News Two-Die Xeon? Leaked Sapphire Rapids Photo Appears to Show Chiplets

intel really should stop trying to put down amd since they seem to do exactly that later on.
Yeah - it's almost becoming a meme.
  • 486 - clocking it higher is expensive. AMD pulls out the DX4-120.
  • Pentium / K6 : Socket 7 is dead at 233 MHz. AMD pulls out the K6-III+ at up to 570 MHz officially.
  • Pentium III : 1 GHz and over is not stable with a short pipelined architecture. AMD pulls out the Athlon.
  • Pentium 4 : Rambus and high frequency is the future. AMD beats it to a pulp with Athlon XP and DDR.
  • Itanium : x86 can't do 64-bit properly. AMD pulls out x86-64.
  • Core 2 : integrated memory controller is useless. AMD enables dual channel DDR/DDR2 on consumer-level stuff (sockets s939 and AM2) - the IMC was already there on all their x86-64 chips.
  • Core 2 Quad up to 7th generation Core i : more than quad core is useless for daily computing. AMD pulls out six-core Phenom, and (much later, Bulldozer was a major blunder in the PC space, although it worked on consoles) Ryzen 1700.
  • Up until now : thread safety is a gimmick. AMD : pulls out Bulldozer for consoles.
  • Up until now : glued cores are bad for performance. AMD pulls out Threadripper (up to 2 dies did work quite nicely) then Zen 2-3 (up to 8 dies in a single package).
What's next? BAR address range resizing is only for big businesses? ECC DDR has no use in the consumer space? Weak embedded graphics are good enough for desktop users?
Oct 5, 2020
As much as this is good news, I feel that Intel (and AMD for that matter but especially Intel) will eventually move back to a monolithic die.......... unless something else happens. ..............


Mar 31, 2014
I highly doubt intel is using chiplets. What they usually do is to combine 2 dies to double the core count
Yes, this.

I bet each of those 'chiplets' is about twice as large as an entire Ryzen CPU. This is probably more like the Pentium D 2021 edition. ;)


Sep 2, 2019
I highly doubt intel is using chiplets. What they usually do is to combine 2 dies to double the core count
In the end this doesn't matter. It is a MCM design one way or the other.
Additionally this would give them the opportunity to double memory channels (and probably also PCIe lanes), if they intend to (?), as they have already done with Cascade Lake AP with 12 memory channels.
Finally it is very likely, that this design (if it's no fake) will simply be an extension to their expertise with Cascade Lake AP.
The main question is: How efficient will be their third process iteration 10nm+++ aka Enhanced SuperFin because this will limit core scaling in the end. The design will be released after Milan and most likely outperform it (in special workloads even by far), but it will be released shortly before Genoa in 5nm, which puts it in a difficult position (because originally Granite Rapids SP in 7nm should have been the competitor for Genoa, but with the 7nm delay, this server will arrive not until the first half of 2023).
Yes, this.
I bet each of those 'chiplets' is about twice as large as an entire Ryzen CPU. This is probably more like the Pentium D 2021 edition. ;)
Of course it will be larger. Already Ice Lake SP will provide 32 to 36 (?) cores on a monolithic die, therefore of course it is significantly larger than a Ryzen CPU with only up to 16 cores. :rolleyes:

Btw: 3900/50X and 5900/50X require ~ 273 mm2 die area.
Skylake SP (in 14nm) as HCC with up to 18 cores has about ~ 428 mm2, therefore you can estimate that an HCC for Ice Lake SP or Sapphire Rapids SP will be significantly smaller and therefore still larger but far away from doubling the area (compared to a Ryzen).

But if you had something different in mind, a 64 core Rome/Milan/Threadripper has a die area of about ~ 1008 mm2. This also means, that a 64 core Sapphire Rapids SP most likely *) will only require a few square millimeters more (if at all) than Milan, because Intel produces these die(s) completely in 10nm whereas AMD still uses a very large 14nm IO-die (the IO-die for Ryzen's is produced in 12nm, the one for Rome in 14nm; currently it is unclear if AMD had possibly shrinked it to 12nm for Milan).
A 32 core Epyc/Threadripper currently requires at least ~ 712 mm2 with four chiplets, therefore it is safe to assume, that a 32 core Intel die will require less area. Already a 28 core Skylake SP in 14nm only requires 602 mm2.

*) "Most likely" because it also depends on the amount of increased caches and possibly "much broader" cores.
Last edited:


May 31, 2019
I highly doubt intel is using chiplets.
Intel presented their intentions to use smaller chiplets for what they call Client 2.0.

AMD's Norrod explained how they need to move to 3D mfg. for performance.
See his presentation video and discussion,38838.html

Intel and TSM both are developing hybrid bonding. Intel announced a hybrid-bonded SRAM test chip recently.



May 23, 2015
Is it just me or does the pinout on the CPU appear to be designed similar to what you'd see in a dual socket board? It seems like it's literally 2 separate CPUs glued together to fit into 1 oversized socket. So just imagine a dual socket setup designed to work with one glued together cpu. So no infinity fabric design on the CPU itself.


Feb 27, 2020
Reactions: HyperMatrix