From WikiChip
Difference between revisions of "amd/microarchitectures/zen 3"
< amd‎ | microarchitectures

(Fixed indentation so bullets appear correctly under "Core" and "Front-end")
 
(26 intermediate revisions by 17 users not shown)
Line 7: Line 7:
 
|manufacturer 2=GlobalFoundries
 
|manufacturer 2=GlobalFoundries
 
|introduction=October 8, 2020
 
|introduction=October 8, 2020
|process=7nm
+
|process=7nm, 12nm
|process 2=12 nm
+
|cores=64
 +
|cores 2=56
 +
|cores 3=48
 +
|cores 4=32
 +
|cores 5=28
 +
|cores 6=24
 +
|cores 7=16
 +
|cores 8=12
 +
|cores 9=8
 +
|cores 10=6
 
|type=Superscalar
 
|type=Superscalar
 
|oooe=Yes
 
|oooe=Yes
Line 44: Line 53:
 
|extension 27=UMIP
 
|extension 27=UMIP
 
|extension 28=CLZERO
 
|extension 28=CLZERO
 +
|extension 29=VAES
 +
|extension 30=VPCLMUL
 
|predecessor=Zen 2
 
|predecessor=Zen 2
 
|predecessor link=amd/microarchitectures/zen 2
 
|predecessor link=amd/microarchitectures/zen 2
Line 56: Line 67:
  
 
== Codenames ==
 
== Codenames ==
[[File:amd zen2-3 roadmap.png|400px|right]]
+
[[File:amd zen2-3 roadmap.png|thumb|right|Zen 3 on the roadmap]]
{{future information}}
 
  
 +
'''Product Codenames:'''
 
{| class="wikitable"
 
{| class="wikitable"
 
|-
 
|-
Line 65: Line 76:
 
| {{amd|Milan|l=core}} || Up to 64/128 || High-end server [[multiprocessors]]
 
| {{amd|Milan|l=core}} || Up to 64/128 || High-end server [[multiprocessors]]
 
|-
 
|-
| {{amd|Genesis Peak|l=core}} || ?/? || Workstation & enthusiasts market processors
+
| {{amd|Chagall|l=core}} || Up to 64/128 || Workstation & enthusiasts market processors
 
|-
 
|-
 
| {{amd|Vermeer|l=core}} || Up to 16/32 || Mainstream to high-end desktops & enthusiasts market processors
 
| {{amd|Vermeer|l=core}} || Up to 16/32 || Mainstream to high-end desktops & enthusiasts market processors
 
|-
 
|-
| {{amd|Cezanne|l=core}} || Up to 8/16 || Mainstream desktop & mobile processors with GPU  
+
| {{amd|Cezanne|l=core}} || Up to 8/16 || Mainstream APUs with GPUs
 +
|}
 +
 
 +
'''Architectural Codenames:'''
 +
{| class="wikitable"
 +
|-
 +
! Arch !! Codename
 +
|-
 +
| Core || Cerebrus
 +
|-
 +
| CCD || Breckenridge
 +
|}
 +
 
 +
== Products ==
 +
{{future information}}
 +
 
 +
{| class="wikitable"
 +
|-
 +
! Processor Series !! Cores/Threads !! Market
 +
|-
 +
| EPYC 7003 "{{amd|Milan|l=core}}" || Up to 64/128 || High-end server [[multiprocessors]]
 +
|-
 +
| {{amd|Trento|l=core}}<!--s/a Milan page--> || ?/? || High-performance computing
 +
|-
 +
| Ryzen Threadripper 5900 "{{amd|Chagall|l=core}}" || Up to 64/128 || Workstation processors
 +
|-
 +
| Ryzen 5000 "{{amd|Vermeer|l=core}}" || Up to 16/32 || Mainstream to high-end desktops & enthusiasts market processors
 +
|-
 +
| Ryzen 5000 APU "{{amd|Cezanne|l=core}}" || Up to 8/16 || Mainstream desktop & mobile processors with integrated GPU
 
|}
 
|}
  
 
== Process technology ==
 
== Process technology ==
Zen 3 will be fabricated on [[TSMC]]'s [[7 nm process|7nm+ process]], the same process used in Zen 2 Refresh processors.  
+
Zen 3 is fabricated on [[TSMC]]'s [[7 nm process|7nm+ process]] for the Core Compute Die (CCD), the same process used in Zen 2 Refresh processors, as well as [[GlobalFoundries]] [[14 nm process|12nm process]] for the Input/Output Die (IOD).
 +
 
 +
Note: Only the APU series of microprocessors retains the monolithic design, so they are fabricated solely on [[TSMC]]'s [[7 nm process|7nm+ process]].
 +
 
 +
== Compiler support ==
 +
{| class="wikitable"
 +
|-
 +
! Compiler !! Arch-Specific || Arch-Favorable
 +
|-
 +
| [[GCC]] || <code>-march=znver3</code> || <code>-mtune=znver3</code>
 +
|-
 +
| [[LLVM]] || <code>-march=znver3</code> || <code>-mtune=znver3</code>
 +
|}
 +
* '''Note:''' Initial support in GCC 10.3 and LLVM 12.0.
  
