You would be right. I may be wrong on the dates but I remember someone mentioning around that time AMD had multiple teams working on it. I followed Zen closely until it launched because their hotchips segment caught my attention. Mike Clark's spill was punctuated by something said earlier. That's all I remember. It's also what allows AMD to bring forth or back tech stacks pertaining to a particular design.Just to make sure I am not in the wrong. I am viewing the Video of AMD Brief at Hotchips of 2016. But if I am not mistaken(which is often), I believe that it was AMD CTO Mark Papermaster that actually confirmed that information(of the Leapfrog Teams) one year after Mike Clark Spilled The Beans on Zen5 on 2018
According to Mark Papermaster, both the Zen 5 and Zen 4 core designs are being developed by two separate leapfrogging design teams. That was back on 2019 when he said that. I am sure that now on 2022 that the leapfrogging design teams are hard at work on Zen6/Zen7
This is what has always puzzled me. How is Intel and AMD aware of each products Months before we get even a whiff from WCCFTECH and Co? Do they have Spies or an agreement to reveal what they have on a roadmap?I think those working in the business might have more knowledge about what their competitors will release than we do.
I would argue that Intel is NOT number one. Actually, it depends. (and remember, I own BOTH).Just an FYI. I was an early adopter of the Zen architecture. A modest R3 1200 in late 2017 or very early 2018. I was on another forum and an AMD employee showed up on the forum. He said "We are mapped out through Zen 4. By the time Zen 2 is released, we will leave intel in the rear view mirror." AMD didn't leave Intel in the rear view mirror until Zen 3 because of the memory latency issue on Zen 2.
When they say they finished Zen 5. That simply means they finished the design architecture of Zen 5. It doesn't mean Zen 5 is taped out, a finished chip or operational.
I do believe that Zen 4 is totally finished, operational and complete and has been for many months. Market conditions would say that AMD should have moved up the release date for Zen 4.
I know a lot of you here are AMD fanboys. The truth is that AMD is currently in 2nd place behind Intel and Alder Lake. Obviously there is no need to panic because Zen 4 is coming. AMD doesn't act as if they are in 2nd place to Intel.
I don't know about you guys but most serious gamers are playing games mostly on 1440p but some are playing 4k setups. The 5800x3D is great for 1080p games and gets smoked by other AMD CPU's as well as very inexpensive i5 Alder Lake chips in every category other than 1080p gaming.
Something a lot of people forget about Intel. They are on 10nm right now and people do not recognize the efficiency gains over the 14nm process they were on for 6 years. That is truly pathetic. If Intel gets down to 7nm, they will realize even more gains. Obviously AMD will make significant gains with the 5nm process.
I really do not think Intel's roadmap will remain because of Zen 4 and AMD. Which is why AMD needs to accelerate their product execution.
I think AMD wants to be a duopoly with Intel in the CPU arena. I do not think Intel has any long term plans of sharing or losing any more market share to AMD. I know that AMD wants to view itself as a luxury brand. The B550 X570 is a PCI-4 one and done motherboard. I know the AM4 platform lasted forever.
It's time for Zen 4, that's my point. Zen 3 is a year and a half old.
None here is to believe that BS. They actually thought they would be against 10nm and 7nm a lot earlier. Let me find the post.Just an FYI. I was an early adopter of the Zen architecture. A modest R3 1200 in late 2017 or very early 2018. I was on another forum and an AMD employee showed up on the forum. He said "We are mapped out through Zen 4. By the time Zen 2 is released, we will leave intel in the rear view mirror." AMD didn't leave Intel in the rear view mirror until Zen 3 because time for Zen 4, that's my point. Zen 3 is a year and a half old.
Just a guess, but I think the way Intel and AMD are able to predict what each other are up isn't due to owning spies or performing clandestine operations, but it's likely something simpler: they are in the same industry, and that means their employees talk to each other outside of work, their employees attend the same conferences, are on the same committees within their industry, and because they have a much better understanding of the trajectory of the leading fabs , they can extrapolate that information to determine what potential products each other are making and thus plan accordingly.This is what has always puzzled me. How is Intel and AMD aware of each products Months before we get even a whiff from WCCFTECH and Co? Do they have Spies or an agreement to reveal what they have on a roadmap?
Nobody here believes that AMD would have said that type of nonsense (as opposed to certain CEO of Intel as of recently), seeing how they were just coming off such bad years. They had to be cautiously optimistic back on those yearsJust an FYI. I was on another forum and an AMD employee showed up on the forum. He said "We are mapped out through Zen 4. By the time Zen 2 is released, we will leave intel in the rearview mirror."
Category: servers... all AMD. I doubt many would argue that.
Desktop gaming. As of now I would have to say a tie in ability, but cheaper and less heat and power for AMD win -- AMD
Desktop productivity. While ADL is fast for a few cores, for full power on 32 threads, you can't touch the 5950x. -- win -- AMD
Laptops. Not really a good follower, but from what I read, again, its almost a tie. -- win -- Tie with the understanding that my knowledge is limited in this area.
