Ryzen vs i7 (Mainstream); Threadripper vs i9 (HEDT); X299 vs X399/TRX40; Xeon vs Epyc

Evidently you guys need it. You sure cannot figure it out yourselves if you think AMD cannot get over 8% IPC in a new CPU architecture but think Intel can get 17% frequency increase, something not done since possibly sandy bridge or earlier.

And I'm talking peak increase on the architecture under water overclock, not Intel leaving room on the table then increasing frequency the following iteration, but the process not really adding more than 100-200MHz amounting to 5% or less.

Literally, nothing has been shown for Intel improving 10nm iterations worth anything. Hell, Intel is resurrecting 22nm Has well designs now! And yet we are to believe Intel is going to pull a rabbit out of it's hat on 10nm++, when they didn't on 10nm to 10nm+. Give me a break! I don't believe in fairy tales, so considering Intel had initial issues going from 22nm to 14nm, delaying it a year, then rehashed skylake on 14nm for three or four years because they can't really get 10nm going right, how am I supposed to believe they will get that gain in frequency on 10nm++ or have a 7nm working process in 2021? I'll believe it when I see it.

AMD has actually been delivering on IPC gains and changes when they claim them the part couple years and TSMC has been nailing the process nodes so hard that they are talking 3nm in 2022, which would be a switch to gate all around, most likely using nanosheet FETs. They have 5nm at 50% plus yields with Intel's 10nm (7nm equivalent at TSMC) yields being in the teens to 20s. So where is your analysis coming from? Certainly not reality.

 

Your reality is wack. My reality is that an AMD platform, as a whole, feels less snappy than a remotely comparable Intel one. You don't see this. I don't care (if you do or not).

I do and don't care for any bellybutton focusing 'analysis' that refuses to accept what is actually real, and not merely theorized.

AMD has been far, far behind for a very long time. I applaud the efforts and gains made by them, but they still have a way to go.

I don't care where AMD or Intel bring the improvements from; I just buy the actual improvements (when they're there, for me and my clients).

This is the reality: Intel's demise has been greatly exaggerated. Stay tuned to see what tech we'll get to play with next and leave the 'analysis' to people who actually know, not the people that can do the simple math of random units to prove their own agenda.

 

So lots of words, lots of conclusions, zero facts and evidence. Good job being your normal self with the level of analysis I've come to expect from you!

For snappiness, do I need to post the date and times Jayztwocentz and others say AMD is as snappy? Maybe I could post the analysis between Wyndell of level1tech and Dr. Cutress from Anandtech basically taking a dump on Intel for a couple hours after the Threadripper release. You are not those people with hard data supporting their conclusions (which they give behind the scenes insight in that two hour conversation).

Also, when did I say this is Intel's demise? Intel is going to weather this storm. Their worst case scenario is FUBARing the 7nm process, something that cannot be known until 2021-22. I said wait and see for them to deliver and don't have faith in them until they deliver, not that they are dying.

A proper reading would say AMD's initial order was too low on Zen 2 parts and keeping more for Epyc has, between the order and Epyc demand, caused limited desktop growth. The server sales will filter into the market share numbers in Q4 through H1 2020. The mobile numbers shocked me as Intel was targeting holding them to 12% and the projection was around 15% and AMD got 14% in the one area Intel truly dominates in.

But, me showing Intel holds over 80% in desktops and mobile, along with over 90% in server says they are not dying. Until AMD holds 40% of the server market, I'll likely agree with that. If Intel doesn't deliver on 7nm, they are in deep poopoo, though. At that point, they WILL need to contract with either TSMC or Samsung for their process IP OR to fab for them.

Intel has some great architectures, but are plagued by process woes. The reason the 8-core rocket lake has fewer cores is because it is a wider core design. They designed those for smaller nodes that save space so that you get the increased performance of going wider without creating significantly larger dies which can reduce yields, etc. That is why, since comet lake is not a back port, rocket lake, even with two fewer cores, should have the increased IPC with sunny cover and the advantages of Willow cove, which did not focus on ST performance. It will be a good chip. But that will be competing with Zen 3, and since this is not a skylake derivative, we don't know heat output or expected power efficiency curve. As such, if Zen 3 does get an IPC boost of 10%+, Intel having a core deficit may harm it. Pricing may harm it. Etc.

Anyone denying Intel's problems (you) seems to be willfully obtuse at this point, just as anyone saying Intel is dead is the same on the opposite end of the spectrum.

Notice you trying to lump me in with another class to diminish and dismiss my analysis. Common tactic for people who lack foundational argumentation.

Come back when you have an actual argument.