 
== Architecture ==
 
== Architecture ==
  
There is very limited information available about the architectural improvements of Zen 3.
+
=== Key changes from {{\\|Zen 2}} ===
 +
* CCD
 +
** Unified 8-core CCX (from 2x 4-Core CCX per CCD)
 +
** 32 MiB L3$ available equally to all cores in CCD.
 +
*** Increased L3 latency (~46 cycles, up from ~40 cycles)
 +
* Core
 +
** Higher [[IPC]] (AMD self-reported +19% IPC)
 +
** Front-end
 +
*** Increased branch prediction bandwidth
 +
*** "zero-bubble" branch prediction
 +
*** L1 BTB doubled from 512 to 1024 entries
 +
*** Improved prefetching
 +
*** Improved µop cache
 +
** Back-end
 +
*** Floating point unit:
 +
**** FMA latency reduced by 1 cycle from 5 to 4.
 +
**** Fifth and sixth dedicated execution ports added for floating point store and FP-to-int transfer, no longer sharing 2nd FADD port.
 +
**** Unified scheduler split into 1 scheduler per FMA/FADD/transfer port set.
 +
**** 256b VAES and VPCLMULDQ support for doubled AES and AES-GCM cryptographic throughput.
 +
**** Hardware implementation of BMI2 PDEP/PEXT bit scatter/gather operations, compared to prior microcode emulation.
 +
*** Integer unit:
 +
**** Integer physical register file increased from 180 to 192 entries
 +
**** Issue increased from 7 (existing 4 ALU and 3 AGU) to 10 with 1 new dedicated branch execution port and 2 separated store data pathways.
 +
**** Schedulers shared between pairs of ALU + AGU/branch ports instead of dedicated for each.
 +
**** Instruction redundancy increased between ports for reduced bottlenecking on a wider variety of instruction streams.
 +
**** 8/16/32/64 bit signed integer division/modulo latency improved from 17/22/30/46 cycles to 10/12/14/20. (Unsigned operations are ~1 cycle faster for some of both old/new cases.) Throughput improves proportionately.
 +
*** Load/store:
 +
**** Load throughput increased from 2 to 3, if not 256b.
 +
**** Store throughput increased from 1 to 2, if not 256b.
 +
**** Store queue increase from 48 to 64 slots.
 +
**** Page table walkers tripled from 2 to 6 for TLB miss handling.
 +
{{expand list}}
  
=== Key changes from {{\\|Zen 2}} ===
+
=== New Instructions ===
 +
Zen 3 introduced the following ISA enhancements:
 +
 
 +
* {{x86|VAES}} - 256-bit Vector AES instructions
 +
** <code>VAESDEC</code> - AES Decryption Round
 +
** <code>VAESDECLAST</code> - AES Last Decryption Round
 +
** <code>VAESENC</code> - AES Encryption Round
 +
** <code>VAESENCLAST</code> - AES Last Encryption Round
 +
* <code>{{x86|VPCLMULQDQ}}</code> - 256-bit Vector Carry-Less Multiplication of Quadwords
 +
* {{x86|PCID}} - Process Context Identifiers
 +
** <code>{{x86|INVPCID}}</code> - Invalidate TLB entry(s) in a specified PCID
 +
* {{x86|INVLPGB}} - Broadcast TLB flushing
 +
** <code>INVLPGB</code> - Invalidate TLB entry(s) with broadcast to all processors
 +
** <code>TLBSYNC</code> - Synchronize TLB invalidations
 +
* {{x86|PKU}} - Memory Protection Keys for Users
 +
** <code>RDPKRU</code> - Read Protection Key Rights
 +
** <code>WRPKRU</code> - Write Protection Key Rights
 +
* {{x86|CET|CET_SS}} - Control-flow Enforcement Technology / Shadow Stack
 +
** <code>CLRSSBSY</code>, <code>INCSSP</code>, <code>RDSSP</code>, <code>RSTORSSP</code>, <code>SAVEPREVSSP</code>, <code>SETSSBSY</code>, <code>WRSS</code>, <code>WRUSS</code>
 +
* {{x86|SME|SEV-SNP}} - 3rd generation Secure Encrypted Virtualization - Secure Nested Paging
 +
** <code>PSMASH</code>, <code>PVALIDATE</code>, <code>RMPADJUST</code>, <code>RMPUPDATE</code>
 +
* {{x86|PSFD}} - Predictive Store Forwarding Disable (Speculation Control MSR)<ref name="amd-psf"/>
 +
 
 +
Sources: <ref name="amd-24593-apm2"/><ref name="amd-24594-apm3"/><ref name="amd-26568-apm4"/>
 +
 
 +
=== Memory Hierarchy ===
 +
==== Data and Instruction Caches ====
 +
* L0 Op Cache:
 +
** 4,096 Ops per core, 8-way set associative
 +
** 8 Op line size
 +
** Parity protected
 +
* L1I Cache:
 +
** 32 KiB per core, 8-way set associative
 +
** 64 B line size
 +
** Parity protected
 +
* L1D Cache:
 +
** 32 KiB per core, 8-way set associative
 +
** 64 B line size
 +
** Write-back policy
 +
** 4-5 cycles latency for Int
 +
** 7-8 cycles latency for FP
 +
** ECC
 +
* L2 Cache:
 +
** 512 KiB per core, 8-way set associative
 +
** 64 B line size
 +
** Write-back policy
 +
** Inclusive of L1
 +
** ≥ 12 cycles latency
 +
** {{abbr|DEC-TED}} ECC, tag & state arrays {{abbr|SEC-DED}}<!--7 check bits for 42 tag bits; AMD-55898-0.50 Sec 3.5-->
 +
* L3 Cache:
 +
** "{{amd|Milan|l=core}}" & "{{amd|Chagall|l=core}}": 32 MiB/CCX, up to 256 MiB total
 +
** "{{amd|Vermeer|l=core}}": 32 MiB/CCX, up to 64 MiB total
 +
** "{{amd|Cezanne|l=core}}": 16 MiB, 8 MiB usable on some SKUs
 +
** Shared by all cores in the {{abbr|CCX}}, configurable<ref name="amd-56375-qos"/>
 +
** 16-way set associative
 +
** 64 B line size
 +
** L2 [[victim cache]]
 +
** Write-back policy
 +
** 46 cycles average load-to-use latency
 +
** DEC-TED ECC, tag array & shadow tags SEC-DED<!--AMD-55898-0.50 Sec 3.5-->
 +
** QoS Monitoring and Enforcement V2.0
 +
 
