9900K with 2 extra cores. We can more accurately extrapolate the potential of that CPU and it's insane power needs vs the Ryzen 3xxx series which we are now pretty well versed in how they are likely to perform.
The difference between your speculation and most of the other people in this thread, is yours has consistently been full of ****.
Unless AMD has openly lied about the IPC increase, and all their partners have been complicit, 7nm is a huge blowout and clocks are terrible (it's done wonders for Vega clocks which wasn't even originally designed to run on it), the new I/O arrangement is borked, the expected improvements in memory latency and speeds are a lie, and expected improvements to IF don't materialise ... then we pretty much have to see AMD overtake Intel in the vast majority of performance metrics, and at significantly lower power draw, at higher yields, and at much lower costs (to AMD - retail price obviously remain to be seen). Just how much better they are remains to be seen. It also remains to be seen whether Intel will have any answer at all before Zen 3 is already on the shelves.
Soldato
Your literally have to be an idiot to think that any company could take and older well developed process,that has very little scope left for improvement and think they could add 25% more cores with no detrimental effect on the power draw or a reduction in clock speed.
We don't live in Harry Potter world, and you can't fudge the physics involved, yes can plaster over it with marketing and brand loyalty or mindshare, but again only an idiot would be unable to extrapolate the likely outcome.
We don't live in Harry Potter world, and you can't fudge the physics involved, yes can plaster over it with marketing and brand loyalty or mindshare, but again only an idiot would be unable to extrapolate the likely outcome.
What remains to be seen is if the mainstream retail and business community pick up Ryzen and the Server community actually pick up Epyc in volume. There is still an overwhelming brand position to overcome.
I think that for enthusiasts like on here we are going to see a large movement, both the brown sticky stuff from Intel loyalists and in sales for the majority. However, to actually do that it has to beat out Intel in gaming benchmarks - and we don't know if it is definitely going to do that yet.
It is going to be really interesting to follow the financial results for Q3 through to Q1 2020 and see how much real market adoption has translated into sales. For all the conversations about where AMD figures on top selling chips on the Rainforest the financial numbers don't show a massive uplift yet. What they are showing is excellent margins and AMD starting to make a decent profit, which is awesome.
I think that for enthusiasts like on here we are going to see a large movement, both the brown sticky stuff from Intel loyalists and in sales for the majority. However, to actually do that it has to beat out Intel in gaming benchmarks - and we don't know if it is definitely going to do that yet.
It is going to be really interesting to follow the financial results for Q3 through to Q1 2020 and see how much real market adoption has translated into sales. For all the conversations about where AMD figures on top selling chips on the Rainforest the financial numbers don't show a massive uplift yet. What they are showing is excellent margins and AMD starting to make a decent profit, which is awesome.
Associate
- Joined
- 27 Apr 2007
- Posts
- 963
The irony being that Intel still have the better IPC and clock speeds after such a terrible 10 years so what does that say about the competition?
AMD have an incredible opportunity right now to really pull ahead and their future looks really bright.
But rewriting history to fit your personal bias is just silly.
Both companies have had periods of major dominance over my 25 years of PC building and right now we clearly seem to be moving into a phase where AMD's designs are significantly better.
But at the same time they have been so bad for the majority of this decade that they flirted with bankruptcy.
Why not keep some perspective? Don't be a hater as it blinds you to the truth and the truth is more important than any CPU that will ever be built.
Associate
Please this has been debunked many times. Ryzen 1xxx got onto IPC parity with Kaby Lake on an inferior frequency node. Ryzen 2xxx got onto IPC parity with Coffee Lake with again an inferior frequency node. With the increase in multi-core requirements, having the highest clock speeds for the average user is less important. The single core clock speeds are purely needed for gaming for the most part. So Ryzen got popular because it offered less bottlenecks for multi-core use.
You think I have a personal bias on this? I had a 3930K for 6 yrs prior to my ThreadRipper. At the time the only thing really better than that would have been the Xeon 1680v2 if I could have got hold of one.
I'm not a hater of Intel. I don't see the benefit in their products for the average user. I see very niche use cases for them currently with the vast majority not needing 5+GHz single core performance. I see AMD likely bringing out a 3700X with a 5GHz boost and 12cores. Something Intel cannot currently compete with as it takes away each advantage Intel has left over AMD. Even if Comet Lake comes out with mindblowing performance it's not going to do it at the same power envelope that AMD can. They may claim 125W TDP for it but it'll pull north of 250W without even thinking about it for an all core load.
Hell I could have bought a new CPU the entire time I had my 3930K. At no point did I need to do so. None of them were a viable upgrade for the money. Even up to Coffee Lake.
Associate
- Joined
- 27 Apr 2007
- Posts
- 963
You completely failed to address the above quote I highlighted from you.
When I see people reduced to berating a company which has totally dominated for ten years due to lack of innovation it makes me suspicious.
Do you buy innovations or the best product available?
AMD's products released in the last 2 years have been very innovative and also very good and the immediate future looks especially good.
I don't care how much glue they use or how many chip-lets as I look at the real world data not the innovations.
Whilst I have ignored AMD for over decade regardless of their innovations I will now consider them again purely based on performance and other pragmatic data.
If one wants to see the truth one needs to be pragmatic and unbiased and ignore frivolities.
Soldato
Ryzen 1 was nowhere near kaby...... it was much closer to haswell.
Ryzen 2 is also nowhere near coffeelake.
Where do some of you get these "facts" from?
Or are we using cinebench as the be and end all?
Ryzen 2 is also nowhere near coffeelake.
Where do some of you get these "facts" from?
Or are we using cinebench as the be and end all?
Soldato
10 years of dominance during which they did what? Sandy lake onwards, bar an IPC boost around Haswell saw each generation still with very similar performance and power use for anyone who actually made use of the "k" chips.