 

That's odd, because the snappiness of both the GL702ZC and Predator Helios 500 (all AMD hw laptops) was pretty high/excellent and I hadn't seen anyone complaining of their desktop Zen systems.
Plus, independent reviews would definitely disagree with you on Intel being more 'snappy'.

Anyway, it is correct that Intel has been messing things up with 10nm repeatedly, and I'd take their claims on IPC and performance gains (not to mention readiness of 10nm) with a grain of salt.
Is it possible they will deliver? Of course its possible, but given their track record, I wouldn't hold my breadth until the reviews are out (and if you check my post history on prior AMD releases, you'd notice I maintained the same attitude).

AMD has been quite consistent (not to mention dependable) in their claims about IPC and performance increases since launching Zen uArch.
And as it was noted before, TSMC reported that 5nm yields are already at (or better than) 50%. So, when compared to Intel, AMD actually has the process advantage (even the existing 7nm is better than Intel's 10nm... 7nm EUV on which Zen 3 is being released in late 2020 on is even better - and Zen 3 was touted to be by AMD itself 'a brand new uArch' - so its quite likely we will be seeing up to 15% increase in IPC... if not potentially more, but like with anything else, I maintain the premise we need to wait and see until its actually released to gauge its performance).

Your conviction regarding Intel's promises doesn't have conclusive evidence to back it though. We have a stack of broken promises, delays, and extremely low yields for 10nm. I just don't see how Intel can ramp up yields of 10nm to be viable for desktops... and if they do, the clock speeds will likely be lower even compared to AMD.
Also worthy of note is that mobile chips with 10nm Intel parts have demonstrated some noteworthy benchmark scores against AMD APUs'... that is, APU's of 12nm using Zen+ (not Zen 2).
So, it will be interesting to see how Zen 2 stacks up against Intel mobile 10nm parts.

 

Intel's Marketing "Benchmark" Claims Shenanigans are baaackk...

First a short translation and report of the most recent marketing shenanigans from Intel (starts at 00:40), then AdoredTV reviews this and the rest of the last years and more Intel BS:

Intel Has COMPLETELY Lost It, 3nm COMING!
Dec 8, 2019
Gamer Meld
Intel making the wildest claims, Radeon RX 5500 XT is real, 5nm is around the corner and 3nm?! Stay tuned...


phoenixtheraver 1 hour ago
"It's so sad how Intel has been beaten so badly, all they got left is to slander their competition with lies and propaganda. It's both sad and really pathetic."

Ramon Schepers 1 hour ago
tsmc: 3nm!
intel: lets move back to 22nm! intelol.

F_S_P 1 hour ago
""Stop it. Get some help." Best meme for describing Intel right now"

Anton Nym 2 hours ago
"I'm still upset with Intel for decades of incremental improvements, unfair, monopolistic business practices, and massive overcharging. Intel has therefore made me a huge AMD fanboy. Go AMD!"

Darrel Madden 46 minutes ago
"Which is exactly why I went all in with Ryzen. I've been in IT for over 30 years and Intel have been dirty for way too long. And my R5 1600 runs just great."

Intel's Disgraceful Marketing Gets Worse
Dec 7, 2019
AdoredTV
...and [Intel] also violates the FTC orders.


Excalis Ogre lord 1 day ago
"I will personally never buy an Intel CPU again."

spaceduck413 1 day ago
"Ahh yes, just when I was thinking MAYBE a 10980xe would work for my use case if it came back in stock before the 3960x, Jim releases a video reminding me why I don't want to give Intel any of my money. Great timing Jim!"

nazgul77de 1 day ago
"Thanks a million for uncovering and summarizing marketing fraud and taking the diligence to report it to the responsible authorities. Let us hope that major OEMs get tired of Intels delivery issues and are not afraid of Intels wrath when offering competition products that customers are asking for. The very very sad thing is the majority of the public will never be aware of the scale of the lies, because the [BS] marketing gets advertised widely thanks to Intel money, not the analytical tech press reports exposing the truth."

JwJw01 1 day ago
" 30:33 Damn you're not holding back. Go get them."

To end on a positive note, here's ThreadRipper 3 killing it against Intel:

Threadripper 3000 Launch Overview - AMD's Victory Lap
Nov 25, 2019
AdoredTV
Intel's pain will be even worse in 2020.

 

The problem with AMD owning HEDT is they actually have to get silicon to market that can be purchased. A limited release just to reviewers as PoC won't cut it for long.

 

I wasn't going to say anything but this was just too funny. You "feel" that AMD is less snappy, but talk about what is "actually real." I've used a desktop/laptop or two over the years and I don't "feel" a snappiness difference between them provided they have similar SSD configurations. You've certainly made up your mind that there is some usability difference between AMD and Intel and that's all you're banking on. I don't see it and I've used a lot of hardware.