 +
==== Translation Lookaside Buffers ====
 +
* ITLB
 +
** 64 entry L1 TLB, fully associative, all page sizes
 +
** 512 entry L2 TLB, 8-way set associative
 +
*** 4-Kbyte and 2-Mbyte pages
 +
** Parity protected
 +
* DTLB
 +
** 64 entry L1 TLB, fully associative, all page sizes
 +
** 2,048 entry L2 TLB, 16-way set associative
 +
*** 4-Kbyte and 2-Mbyte pages, PDEs to speed up table walks
 +
** Parity protected
 +
 
 +
All caches and TLBs are competitively shared in multi-threaded mode.
 +
 
 +
==== System DRAM ====
 +
* EPYC 7003 "{{amd|Milan|l=core}}":
 +
** 8 channels per socket, up to 16 DIMMs, max. 4&nbsp;TiB
 +
** Up to PC4-25600L (DDR4-3200)
 +
** {{abbr|SR}}/{{abbr|DR}} {{abbr|RDIMM}}, {{abbr|4R}}/{{abbr|8R}} {{abbr|LRDIMM}}, {{abbr|3DS DIMM}}, {{abbr|NVDIMM-N}}
 +
** ECC supported (x4, x8, x16, chipkill)<!--AMD-55898-0.50 Sec 3.7-->
 +
** DRAM bus parity and write data CRC options<!--ibid-->
 +
 
 +
* Ryzen Threadripper 5900 "{{amd|Chagall|l=core}}":
 +
** 8 channels, up to 8 DIMMs, max. 2&nbsp;TiB
 +
** Up to PC4-25600L (DDR4-3200)
 +
** SR/DR {{abbr|UDIMM}}, RDIMM, LRDIMM, 3DS DIMM
 +
** ECC supported
 +
 
 +
* Ryzen 5000 "{{amd|Vermeer|l=core}}":
 +
** 2 channels, up to 4 DIMMs, max. 128&nbsp;GiB
 +
** Up to PC4-25600U (DDR4-3200 UDIMM), ECC supported
 +
 
 +
* Ryzen 5000 APU "{{amd|Cezanne|l=core}}":
 +
** {{amd|Socket AM4|l=pack}}:
 +
*** 2 channels, up to 4 DIMMs, max. 128&nbsp;GiB
 +
*** Up to PC4-25600U (DDR4-3200 UDIMM), ECC supported ("PRO" models)
 +
** {{amd|FP6|FP6 package|l=pack}}, DDR4 mode:
 +
*** 2 × 64-bit channels, up to 2 DIMMs, max. 64&nbsp;GiB
 +
*** Up to PC4-25600U (DDR4-3200 UDIMM), ECC supported(?)
 +
** FP6 package, LPDDR4 mode:
 +
*** 4 × 32-bit channels, max. 32&nbsp;GiB
 +
*** Up to LPDDR4X-4266
  
* +19% IPC
+
Sources: <ref name="amd-56375-qos"/><ref name="amd-56665-sog-19h"/><ref name="amd-55898-ppr-1901-0.35"/><ref name="amd-55898-ppr-1901-0.50"/><ref name="amd-56178-mdg-fp6"/>
* Unified 8-core CCX with 32 MiB L3$ available to all 8 cores equally. Latency increased by roughly 7 cycles (18%) to an average of 46 cycles.
 
* Integer unit:
 
** Integer physical register file increased from 180 to 192 entries
 
** Issue increased from 7 (existing 4 ALU and 3 AGU) to 10 with 1 new dedicated branch execution port and 2 separated store data pathways.
 
** Schedulers shared between pairs of ALU + AGU/branch ports instead of dedicated for each.
 
** Instruction redundancy increased between ports for reduced bottlenecking on a wider variety of instruction streams.
 
** 8/16/32/64 bit signed integer division/modulo latency improved from 17/22/30/46 cycles to 10/12/14/20. (Unsigned operations are ~1 cycle faster for some of both old/new cases.) Throughput improves proportionately.
 
* Floating point unit:
 
** FMA latency reduced by 1 cycle from 5 to 4.
 
** Fifth and sixth dedicated execution ports added for floating point store and FP-to-int transfer, no longer sharing 2nd FADD port.
 
** Unified scheduler split into 1 scheduler per FMA/FADD/transfer port set.
 
** 256b VAES and VPCLMULDQ support for doubled AES and AES-GCM cryptographic throughout.
 
** Hardware implementation of BMI2 PDEP/PEXT bit scatter/gather operations, compared to prior microcode emulation.
 
* Load/store:
 
** Load throughput increased from 2 to 3, if not 256b.
 
** Store throughput increased from 1 to 2, if not 256b.
 
** Store queue increase from 48 to 64 slots.
 
** Page table walkers tripled from 2 to 6 for TLB miss handling.
 
* Improved prefetching
 
* Increased branch prediction bandwidth
 
** "zero-bubble" branch prediction
 
** L1 BTB doubled from 512 to 1024 entries
 
* Improved µop cache
 
{{expand list}}
 
  
 
== All Zen 3 Chips ==
 
== All Zen 3 Chips ==
  
<!-- NOTE:  
+
<!-- NOTE:
 
           This table is generated automatically from the data in the actual articles.
 
           This table is generated automatically from the data in the actual articles.
 