We had 4.8ghz+ Sandy bridge from Jan 2011, it took till 6th gen (Including Haswell's IPC uplift and Kaby's node shrink) to have something that genuinely beat it rather than just another 200mhz added to otherwise a very similar chip.
AMD were in a terrible place the whole time. Intel's dominance clearly wasn't from a position of pushing and innovation, it was from an absence of competition. They spent 6 years (until Ryzen release in Feb 2017) doing a wee clock speed increase with a very nicely segmented processor market that I'm sure turned them a VERY pretty penny. All that time and here we are, 2019, they're still pushing exactly the same architecture with ONE node shrink.
Where the hell are you guys spinning this tale of amazing innovation and progress from?!
You can't talk about ignoring bias and heap praise on Intel for doing next to nothing for close to a decade because their rival had properly tanked (generally due to Intel dirty tricks but AMD made some seriously bad choices too) that's blatant hypocrisy.
Come 2017 and Ryzen... ohhh, Intel suddenly have a 6 core chip (on the same node). Come 2018 oooh, Intel can do an 8 core chip too (on the same node, now at furnace levels of power/heat).
Where's the innovation in all this? Dominance, absolutely. It completely wasn't of Intel's own making though, the parts of it that were, were dirty tricks and backroom deals. That's absolutely devoid of any reason for celebration.
AMD's come back, they came in with a poorer (but smaller) litho, 2 gens later they've changed architecture already. Chiplets is clearly where Intel will follow. That they do nothing but minor clock bumps for nearly 10 years followed by cribbing the design of their competition. Clearly dominant and innovative, yes.
We had 4.8ghz+ Sandy bridge from Jan 2011, it took till 6th gen (Including Haswell's IPC uplift and Kaby's node shrink) to have something that genuinely beat it rather than just another 200mhz added to otherwise a very similar chip.
AMD were in a terrible place the whole time. Intel's dominance clearly wasn't from a position of pushing and innovation, it was from an absence of competition. They spent 6 years (until Ryzen release in Feb 2017) doing a wee clock speed increase with a very nicely segmented processor market that I'm sure turned them a VERY pretty penny. All that time and here we are, 2019, they're still pushing exactly the same architecture with ONE node shrink.
Where the hell are you guys spinning this tale of amazing innovation and progress from?!
You can't talk about ignoring bias and heap praise on Intel for doing next to nothing for close to a decade because their rival had properly tanked (generally due to Intel dirty tricks but AMD made some seriously bad choices too) that's blatant hypocrisy.
Come 2017 and Ryzen... ohhh, Intel suddenly have a 6 core chip (on the same node). Come 2018 oooh, Intel can do an 8 core chip too (on the same node, now at furnace levels of power/heat).
Where's the innovation in all this? Dominance, absolutely. It completely wasn't of Intel's own making though, the parts of it that were, were dirty tricks and backroom deals. That's absolutely devoid of any reason for celebration.
AMD's come back, they came in with a poorer (but smaller) litho, 2 gens later they've changed architecture already. Chiplets is clearly where Intel will follow. That they do nothing but minor clock bumps for nearly 10 years followed by cribbing the design of their competition. Clearly dominant and innovative, yes.
Last edited:
Associate
Cinebench, CPUz and a few other tests of IPC show that at the same frequency Kaby = Ryzen 1 and Coffee = Ryzen 2 within margins of error. How people still don't get this I do not know. The advantage Intel has is purely frequency for those generations. Found that out as one of the things I do a lot of testing for is iRacing which is heavily dependent on single
@smilingcrow I buy as and when I need to buy tech. I buy what's good for the time period I am purchasing. If there's some large change close to release I'll wait for that instead of wasting money on something that's already out of date. As I said I had a 3930K for 6yrs before getting my current Threadripper. There was no point in replacing the 3930K due to the lack of extra benefit from anything released by Intel during those 6yrs. AMD had nothing during that time either so I'm finding your position slightly odd. What do I need to address about Intel's stagnation and lack of innovation to provide a product worth updating to over the course of 8yrs in from Sandy Bridge to Coffee Lake? I'm at the point of looking for a 1680v2 for that old rig so I can use it again as it's still fine other than the dead CPU (killed by a broken LLC setting on my old Gigabyte board).
Loyalty to a company, whichever one, who exist purely to remove cash from your wallet is a weird concept.
I purchase pc components exactly the same as any other expensive electrical products. Price to performance, that's it. I'm typing this on a pocophone for that very reason.
I purchase pc components exactly the same as any other expensive electrical products. Price to performance, that's it. I'm typing this on a pocophone for that very reason.
I agree with the sentiment however I just need to offer a counter to the "only exist to make money" meme. At the level of the average employee this may be true but the people doing the real work, the creative ones. They're doing it for the passion they have for whatever they're producing.
Passion, really? I work in the media, a creative industry. I work purely for the money. I don't mind the work, which is a bonus, but I work for those who pay the most. Passion doesn't pay the mortgage.
Permabanned
Dam right
That's exactly I mean. Not everyone has the passion some are just mercenaries, like you and I. That's not everyone and you're not spending every waking moment on a business you don't care about. I work 9-5 because I value time more to spend it on things I enjoy i.e. not work. Others find their work is who they are and what they do, they have a passion for it.
Absolutely agree.
Still the likes of intel, AMD and Nvidia exist to make a profit. They take our money and put it in theirs. Which I've not got a problem with, it's our choice. But why would anyone have loyalty to this? They don't care about consumers apart from how much they can squeeze out of us. Of course we hear the usual nonsense about its 'all about us', in reality it's all about the bottom line. Trying to engender brand loyalty is a marketing con, don't fall for it.