I think blindly supporting any company is a silly thing. I'm glad AMD is on top for the first time since the Opteron 165 days, it is helping them gain market share which should help them remain competitive in the future. Intel has been stagnate for far too long.

 

He has to admit the lack of 2019 TR sales is the absence of 3960x and 3970x in the market for sale.

 

intel back porting so we might get to see golden cove on 10nm+++, this is a smart move as usually they have nothing for desktop except refreshes. @tilleroftheearth looking forward to see what intel can bring again by 2021.

if true, golden cove 30% IPC over CFL while getting the good old 5ghz, its hard to settle for less now unless its a big difference in value and efficiency.

 

It's nothing but Intel's marketing BS. Intel's "plans" are fantasies at this point without anything to back it up. Intel prognostications aren't believable any longer, and haven't been for years now.

Intel's 10nm has failed to arrive in any high volume useful form for 6+ years and Intel is still failing to deliver desktop / server 10nm CPU's that are as competitive as their own 14nm CPU's.

With Intel giving up on having a majority share of mainstream CPU's (like Intel has a choice!!) => Intel has lost it and to pretend to be able to predict anything past next week is a joke.

This Intel BS is a bunch of baloney to dazzle and confuse the unwary. I can't believe Ian is repeating Intel's BS seriously => "Intel Expects" BS: Intel doesn't have a clue.

Intel’s Manufacturing Roadmap from 2019 to 2029: Back Porting, 7nm, 5nm, 3nm, 2nm, and 1.4 nm
by Dr. Ian Cutress on December 10, 2019 2:35 PM EST
https://www.anandtech.com/show/15217/intels-manufacturing-roadmap-from-2019-to-2029

Enthusiasts Aren't Stupid!

https://www.reddit.com/r/Amd/comments/e8ien0/enthusiasts_arent_stupid/

 

This will change. AMD wasn’t competitive before lately. Just look at history. Same for Nvidia... They won’t do amazing things before some steal their sales. I expect you neither sit with the engineers table.

 

Some Intel vs 3950x + ThreadRipper 3 Benchmarks...

i9 10900X vs 10920X vs 10940X vs 10980XE vs Ryzen TR 3950X vs 3960X vs 3970X
Premiered Nov 26, 2019
Benchmark PC Tech
i9 10900X vs i9 10920X vs i9 10940X vs i9 10980XE vs Ryzen 9 3950X vs Ryzen Threadripper 3960X vs Ryzen Threadripper 3970X Complete Roundup CASCADE LAKE-X vs RYZEN THREADRIPPER 3000 GEN BENCHMARK

 

Notice how they are talking about things I discussed on process and where the industry was headed this past summer and last year. Only thing they didn't mention much of is a possible fully-depleted silicon on insulator (like sapphire), or a mixed situation of FDSoI with finFET or GAA or both.

Intel seems like they released that more to calm investors by saying they have a long standing plan. This is smart, as forward looking statements have safe harbor protections, so long as they do not embed current information or base it on known past performance (which would be teased out to have protections for projections but not for misstatements or omissions on past performance).

Either way, I mentioned Intel is releasing comet S in April. If true, rocket may not come to market this year, which back ports likely Willow cove, not sunny (which is only 18% IPC). We do not know what Willow will bring, but Intel already said they did NOT focus on ST performance, so likely not a lot of extra IPC.. If that then comes out in 2021 instead of 2020, it will be up against Zen 4, not Zen 3, which means AMD will have two generations worth of IPC stacking, which zen 2 is already ahead on IPC and performance is roughly equal to Intel with higher frequency. If AMD pulls off the 10-17% IPC with Zen 3, and Intel releases rocket against it, then Intel may have a slight performance lead. If it is going against Zen 4, then Intel's rocket lake will be like an FX processor: high frequency, low performance. I'm hoping that won't be the case, but known info says that would be the scenario and that would really suck.

If Intel instead cannot get rocket out until 2021, they must go ocean/golden cove to be competitive! That would be closer to ole's estimate and would keep them in the game. But the rumor says it is Willow cove, and that may not be good enough.

 

with zen 3, it is a dead end platform which not that worthy of consideration. but with intel, willow cove or golden cove will be back ported at that time to 10nm+++ and its very easy to see a 4.8-5 ghz cpu all cores mind you at a much higher IPC.

on the other hand, zen4 is gonna have to compete with that at 5nm. no doubt it'll probably be more efficient, at a much lower clocks too. i'd guess around the 4-4.1ghz we see today, vs a similar or stronger IPC CPU from intel at 4.8ghz, we already know how that will turn out.