           If a microprocessor is missing from the list, an appropriate article for it needs to be
 
           If a microprocessor is missing from the list, an appropriate article for it needs to be
Line 185: Line 347:
 
* AMD 'Tech Day', February 22, 2017
 
* AMD 'Tech Day', February 22, 2017
 
* AMD 2017 Financial Analyst Day, May 16, 2017
 
* AMD 2017 Financial Analyst Day, May 16, 2017
 +
 +
== References ==
 +
<references>
 +
<ref name="amd-psf">{{cite techdoc|title=White Paper: Security Analysis of AMD Predictive Store Forwarding|url=https://www.amd.com/system/files/documents/security-analysis-predictive-store-forwarding.pdf|publ=AMD|date=2021-03}}</ref>
 +
<ref name="amd-24593-apm2">{{cite techdoc|title=AMD64 Architecture Programmer’s Manual Volume 2: System Programming|url=https://www.amd.com/system/files/TechDocs/24593.pdf|publ=AMD|pid=24593|rev=3.37|date=2021-03}}</ref>
 +
<ref name="amd-24594-apm3">{{cite techdoc|title=AMD64 Architecture Programmer’s Manual Volume 3: General-Purpose and System Instructions|url=https://www.amd.com/system/files/TechDocs/24594.pdf|publ=AMD|pid=24594|rev=3.32|date=2021-03}}</ref>
 +
<ref name="amd-26568-apm4">{{cite techdoc|title=AMD64 Architecture Programmer’s Manual Volume 4: 128-Bit and 256-Bit Media Instructions|url=https://www.amd.com/system/files/TechDocs/26568.pdf|publ=AMD|pid=26568|rev=3.24|date=2020-05}}</ref>
 +
<ref name="amd-56375-qos">{{cite techdoc|title=AMD64 Technology Platform Quality of Service Extensions|url=https://developer.amd.com/wp-content/resources/56375.pdf|publ=AMD|pid=56375|rev=1.02|date=2020-10}}</ref>
 +
<ref name="amd-56665-sog-19h">{{cite techdoc|title=Software Optimization Guide for AMD Family 19h Processors (PUB)|url=https://www.amd.com/system/files/TechDocs/56665.zip|publ=AMD|pid=56665|rev=3.00|date=2020-11}}</ref>
 +
<ref name="amd-55898-ppr-1901-0.35">{{cite techdoc|title=Preliminary Processor Programming Reference (PPR) for AMD Family 19h Model 01h, Revision B1 Processors|url=https://www.amd.com/system/files/TechDocs/55898_pub.zip|publ=AMD|pid=55898|rev=0.35|date=2021-02-05}}</ref>
 +
<ref name="amd-55898-ppr-1901-0.50">{{cite techdoc|title=Preliminary Processor Programming Reference (PPR) for AMD Family 19h Model 01h, Revision B1 Processors|publ=AMD|pid=55898|rev=0.50|date=2021-05-27}}</ref>
 +
<ref name="amd-56178-mdg-fp6">{{cite techdoc|title=FP6 Processor Motherboard Design Guide|publ=AMD|pid=56178|rev=1.03|date=2020-01}}</ref>
 +
</references>
  
 
== See Also ==
 
== See Also ==
* AMD {{\\|Zen}}
+
* AMD {{\\|Zen}}, {{\\|Zen 2}}, {{\\|Zen 4}}
 
* Intel {{intel|Tigerlake|l=arch}}
 
* Intel {{intel|Tigerlake|l=arch}}
 
* Read also: [https://www.anandtech.com/print/16214/amd-zen-3-ryzen-deep-dive-review-5950x-5900x-5800x-and-5700x-tested AMD Zen 3 Ryzen Deep Dive Review]
 
* Read also: [https://www.anandtech.com/print/16214/amd-zen-3-ryzen-deep-dive-review-5950x-5900x-5800x-and-5700x-tested AMD Zen 3 Ryzen Deep Dive Review]
 +
* Read here: [https://techmotherboard.com/best-zen-3-cpu/ AMD Zen 3 Reviews]

Latest revision as of 21:13, 6 April 2023

Edit Values
Zen 3 µarch
General Info
Arch TypeCPU
DesignerAMD
ManufacturerTSMC, GlobalFoundries
IntroductionOctober 8, 2020
Process7nm, 12nm
Core Configs64, 56, 48, 32, 28, 24, 16, 12, 8, 6
Pipeline
TypeSuperscalar
OoOEYes
SpeculativeYes
Reg RenamingYes
Stages19
Decode4-way
Instructions
ISAx86-64
ExtensionsMOVBE, MMX, SSE, SSE2, SSE3, SSSE3, SSE4A, SSE4.1, SSE4.2, POPCNT, AVX, AVX2, AES, PCLMUL, FSGSBASE, RDRND, FMA3, F16C, BMI, BMI2, RDSEED, ADCX, PREFETCHW, CLFLUSHOPT, XSAVE, SHA, UMIP, CLZERO, VAES, VPCLMUL
Succession

Zen 3 is a microarchitecture developed by AMD as a successor to Zen 2. It was publicly released on October 8, 2020. Mainstream Desktop processors hit shelves on November 5, 2020.

History[edit]

amd zen future roadmap.jpg

Zen 3 was formally disclosed in a roadmap by Lisa Su, AMD's CEO, during AMD's Tech Day in February of 2017. Zen 3 will be the 3rd iteration of the Zen microarchitecture. On Investor's Day in May 2017 Jim Anderson, AMD Senior Vice President, confirmed that Zen 3 is set to utilize 7nm+ process.