HEDT: All AMD
Not sure how reliable this is, but back in the FX days, as in NVidia, I'd been told by someone online back when message boards were thriving that sometimes these companies send engineering samples to a competitor for some reason. I'm not entirely sure what given that it's been at least 15 years or more since I was told that.I could totally believe the employees of either company talking smack about each other regardless of whether they had any kind of actual knowledge or not.
They may have said it just because Zen 2 moved to TSMC and Intel was still having issues with 10nm. For the first time AMD would actually have a process advantage. Even in the past when AMD did have a good design, Intel had a significant process advantage.
There's also any number of different contexts to claim leaving Intel in the dust. For HEDT and server it was pretty clear that AMD gained an strong advantage. Desktop was more of a split depending on whether you cared more about gaming or productivity.
I could believe that AMD employees were confident enough to boast. After the long years of Bulldozer and its descendants I couldn't blame them for wanting to either.
What do you mean? Can you expound on this?AMD With Leapfrog Design teams executing the road map to de T would still have been at a disadvantage if Intel had been able to bring 10nm on 2018 and 7nm on 2020.
Naples vs Skylake, Rome vs Ice Lake and Milán vs Sapphire Rapids is a different Ball Game than Rome beating Skylake and Milan/Milan-X Curb Stomping Ice Lake. Sapphire Rapids will also be just in time to catch a massive beating from Genoa/BergamoWhat do you mean? Can you expound on this?
I get that but your original post implied that AMD leapfrogging wouldn't be a match for Intel if Intel were capable of bringing their fabs online for high yields. The implication being that the node would make a large difference in performance? You're looking at a 4 year average for a design to come to fruition.Naples vs Skylake, Rome vs Ice Lake and Milán vs Sapphire Rapids is a different Ball Game than Rome beating Skylake and Milan/Milan-X Curb Stomping Ice Lake. Sapphire Rapids will also be just in time to catch a massive beating from Genoa/Bergamo
32 of those small cores is about equal to 8 big cores, so thats the equivalent of 16 strong cores. AMD has 16 pretty strong cores, and I know This redwood/crestmont will be stronger than alder lake, but against Zen 4 ? Or Zen 5 ? I don't know that Intels troubles are over or dominance will be theirs any time soon.I get that but your original post implied that AMD leapfrogging wouldn't be a match for Intel if Intel were capable of bringing their fabs online for high yields. The implication being that the node would make a large difference in performance? You're looking at a 4 year average for a design to come to fruition.
Node aside, the design has to be good. If Intel's fabs weren't delayed, I'm not fully convinced they stood a fighting change against AMD. Alderlake is impressive until you consider how it got there, especially over a 1+ year old processor. Raptor Lake is not going to make a huge difference, IIRC, but Gelsinger stated Meteor Lake is when Intel's bad tides will truly turn around.
That's on Intel 4 (7nm) is going to use the Redwood cove for the Ps and the new Crestmont cores for the Es with a 40 core 8+32 layout. I believe those will be the first desktop processors to use Arc as its iGPU.
I've got more faith in that launching on schedule with the Arc than their discreet cards or later nodes.
Potato, potahto. Novel construction of larger semiconductor solutions, marketing gimmicks.Just for giggles, I decided to check out what UserBenchmark had to say about the 5800X3D. It's as good as I had expected.
View attachment 60116
Well, that sure locks my opinion of that site. Trailer trash reporting. I suppose they would also badmouth the Milan-X, which this is based off of, and Microsoft as well as others have jumped on that bandwagon big time.Just for giggles, I decided to check out what UserBenchmark had to say about the 5800X3D. It's as good as I had expected.
View attachment 60116
I've played around with FS a bit just flying over my own city, and if there is a game you really don't need 100fps for, it is this one (RTS games also apply here). Unless you're right up close to the ground, the terrain moves so gradually underneath you that even at 60fps the experience is surreal (with max settings and HDR).
Intel New it had 10nm troubles before 14nm even began ramping Up. Even a somewhat delayed 10nm Xeon roadmap had Ice Lake X Scheduled release date of about 2019 and Sapphire Rapids at 2021. Lets say that Intel could have pulled that Up. AMD Would have been playing catchup and on the defense most of the time.I get that but your original post implied that AMD leapfrogging wouldn't be a match for Intel if Intel were capable of bringing their fabs online for high yields. The implication being that the node would make a large difference in performance? You're looking at a 4 year average for a design to come to fruition.
Well, that sure locks my opinion of that site. Trailer trash reporting. I suppose they would also badmouth the Milan-X, which this is based off of, and Microsoft as well as others have jumped on that bandwagon big time.
Intel knew it had 10nm problems before 14nm because they'd began working on 10nm years before. 14nm itself was delayed and then had yield problems, just as 22nm had before it. It wasn't until 2017 or 2018 when Intel finally admitted they had become too aggressive over the years with their goals and that's how they got out of lockstep with their teams.Intel New it had 10nm troubles before 14nm even began ramping Up. Even a somewhat delayed 10nm Xeon roadmap had Ice Lake X Scheduled release date of about 2019 and Sapphire Rapids at 2021. Lets say that Intel could have pulled that Up. AMD Would have been playing catchup and on the defense most of the time.