 

You are missing it. Willow doesn't focus on ST. Single threaded performance equals IPC. Since rocket is now planned 2021, it is going against Zen 4. So AMD will roll up two more gens of IPC compared to Intel. Rocket lake is 14nm, not 10nm. So why are you bringing up 10nm, which only applies to the iGP on rocket lake? Also, there is nothing to suggest Intel will hit 5GHz on 10nm, which they will be around 4GHz if 10nm+ says anything.

Zen 4 is on TSMC 5nm. Zen 5 will either be on 5nm+ or 3nm, considering TSMC is now saying 3nm may be ready for 2022. So Intel is competing with an EUV 5nm node shrink and two gens of IPC increases when they released rocket, which is Willow cove on 14nm.

So your math doesn't add up!

 

amd hasnt mention anything about zen 4 IPC though. unless 3D cache stacking is used i dont expect all that much. also in 2021 it'll be golden cove not willow cove so no problem here with intel's IPC, assuming it is a major leap in IPC.

 

No, do a simple Google search. Rocket is 14nm Willow cove, NOT golden cove. So some research before you speak. PCGamesN has an article talking about it from 8 days ago!

Also, Zen 4 is a full redesign. Considering the IPC gain on Zen 2, then rumors of 10-17% IPC gain on Zen 3, another overhaul for Zen 4 plus switching to DDR5 increasing memory bandwidth, they will likely have another 10-15% IPC there. So that is easily 23%+ on IPC over Zen 2. So there is always a chance Intel's speed, which was neutralized by Zen 2 IPC, results in Intel getting 18-26% IPC gain and them being equal.

But, if AMD does get closer to 15% per generation, then AMD could have a larger 5% IPC lead over Intel. If they hit 7% performance over Intel, that was the industry standard average of IPC for new core architectures before the Zen/sunny cover era. That is starting to move toward FX land.

 

why do i care about rocket lake? i am talking about golden cove in general. intel's slides showing backport to 10nm+++ with a new arch that is meant for 7nm, not 10nm arch back porting to 14nm+++, and they'd obviously only do that for 1st gen 7nm desktop/server parts so have best yields + performance.

 

You are going to believe one Intel slide when rocket lake S is their mainstream desktop offering in 2021 and they haven't met their process roadmaps in 5 years. Good job, cause that makes sense. Also, Intel often does not produce on the node until 6+ months after it is allegedly volume production ready, meaning your thought is about their 2022 processor which will be against Zen 5, not 2021.

 

yea i'll believe the backport of 1st gen 7nm arch design to 10nm+++ cause that actually does sound reasonable. i mean we are at 10nm+ or ++ at the moment so it doesnt seem all that impossible. for the 7nm onward i have doubts but thats so far down the future no point talking about it.

 

That's cool, but I'm trying to get your timeline right. What you want is in 2022. That will likely be Zen 5 on 5nm+ with potentially a larger use of EUV. To put this into perspective, the 5nm non-plus will have 14 out of a total 59 possible layers done with EUV. It will have a transistor density around 173.1 MTx/mm^2. That is a 1.37x transistor density over Samsung and will be used by Apple, AMD, and HiSilicon in 2021. The plus variant is a refinement. I'm being AMD will use the refinement, waiting for 3nm to mature for 2023.

That is double the density of transistors compared to Intel's 10nm node. That will allow AMD to go wider, which helps IPC.

So even if Intel is using 10nm+++ at that time, they will have a process deficit in 2022. But, golden cove will have another large IPC gain as they are focusing on ST for that generation.

The only question, and where we some down on different sides, is that you believe Intel will get to 5GHz, whereas I think Intel will be at 4.3-4.6 GHz. We can agree to disagree.

But 10nm was allow AF and Intel killed it (also very low yields). 10nm+ is 900MHz-1.1GHz slower than 14nm+++. Intel was always over 4.4GHz on 14nm, seen in Broad well, with skylake hitting 4.8, 14nm+ kaby hitting 5.0, and now coffee refresh up to 5.3 if you can keep it cool. So it isn't out of the question. But Intel also had a large increase between broad well and skylake, whereas we have little to go on but it seems there wasn't much of any improvement in frequency from cannon to ice lake on frequency. So getting an additional GHz just doesn't make sense to me. Intel isn't magic, just like mortgages are not immune to the forces of the economy (which believing such gave us the 2008 recession).

 

Where do you come up with these flights of fantasy?

Zen 3 will be in the entire line up of Ryzen CPU's, from APU's to Epyc.