Codenames[edit]

Zen 3 on the roadmap

Product Codenames:

Core C/T Target
Milan Up to 64/128 High-end server multiprocessors
Chagall Up to 64/128 Workstation & enthusiasts market processors
Vermeer Up to 16/32 Mainstream to high-end desktops & enthusiasts market processors
Cezanne Up to 8/16 Mainstream APUs with GPUs

Architectural Codenames:

Arch Codename
Core Cerebrus
CCD Breckenridge

Products[edit]

Symbol version future.svg Preliminary Data! Information presented in this article deal with future products, data, features, and specifications that have yet to be finalized, announced, or released. Information may be incomplete and can change by final release.
Processor Series Cores/Threads Market
EPYC 7003 "Milan" Up to 64/128 High-end server multiprocessors
Trento  ?/? High-performance computing
Ryzen Threadripper 5900 "Chagall" Up to 64/128 Workstation processors
Ryzen 5000 "Vermeer" Up to 16/32 Mainstream to high-end desktops & enthusiasts market processors
Ryzen 5000 APU "Cezanne" Up to 8/16 Mainstream desktop & mobile processors with integrated GPU

Process technology[edit]

Zen 3 is fabricated on TSMC's 7nm+ process for the Core Compute Die (CCD), the same process used in Zen 2 Refresh processors, as well as GlobalFoundries 12nm process for the Input/Output Die (IOD).

Note: Only the APU series of microprocessors retains the monolithic design, so they are fabricated solely on TSMC's 7nm+ process.

Compiler support[edit]

Compiler Arch-Specific Arch-Favorable
GCC -march=znver3 -mtune=znver3
LLVM -march=znver3 -mtune=znver3
  • Note: Initial support in GCC 10.3 and LLVM 12.0.

Architecture[edit]

Key changes from Zen 2[edit]

  • CCD
    • Unified 8-core CCX (from 2x 4-Core CCX per CCD)
    • 32 MiB L3$ available equally to all cores in CCD.
      • Increased L3 latency (~46 cycles, up from ~40 cycles)
  • Core
    • Higher IPC (AMD self-reported +19% IPC)
    • Front-end
      • Increased branch prediction bandwidth
      • "zero-bubble" branch prediction
      • L1 BTB doubled from 512 to 1024 entries
      • Improved prefetching
      • Improved µop cache
    • Back-end
      • Floating point unit:
        • FMA latency reduced by 1 cycle from 5 to 4.
        • Fifth and sixth dedicated execution ports added for floating point store and FP-to-int transfer, no longer sharing 2nd FADD port.
        • Unified scheduler split into 1 scheduler per FMA/FADD/transfer port set.
        • 256b VAES and VPCLMULDQ support for doubled AES and AES-GCM cryptographic throughput.
        • Hardware implementation of BMI2 PDEP/PEXT bit scatter/gather operations, compared to prior microcode emulation.
      • Integer unit:
        • Integer physical register file increased from 180 to 192 entries
        • Issue increased from 7 (existing 4 ALU and 3 AGU) to 10 with 1 new dedicated branch execution port and 2 separated store data pathways.
        • Schedulers shared between pairs of ALU + AGU/branch ports instead of dedicated for each.
        • Instruction redundancy increased between ports for reduced bottlenecking on a wider variety of instruction streams.
        • 8/16/32/64 bit signed integer division/modulo latency improved from 17/22/30/46 cycles to 10/12/14/20. (Unsigned operations are ~1 cycle faster for some of both old/new cases.) Throughput improves proportionately.
      • Load/store:
        • Load throughput increased from 2 to 3, if not 256b.
        • Store throughput increased from 1 to 2, if not 256b.
        • Store queue increase from 48 to 64 slots.
        • Page table walkers tripled from 2 to 6 for TLB miss handling.

This list is incomplete; you can help by expanding it.

New Instructions[edit]

Zen 3 introduced the following ISA enhancements:

  • VAES - 256-bit Vector AES instructions
    • VAESDEC - AES Decryption Round
    • VAESDECLAST - AES Last Decryption Round
    • VAESENC - AES Encryption Round
    • VAESENCLAST - AES Last Encryption Round
  • VPCLMULQDQ - 256-bit Vector Carry-Less Multiplication of Quadwords
  • PCID - Process Context Identifiers
    • INVPCID - Invalidate TLB entry(s) in a specified PCID
  • INVLPGB - Broadcast TLB flushing
    • INVLPGB - Invalidate TLB entry(s) with broadcast to all processors
    • TLBSYNC - Synchronize TLB invalidations
  • PKU - Memory Protection Keys for Users
    • RDPKRU - Read Protection Key Rights
    • WRPKRU - Write Protection Key Rights
  • CET_SS - Control-flow Enforcement Technology / Shadow Stack
    • CLRSSBSY, INCSSP, RDSSP, RSTORSSP, SAVEPREVSSP, SETSSBSY, WRSS, WRUSS
  • SEV-SNP - 3rd generation Secure Encrypted Virtualization - Secure Nested Paging
    • PSMASH, PVALIDATE, RMPADJUST, RMPUPDATE
  • PSFD - Predictive Store Forwarding Disable (Speculation Control MSR)[1]

Sources: [2][3][4]

Memory Hierarchy[edit]

Data and Instruction Caches[edit]

