Hardware leaker claims that AMD may limit its Ryzen 7000 3D V-Cache models to just two six-core and eight-core models.
Ryzen 7000 3D V-Cache CPUs Could Be Limited To 6-Core, 8-Core Configurations : Read more
The first irony of this 3D V-Cache article being...
When AMD figured out how great 3D V-Cache was with the 5800X3D, why didn't they incorporated that into the baseline design of the next generation of CPUs?
Oh. Right. Why build it better
(or standardized) from the start when you can milk everyone along? Got it.
That aside, the inescapable truth about AMD Zen4 desktop processors remains; and in light of these "leaks"
as witnesses view the various samples that we all knew were coming roll off the end of the production line, needs to said:
The continuation of employing the current, completely counter-productive AM5 Integrated Heat Spreader (IHS) packaging and mounting positions dictated by AM4 designs, now seems simply unfathomable; but worse, under the ridiculous premise, "
to maintain backward compatibility with AM4 cooling" especially with 3D V-Cache as the existing designs have quickly
(in fact, all too quickly) demonstrated to yield significant and measurable negative impacts across every engineering path relative to each "new" underlying technology being employed both by the CPU and motherboard and component OEMs and what bothers me most is that everybody (and I do mean
everybody) knows it; but nobody is willing to say anything publicly.
Put simply,
PCIe 5 (now over 3 1/2 years old) and DDR5 are each being significantly constrained as their specifications are being forced to accommodate AMD's AM4 socket positioning and dimension issues as well as AM5 IHS and core relative height, width and depth positions inside those AM4 socket physical constraints in addition to the AM4 socket's exterior constraints to accommodate AM4 cooler mounting positions.
Herein lies another problem in that AM5 specifications for associated underlying technologies never actually considered: Necessitating incompatible AM4 pathways, and in fact they assumed those paths would be clear of physical obstructions to support the shorter trace paths
required to support the aforementioned new PCIe and DDR Standards and for their accompanying components and devices not to mention other pins for things like 10gb+ LAN, PCIe 5 M.2, Thunderbolt 4, USB 4, etc..) in accordance with their published Standards. Instead, what consumers ended up getting were motherboards forced to piece together pathways outside their normal design specifications with systems overheating from the CPU outward while stressing every single component that resides on the CPU itself, on the motherboard, in each PCIe slot or otherwise occupying space within the system chassis while drawing much more power than intended by design, only to hit a thermal throttle that itself then serves to define the baseline parameters of the CPU and as well as other component performances all while generating more heat than planned and using signifantly more power than would otherwise be necessary to achieve an even higher planned baseline performance that would be more stable, even when utilizing all of the exact motherboard, RAM, Storage, PCIe and power supply hardware components.
The lengths to which the supply chain OEMs have gone to achieve what they already have given these constraints has been extreme; but people investing in any X3D CPU or motherboard should pay particular attention to these facts as that's a pretty damn big investment and again, there's a lot more going on here than just an IHS issue.
From an engineering perspective, everything about this is ludicrous including the fact that the AM4 coolers themselves were never built to accommodate a CPU IHS running at 90c over such a large contact area for such extended periods of time, rendering them ineffective to the task in the first place!
For these basic "engineering 101" rational reasons, until AMD drops this ridiculous IHS and AM4 compatibility nonsense, I will be skipping the entire AM5 generation of products as this kind of investment falls into the exact same category of why I'll never buy a 40 Series Nvidia Card:
I don't want my house to burn down just because some idiot had a stupid idea that should have been considered and shot-down in the first engineering meeting or in the worst case scenario, at least tested on a bench after a sample was made so everyone could say at once, "Yeah, this would be a pretty stupid thing to do!".
You don't build a V-8 Engine with mounts designed only to fit inside a Tesla Frunk, and then insist that all car makers employing it only use a chainsaw exhaust pipe that must stick out through the sunroof. That's what AMD did here and it's also why you'll see Intel platform hardware yielding better PCIe 5 and DDR component performance, with fewer errors, lower latency and using less power to do it with longer lasting parts that are unstressed by AMD's self-imposed design constraints and these lessons that Intel will not be forced to learn will carry over to Intel having better PCIe 6 (yeah, it was finalized almost a year ago) and DDR6 while AMD systems languish behind self-imposed design constraints that are pointless, expensive both in the short and long-term for consumers. There's a lot more going on here than just the negative impacts of having a thicker IHS that exists for all the wrong reasons, a lot more.
Will Zen4 laptop designs be forced down some similar dumb rabbit hole? Only time will tell.
Facebook
Twitter
Instagram