Along the way Zen 3 will offer a big upgrade for EOL AM4's either then or down the road for owners of AM4 computers.

I upgraded my AM2+ with the last best CPU available for that socket and it's still running fine 12+ years later.

As for the Zen 3 ThreadRipper 3 and Epyc 3, who knows how much longer those sockets will continue or EOL as of today?

One of the benefits of buying EOL motherboards and their last CPU's is the stability and long life they can provide. All the platform bugs are worked out, the prices have leveled off to reasonable prices, and if you do a good job on the build you've got a strong long term computer for many years of service.

Buying the cutting edge all the time isn't a good idea for most people, as that's where the most problems and disappoints reside. Waiting for all of those bugs to get firmware updates gives you a stable platform to build on.

In my personal life I prefer the stability and reliability of buying a platform that has been fixed and provides little or no distraction or drama.

For work, that's a different story on one hand we have to keep at the cutting edge for competitive equality, but my penchant for stability and price vs performance focus is also of great value for companies that want the best bang for their bucks and low license, support, and service costs.

It doesn't hurt to save as much money for the company as you cost them, on top of making money for them by keeping them competitive.

Giving up on Zen 3 because it's EOL on AM4 is childish and not giving due consideration to building systems with stability and long life.

This time you are invoking Intel's "fantasies", which likely will never happen just like all of Intel's 10nm fantasies have consistently failed to deliver in the end.

Anything beyond that is even fuzzier speculation. Zen 4 may bring many new technologies, but with those new technologies will be a time of instability while the bugs get worked out.

Reading your post again I gotta believe you really are lost in Intel's BS, when will you wake up and realize Intel is full of themselves and aren't delivering anything worth buying? - now or then, Intel's got a long way to go to successfully make it out of the woods.

 

Amd's hug problem is no one has full machine offerings right now. With the supply issues no OEM's are going to put out preconfigured machines. you need to be able to walk into B&M stores and have a good choice of machines. The same for business catalogs etc.. DIY's will never yield a significant overall market share.

 

i am saying these on the assumption zen 5 will have similar IPC as golden cove. cause if it isn't theres no need to go AMD since intel has dropped price while offering higher clocks. zen3 will have about 15%? IPC over CFL if lucky so zen 4-5 hopefully will be another 15%. with IPC matched, obviously i will go with higher clocks even if i pay a bit more.

 

Talking about Zen 5 and Golden Cove right now strikes me as more than a bit speculative.

 

Shut it, because you look bad when you speak without hard data.



And yes, zen 3 is over 15% IPC over coffee lake considering Zen 2 is already like 6-8% over it and zen 3 integer is 10% over that. Intel has lower IPC until ice lake.

Further, look at those yields. 5nm over 80% yields. Intel can't get 10nm working and 7nm who knows what is happening.

 

https://www.anandtech.com/show/14605/the-and-ryzen-3700x-3900x-review-raising-the-bar/6

Actual IPC results for Zen 2. Now stop talking without facts.

 

It is. He doesn't like AMD pulling ahead and has justified waiting until then, but is claiming that Intel will have what is scheduled for 2022 in 2021, then is comparing what should be available then to Zen 3 coming out 2020, then saying Intel wins. It is the most absurd thing I've heard in a long while.

 

i said zen2 is 6-7% IPC ahead. guess you can't read, goes to show you let your emotion gets to you. now shut it

 

And you are using the most conservative estimates of per generation IPC gains for AMD while assuming the best case scenario for Intel, including a possibly unrealistic frequency on the 10nm refinement.

Reports already have closer to 16-23% IPC of Zen 3 over the current skylake based cores. Two more generations with larger/wider core designs is at minimum 14%, but more likely closer to 20-25%, with an outsized amount up to 32% IPC, added to the minimal 16% IPC of the Zen 3, and you would have closer to 36-41% IPC over that, with the outsize level of up to 48% or so. Also, even this isn't proper, because the year over year IPC changes are actually multiplied, similar to compound interest, meaning that the IPC increase over a current gen will be higher than what I put here as additive.

You then assume Intel will hit 5GHz on 10nm+++, which I'd like to know the basis for that belief, because nothing suggests they will hit that. I'd say 4.4-4.6GHz would be more reasonable, and even at that, it will be a furnace. That places it at only a 5-10% frequency advantage, assuming AMD doesn't increase frequency with the newer nodes, which would likely be 5nm+ in 2022 against your theoretical 10nm+++ golden cove CPU which you assume to be around 30% more IPC over current core uarch.

As I said, Intel will be competitive in 2022, likely not before. They will also likely be on chiplets by then, meaning this also does not fully account for the latency hits they will have in the transition.