  • L0 Op Cache:
    • 4,096 Ops per core, 8-way set associative
    • 8 Op line size
    • Parity protected
  • L1I Cache:
    • 32 KiB per core, 8-way set associative
    • 64 B line size
    • Parity protected
  • L1D Cache:
    • 32 KiB per core, 8-way set associative
    • 64 B line size
    • Write-back policy
    • 4-5 cycles latency for Int
    • 7-8 cycles latency for FP
    • ECC
  • L2 Cache:
    • 512 KiB per core, 8-way set associative
    • 64 B line size
    • Write-back policy
    • Inclusive of L1
    • ≥ 12 cycles latency
    • DEC-TED ECC, tag & state arrays SEC-DED
  • L3 Cache:
    • "Milan" & "Chagall": 32 MiB/CCX, up to 256 MiB total
    • "Vermeer": 32 MiB/CCX, up to 64 MiB total
    • "Cezanne": 16 MiB, 8 MiB usable on some SKUs
    • Shared by all cores in the CCX, configurable[5]
    • 16-way set associative
    • 64 B line size
    • L2 victim cache
    • Write-back policy
    • 46 cycles average load-to-use latency
    • DEC-TED ECC, tag array & shadow tags SEC-DED
    • QoS Monitoring and Enforcement V2.0

Translation Lookaside Buffers[edit]

  • ITLB
    • 64 entry L1 TLB, fully associative, all page sizes
    • 512 entry L2 TLB, 8-way set associative
      • 4-Kbyte and 2-Mbyte pages
    • Parity protected
  • DTLB
    • 64 entry L1 TLB, fully associative, all page sizes
    • 2,048 entry L2 TLB, 16-way set associative
      • 4-Kbyte and 2-Mbyte pages, PDEs to speed up table walks
    • Parity protected

All caches and TLBs are competitively shared in multi-threaded mode.

System DRAM[edit]

  • EPYC 7003 "Milan":
    • 8 channels per socket, up to 16 DIMMs, max. 4 TiB
    • Up to PC4-25600L (DDR4-3200)
    • SR/DR RDIMM, 4R/8R LRDIMM, 3DS DIMM, NVDIMM-N
    • ECC supported (x4, x8, x16, chipkill)
    • DRAM bus parity and write data CRC options
  • Ryzen Threadripper 5900 "Chagall":
    • 8 channels, up to 8 DIMMs, max. 2 TiB
    • Up to PC4-25600L (DDR4-3200)
    • SR/DR UDIMM, RDIMM, LRDIMM, 3DS DIMM
    • ECC supported
  • Ryzen 5000 "Vermeer":
    • 2 channels, up to 4 DIMMs, max. 128 GiB
    • Up to PC4-25600U (DDR4-3200 UDIMM), ECC supported
  • Ryzen 5000 APU "Cezanne":
    • Socket AM4:
      • 2 channels, up to 4 DIMMs, max. 128 GiB
      • Up to PC4-25600U (DDR4-3200 UDIMM), ECC supported ("PRO" models)
    • FP6 package, DDR4 mode:
      • 2 × 64-bit channels, up to 2 DIMMs, max. 64 GiB
      • Up to PC4-25600U (DDR4-3200 UDIMM), ECC supported(?)
    • FP6 package, LPDDR4 mode:
      • 4 × 32-bit channels, max. 32 GiB
      • Up to LPDDR4X-4266

Sources: [5][6][7][8][9]

All Zen 3 Chips[edit]