So you acting like AMD will barely be competitive or be crushed MAKES NO SENSE! It is what the match should have been this year if Intel had a working 10nm process, etc.

Another important point is that at 5nm, the density used for HPC will be closer to 100 million transistors, whereas Intel's 10nm variant will be closer to the 60M range. If my estimate is right, AMD will be reducing the core die size to around 51.28mm^2, which is a nearly 33% reduction in size while adding transistors over the current 7nm process. That would give just over a 50% yield on the current defect density and increase the number of dies per wafer to 616 good dies on the current risk production levels. considering 15-18 months before using this node, 21 months at the farthest, AMD will be able to have a descent amount for market. Can the same be said of Intel considering their current "volume" yields are SO LOW that they introduced Ice Lake on mobile with about 1/3d the normal level of SKUs and even then they are on the market in very low quantities and the 14nm++++ chips outperform them. Same with next year's server chips with high core count. They will only get a tiny amount of usable chips, which is why Cooper Lake will be the high performance variant. If that is the case, on what basis do you think that in 2 years, Intel will solve both the yields AND the frequency issue to correct their issues with volume to market to alleviate the shortage (other than having mobile theoretically on a 7nm node, while trying to have both server and desktop on 10nm nodes)?

Further, why are you assuming Intel will hit another 15-18% IPC increase for golden cove? Please enlighten me on that.

Source article on TSMC 5nm yields used to estimate current yields for 5nm using this defect density:
https://www.anandtech.com/show/15219/early-tsmc-5nm-test-chip-yields-80-hvm-coming-in-h1-2020

 

All right, let's be a little careful here.

AMD's statements about Zen3 gains are "right in line with what you would expect from an entirely new architecture". That statement's vague, and it's qualified by noting that Zen2 was a bigger gain than would normally be expected from an evolutionary upgrade, but unless something else has come out that I didn't spot, 15-20% isn't entirely unreasonable to speculate on. Although this earlier piece was speculating 8% IPC uplift plus maybe 200 MHz frequency.

Speculating beyond that to Zen4 and Zen5 is rather more of a stretch. AMD has certainly done a superb job backing up its claims for Zen2, but my comment above about Zen5 and Golden Cove applies here, as well. Squeezing out gains isn't going to get any easier as chips get faster and architectures are better tuned.

Besides which, no one here is making purchasing decisions based on what Zen4 or Zen5 will be. That's just too far in the future to be very relevant. At least for mainstream, the chipsets will be different; the memory and I/O platforms will also be different, if not by Zen4, certainly by Zen5.

My own plan this spring was to wait for Zen2 and get something between 8-16 cores depending upon where the prices wound up. I had to replace my i7-5820K and went with a 2700X, which changed my equation somewhat. There's not a lot of point for me to get a 3700X or 3800X now; it's certainly a very nice chip, but it's not that radically faster than the 2700X. The 3900X is still expensive and in very short supply. So for me the Zen3 question is somewhat relevant (since I presumably won't have to replace my motherboard). But it probably means I'm not going to upgrade my 2700X for a while unless the bottom really falls out of the market for the 3900X (or AMD releases an end user 3900 at a good discount from the 3900X). More likely I'll wait until the Ryzen 4000 chps are out and see what's actually going on, and then decide.

 

You misunderstand what I'm referencing for Zen 3. I'm relying on the reports of 10% integer IPC and 17% average integer+FP that has been leaked, which is from the same outlet that has the 13% IPC increase leak last fall, which was fairly accurate. I then added AMDs Zen 2 IPC advantage of 6-7% over coffee to the 10% integer leak to come to the 16-17% amount over coffee lake. The average is 7-8% IPC, so going with the 10% integer IPC is not unreasonable for the estimate of Zen 3 over Zen 2. It is the reference point that matters. And 10% is roughly in line generationally, matching the statement from AMD.

(Edit: so the 16-23% IPC range was based on the leaked integer 10% and integer+FP 17% IPC, whereas the FP 32% with some up to 50% was excluded, which 6+17=23%)

Beyond that, I'm basing it on my own assumptions on effects of where I believe AMD will go with their architecture. Going three years out is fully speculative as no one knows the uarch changes nor fully the yields or frequencies achieved on future process nodes.

Instead, my point was to show he was casting AMD in a very conservative analysis on upcoming products while raining down an overly optimistic analysis on Intel's products considering Intel has had issues delivering to date.

Edit:



Edit 2: your October article references redgamingtech as the source. Here is redgamingtech from Dec. 6, 2019 giving the higher numbers I am using. They were the ones with the 13% IPC estimate for Zen 2 last fall, which was slightly lower than the final 15-17% IPC which materialized later. As such, I will estimate on their numbers again as I've no reason not to.

On nvidia, rgt got it wrong for 20 series, but so did most of the tech press. I believe Nvidia uses subterfuge, feeding contradictory information down the line to see where it flushes out, along with instead of creating true leaks to generate hype, they use multiple variants on information so people are talking about which rumor is correct.

 

In addition to the update from last week, RGT has an analysis of IPC leaks from today as well.

 

sounds like AMD gonna be in back in playing catch up game real soon if this keeps up. assuming the leak is correct, and I have high doubts it will be, zen 3 will be about 20% faster than CFL while running at a max 4.3ghz at best at 16 cores and coming in late 2020. i mean just icelake alone is ~15% higher, when golden cove and 10nm++ or +++ hits in late 2021 or early 2022 AMD will be behind.

 

Are you really that ignorant? Next year on desktop is another 14nm++++ sylake called comet lake. And there is another Intel vulnerability that the fix requires locking the voltage at stock voltage.

That 10-core is going against a 16-core and will likely be between 4.8-5.0 GHz. If AMD at 4.3GHz is around 20-23% higher IPC, that would be AMD winning on outright performance AND having more cores. That is called a win.

Ice lake is only on mobile chips and caps at 3.9GHz, which is lower than Zen 3 speeds and has a lower IPC. That is called losing.

AMD will be updating to a new socket, get improved I/O, a new memory standard and interface (AMD has stated publicly lack of memory bandwidth is kneecapping their chips, so they are planning to increase that; this likely means increasing memory channels on server to 12-16 channels, HEDT to 8 channels (see TRX80), and potentially even quad channel for mainstream, if not adding HBM on package), plus a new generation with IPC uplift for Zen 4.

Now, where is your data on golden cove IPC? Where is your data on 10nm++ being able to even get passed 4GHz, needles trying to suggest 5GHz? Even if you add the 17-18% IPC for sunny cove, which means AMD still would be up 2-5% IPC while running at equal to faster speeds than what we have seen on 10nm+, you then have another jump in IPC, which you assume is 7-8% even though AMD did 53%, 3-4%, 15-17%, and is on track for another 10-17% on Zen 3, depending on if talking pure integer or overall mixed workloads, all while having more ability to do greater changes like going much wider due to a 1.8x density increase going to 5nm, which allows for much higher IPC and who's yields on a pure die shrink would already yield as many dies per water as what they get on 7nm DUV.

You assume Intel will fix the frequency and then give Intel another 18% IPC, which has no current basis and only then would they be competitive with a likely Zen 4, especially if 10nm+++ runs at 4.0-4.4GHz, all while coming out in 2022 after AMD has already confirmed zen 4 for 2021 and everyone already knows Intel is doing sunny or Willow cove on 14nm++++ for 2021, meaning your speculative Intel chip would need another Zen generation, zen 5, as the direct head to head, not Zen 4.

Hmmm. Seems you likely need someone to map this out for you.

 

im not talking about next yr, but in late 2021 and early 2022. I would definitely want a 5ghz+ CPU with higher IPC than a low frequency CPU with the same IPC.

 

Once again, where are you getting 5GHz from?

 

5 years ago Intel was gonna deliver 10nm desktop / server CPU's in 2 years.

3 years ago Intel was gonna deliver 10nm desktop / server CPU's in 2 years.

Last Intel said we were gonna get 10nm desktops and server CPU's at the end of this year in 2019... and here's a few gimped 2 core 10nm ULV CPU's - oh, and Intel added that they have disabled the onboard IGPU in those gimped 10nm ULV CPU's.

In the middle of this year 2019, just when we were eagerly expecting announcements for the 10nm desktop and server CPU shipments this year, Intel told us that they are now going to ship 10nm desktop and server CPU's at the end of 2021 or early 2022... in another 2 years, and here's a few gimped ULV 10nm 4 core CPU's.

Intel hasn't been able to get 10nm to work well enough to replace 14nm on desktops and servers, and at the rate Intel are progressing with 10nm it might be another 10 years before they work up to a 16c 10nm desktop / server CPU.

And, I wouldn't believe Intel on anything they are about to say about 7nm either.

Intel predictions for new process deliver can't be believed because Intel has proven that they can't deliver the truth about where they are in the process.

When Intel actually delivers 10nm / 7nm desktop / server CPU's we can buy, that's the time to spend time on Intel.

Until then, AMD is putting on a great show, pull up a chair and enjoy.

 

from 10nm++ or +++, its definitely possible.

 

Source?