 List of all Zen 3-based Processors
 ProcessorFeatures
ModelPriceProcessLaunchedFamilyCoreCTTDPL3BaseTurboMax MemSMTSEVSMETSME
 Uniprocessors
7313P$ 913.00
€ 821.70
£ 739.53
¥ 94,340.29
15 March 2021EPYCMilan1632155 W
155,000 mW
0.208 hp
0.155 kW
128 MiB
131,072 KiB
134,217,728 B
0.125 GiB
3 GHz
3,000 MHz
3,000,000 kHz
3.7 GHz
3,700 MHz
3,700,000 kHz
4,096 GiB
4,194,304 MiB
4,294,967,296 KiB
4,398,046,511,104 B
4 TiB
7443P$ 1,337.00
€ 1,203.30
£ 1,082.97
¥ 138,152.21
15 March 2021EPYCMilan2448200 W
200,000 mW
0.268 hp
0.2 kW
128 MiB
131,072 KiB
134,217,728 B
0.125 GiB
2.85 GHz
2,850 MHz
2,850,000 kHz
4 GHz
4,000 MHz
4,000,000 kHz
4,096 GiB
4,194,304 MiB
4,294,967,296 KiB
4,398,046,511,104 B
4 TiB
7543P$ 2,730.00
€ 2,457.00
£ 2,211.30
¥ 282,090.90
15 March 2021EPYCMilan3264225 W
225,000 mW
0.302 hp
0.225 kW
256 MiB
262,144 KiB
268,435,456 B
0.25 GiB
2.8 GHz
2,800 MHz
2,800,000 kHz
3.7 GHz
3,700 MHz
3,700,000 kHz
4,096 GiB
4,194,304 MiB
4,294,967,296 KiB
4,398,046,511,104 B
4 TiB
7713P$ 5,010.00
€ 4,509.00
£ 4,058.10
¥ 517,683.30
15 March 2021EPYCMilan64128225 W
225,000 mW
0.302 hp
0.225 kW
256 MiB
262,144 KiB
268,435,456 B
0.25 GiB
2 GHz
2,000 MHz
2,000,000 kHz
3.675 GHz
3,675 MHz
3,675,000 kHz
4,096 GiB
4,194,304 MiB
4,294,967,296 KiB
4,398,046,511,104 B
4 TiB
5300G7 nm
0.007 μm
7.0e-6 mm
13 April 2021Ryzen 3Cezanne4865 W
65,000 mW
0.0872 hp
0.065 kW
8 MiB
8,192 KiB
8,388,608 B
0.00781 GiB
4 GHz
4,000 MHz
4,000,000 kHz
4.2 GHz
4,200 MHz
4,200,000 kHz
5300GE7 nm
0.007 μm
7.0e-6 mm
13 April 2021Ryzen 3Cezanne4835 W
35,000 mW
0.0469 hp
0.035 kW
8 MiB
8,192 KiB
8,388,608 B
0.00781 GiB
3.6 GHz
3,600 MHz
3,600,000 kHz
4.2 GHz
4,200 MHz
4,200,000 kHz
5400U7 nm
0.007 μm
7.0e-6 mm
12 January 2021Ryzen 3Cezanne4815 W
15,000 mW
0.0201 hp
0.015 kW
8 MiB
8,192 KiB
8,388,608 B
0.00781 GiB
2.6 GHz
2,600 MHz
2,600,000 kHz
4 GHz
4,000 MHz
4,000,000 kHz
PRO 5350G7 nm
0.007 μm
7.0e-6 mm
1 June 2021Ryzen 3Cezanne4865 W
65,000 mW
0.0872 hp
0.065 kW
8 MiB
8,192 KiB
8,388,608 B
0.00781 GiB
4 GHz
4,000 MHz
4,000,000 kHz
4.2 GHz
4,200 MHz
4,200,000 kHz
PRO 5350GE7 nm
0.007 μm
7.0e-6 mm
1 June 2021Ryzen 3Cezanne4835 W
35,000 mW
0.0469 hp
0.035 kW
8 MiB
8,192 KiB
8,388,608 B
0.00781 GiB
3.6 GHz
3,600 MHz
3,600,000 kHz
4.2 GHz
4,200 MHz
4,200,000 kHz
PRO 5450U7 nm
0.007 μm
7.0e-6 mm
16 March 2021Ryzen 3Cezanne4815 W
15,000 mW
0.0201 hp
0.015 kW
8 MiB
8,192 KiB
8,388,608 B
0.00781 GiB
2.6 GHz
2,600 MHz
2,600,000 kHz
4 GHz
4,000 MHz
4,000,000 kHz
5600G$ 259.00
€ 233.10
£ 209.79
¥ 26,762.47
7 nm
0.007 μm
7.0e-6 mm
13 April 2021Ryzen 5Cezanne61265 W
65,000 mW
0.0872 hp
0.065 kW
16 MiB
16,384 KiB
16,777,216 B
0.0156 GiB
3.9 GHz
3,900 MHz
3,900,000 kHz
4.4 GHz
4,400 MHz
4,400,000 kHz
5600GE7 nm
0.007 μm
7.0e-6 mm
13 April 2021Ryzen 5Cezanne61235 W
35,000 mW
0.0469 hp
0.035 kW
16 MiB
16,384 KiB
16,777,216 B
0.0156 GiB
3.4 GHz
3,400 MHz
3,400,000 kHz
4.4 GHz
4,400 MHz
4,400,000 kHz
5600H7 nm
0.007 μm
7.0e-6 mm
12 January 2021Ryzen 5Cezanne61245 W
45,000 mW
0.0603 hp
0.045 kW
16 MiB
16,384 KiB
16,777,216 B
0.0156 GiB
3.3 GHz
3,300 MHz
3,300,000 kHz
4.2 GHz
4,200 MHz
4,200,000 kHz
5600HS7 nm
0.007 μm
7.0e-6 mm
12 January 2021Ryzen 5Cezanne61235 W
35,000 mW
0.0469 hp
0.035 kW
16 MiB
16,384 KiB
16,777,216 B
0.0156 GiB
3 GHz
3,000 MHz
3,000,000 kHz
4.2 GHz
4,200 MHz
4,200,000 kHz
5600U7 nm
0.007 μm
7.0e-6 mm
12 January 2021Ryzen 5Cezanne61215 W
15,000 mW
0.0201 hp
0.015 kW
16 MiB
16,384 KiB
16,777,216 B
0.0156 GiB
2.3 GHz
2,300 MHz
2,300,000 kHz
4.2 GHz
4,200 MHz
4,200,000 kHz
5600X$ 299.00
€ 269.10
£ 242.19
¥ 30,895.67
7 nm
0.007 μm
7.0e-6 mm
, 12 nm
0.012 μm
1.2e-5 mm
5 November 2020Ryzen 5Vermeer61265 W
65,000 mW
0.0872 hp
0.065 kW
32 MiB