I did the math and it seems less likely now more than ever.

Let's examine Intel's 14nm line and average overclock.

First you had Broad well C with about a 4.4GHz OC. Skykake had 4.8GHz. kaby was 5.0. Then coffee was 5.1-5.3, as each subsequent CPU has been. The old metric was 100-200MHz reduction per two cores and Intel's binning has gotten better. That is 900MHz from 14nm to 14nm+++.

With cannon lake, you had a Max boost of 3.3. With ice lake, it is 3.9. This is after Zen when Intel stopped leaving so much room on the table for OCing. This is also dual and quad core mobile chips, which is where 14nm started on core counts, and that 600MHz is with one new architecture and one process iteration, the same as the 600MHz jump seen from Broad well to kaby lake.

Even assuming up to a 500MHz boost over what mobile saw, like the comet lake mobile boosting to 4.8GHz and 9900KS (extreme binning), which has a boost of 5.3GHz.

So, taken together, it looks like a desktop variant would be in the mid-4GHz range, considering the additional cores and refinements as seen on 14nm.

Could something happen to make this estimate higher or lower? Sure. But pure wishful thinking that Intel will magically get 10nm to 5GHz isn't going to get it there, and you are not presenting any evidence or argument to contradict my points, instead referencing and repeating that it will be two more iterations, which doesn't help your point.

Further, even with that, I don't see Intel's 9th gen quad core going significantly further than the 8 core chips on air and AIOs.

Please, enlighten me to your proof.

 

Bearded Hardware OCing 3970X on water.

 

Ho-Ho-Ho, Merry Xmas!!, have you been nice? Then you get a nice Ryzen CPU - been bad, then you get a lump of "Intel"...

Who would want Intel on Xmas day? - with Vulnerabilities, thermal throttling from locked voltage adjustments, and who knows what else will come to haunt Intel in 2020??

What else could Linus do but recommend an all Ryzen Xmas?

Holiday CPU Buyers Guide - 2019
Dec 14, 2019
Linus Tech Tips

 

See:
https://www.tomshardware.com/news/amd-ryzen-9-3900x-wprime-world-record

More AMD catch up happening.

This time, to a 2017 processor Intel record.

Go AMD!

 

Leaked Intel Six Core CPU Reveals a New Architecture Coming Soon tomshardware.com | today

Overall, it's hard to make any firm conclusions based on this result. The possibilities are many and the evidence is very thin. The one certainty is that this CPU is not a Skylake derivative. What isn't certain is whether or not this is 14nm Rocket Lake with a backported core, 10nm Tiger Lake for desktop, or something totally different, like Ice Lake for server.

 

" Intel has already revealed it has the option of backporting architectures intended for 7nm onto the 10nm node, so the possibility of Rocket Lake being a 14nm architecture with designs intended for 10nm CPUs is certainly there. "

Man, Intel is so whacked out on expecting failure that they are preparing for backporting failed 7nm designs to the failed 10nm process - which is itself a failed process for them for so long and still won't allow full speed full cored full power CPU's.

I would imagine there would be experiments by Intel - trying one focused portion of the architecture or another in test CPU's as they push forward on 10nm or 7nm that wouldn't be a practical production sku but instead is an attempt to put what they can into a testbed chip - and that may be what is making it out into the benchmark test databases.

Of course if that's all Intel can do, maybe they will try to ship it and see if anyone will buy it. Intel will likely reach that point again - like the 2 core 10nm ULV CPU's without IGP.

 

I mean we should wait with the final judgement until we see the end results. Could be nice and it can be fiasco. If it's nice, AMD will see it's stock shares follow the waterflood, similar as "after you have clicked" the button on top of the toilet. How long was Adam and Eve in Paradise?

And not all will come up again. Maybe after several years as last time. Time will tell.

 

Looked at another way, this is a consumer-level AMD processor that costs $500 against an Intel one that even now goes for more like $1300 (and whose successor isn't really a whole lot faster).

It might be interesting to see what a 3950X would do, or a Zen 2 Threadripper (which is a better comparison against an HEDT).

 

@rlk sure, fair points. It doesn't change the glaring fact that two years later, AMD is 1% better than what Intel was two long years ago.

Price points are not interesting to me in a historical context. Sure, it matters when I'm buying something 'now', but if we're showcasing ultimate 'performance' (at least for a single metric like here) - then $$$$$ be damned.

There is always the 'what if' of doing something different... but all that matters is what is actually done.

Can't wait for a real head-to-head in a future showdown when AMD and Intel both offer the world what they consider state-of-the-art at roughly the same time point. That is when I'll feel things getting real and interesting.