32,768 KiB
33,554,432 B
0.0313 GiB
3.7 GHz
3,700 MHz
3,700,000 kHz
4.6 GHz
4,600 MHz
4,600,000 kHz
128 GiB
131,072 MiB
134,217,728 KiB
137,438,953,472 B
0.125 TiB
PRO 5650G7 nm
0.007 μm
7.0e-6 mm
1 June 2021Ryzen 5Cezanne61265 W
65,000 mW
0.0872 hp
0.065 kW
16 MiB
16,384 KiB
16,777,216 B
0.0156 GiB
3.9 GHz
3,900 MHz
3,900,000 kHz
4.4 GHz
4,400 MHz
4,400,000 kHz
PRO 5650GE7 nm
0.007 μm
7.0e-6 mm
1 June 2021Ryzen 5Cezanne61235 W
35,000 mW
0.0469 hp
0.035 kW
16 MiB
16,384 KiB
16,777,216 B
0.0156 GiB
3.4 GHz
3,400 MHz
3,400,000 kHz
4.4 GHz
4,400 MHz
4,400,000 kHz
PRO 5650U7 nm
0.007 μm
7.0e-6 mm
16 March 2021Ryzen 5Cezanne61215 W
15,000 mW
0.0201 hp
0.015 kW
16 MiB
16,384 KiB
16,777,216 B
0.0156 GiB
2.3 GHz
2,300 MHz
2,300,000 kHz
4.2 GHz
4,200 MHz
4,200,000 kHz
5700G$ 359.00
€ 323.10
£ 290.79
¥ 37,095.47
7 nm
0.007 μm
7.0e-6 mm
13 April 2021Ryzen 7Cezanne81665 W
65,000 mW
0.0872 hp
0.065 kW
16 MiB
16,384 KiB
16,777,216 B
0.0156 GiB
3.8 GHz
3,800 MHz
3,800,000 kHz
4.6 GHz
4,600 MHz
4,600,000 kHz
5700GE7 nm
0.007 μm
7.0e-6 mm
13 April 2021Ryzen 7Cezanne81635 W
35,000 mW
0.0469 hp
0.035 kW
16 MiB
16,384 KiB
16,777,216 B
0.0156 GiB
3.2 GHz
3,200 MHz
3,200,000 kHz
4.6 GHz
4,600 MHz
4,600,000 kHz
58007 nm
0.007 μm
7.0e-6 mm
, 12 nm
0.012 μm
1.2e-5 mm
12 January 2021Ryzen 7Vermeer81665 W
65,000 mW
0.0872 hp
0.065 kW
32 MiB
32,768 KiB
33,554,432 B
0.0313 GiB
3.4 GHz
3,400 MHz
3,400,000 kHz
4.6 GHz
4,600 MHz
4,600,000 kHz
128 GiB
131,072 MiB
134,217,728 KiB
137,438,953,472 B
0.125 TiB
5800H7 nm
0.007 μm
7.0e-6 mm
12 January 2021Ryzen 7Cezanne81645 W
45,000 mW
0.0603 hp
0.045 kW
16 MiB
16,384 KiB
16,777,216 B
0.0156 GiB
3.2 GHz
3,200 MHz
3,200,000 kHz
4.4 GHz
4,400 MHz
4,400,000 kHz
5800HS7 nm
0.007 μm
7.0e-6 mm
12 January 2021Ryzen 7Cezanne81635 W
35,000 mW
0.0469 hp
0.035 kW
16 MiB
16,384 KiB
16,777,216 B
0.0156 GiB
2.8 GHz
2,800 MHz
2,800,000 kHz
4.4 GHz
4,400 MHz
4,400,000 kHz
5800U7 nm
0.007 μm
7.0e-6 mm
12 January 2021Ryzen 7Cezanne81615 W
15,000 mW
0.0201 hp
0.015 kW
16 MiB
16,384 KiB
16,777,216 B
0.0156 GiB
1.9 GHz
1,900 MHz
1,900,000 kHz
4.4 GHz
4,400 MHz
4,400,000 kHz
5800X$ 449.00
€ 404.10
£ 363.69
¥ 46,395.17
7 nm
0.007 μm
7.0e-6 mm
, 12 nm
0.012 μm
1.2e-5 mm
5 November 2020Ryzen 7Vermeer816105 W
105,000 mW
0.141 hp
0.105 kW
32 MiB
32,768 KiB
33,554,432 B
0.0313 GiB
3.8 GHz
3,800 MHz
3,800,000 kHz
4.7 GHz
4,700 MHz
4,700,000 kHz
128 GiB
131,072 MiB
134,217,728 KiB
137,438,953,472 B
0.125 TiB
5800X3D$ 449.00
€ 404.10
£ 363.69
¥ 46,395.17
7 nm
0.007 μm
7.0e-6 mm
, 12 nm
0.012 μm
1.2e-5 mm
20 April 2022Ryzen 7Vermeer816105 W
105,000 mW
0.141 hp
0.105 kW
96 MiB
98,304 KiB
100,663,296 B
0.0938 GiB
3.4 GHz
3,400 MHz
3,400,000 kHz
4.5 GHz
4,500 MHz
4,500,000 kHz
128 GiB
131,072 MiB
134,217,728 KiB
137,438,953,472 B
0.125 TiB
PRO 5750G7 nm
0.007 μm
7.0e-6 mm
1 June 2021Ryzen 7Cezanne81665 W
65,000 mW
0.0872 hp
0.065 kW
16 MiB
16,384 KiB
16,777,216 B
0.0156 GiB
3.8 GHz
3,800 MHz
3,800,000 kHz
4.6 GHz
4,600 MHz
4,600,000 kHz
PRO 5750GE7 nm
0.007 μm
7.0e-6 mm
1 June 2021Ryzen 7Cezanne81635 W
35,000 mW
0.0469 hp
0.035 kW
16 MiB
16,384 KiB
16,777,216 B
0.0156 GiB
3.2 GHz
3,200 MHz
3,200,000 kHz
4.6 GHz
4,600 MHz
4,600,000 kHz
PRO 5850U7 nm
0.007 μm
7.0e-6 mm
16 March 2021Ryzen 7Cezanne81615 W
15,000 mW
0.0201 hp
0.015 kW
16 MiB
16,384 KiB
16,777,216 B
0.0156 GiB
1.9 GHz
1,900 MHz
1,900,000 kHz
4.4 GHz
4,400 MHz
4,400,000 kHz
59007 nm
0.007 μm
7.0e-6 mm
, 12 nm
0.012 μm
1.2e-5 mm
12 January 2021Ryzen 9Vermeer122465 W
65,000 mW
0.0872 hp
0.065 kW
64 MiB
65,536 KiB
67,108,864 B
0.0625 GiB
3 GHz
3,000 MHz
3,000,000 kHz
4.7 GHz
4,700 MHz
4,700,000 kHz
128 GiB
131,072 MiB
134,217,728 KiB
137,438,953,472 B
0.125 TiB
5900HS7 nm
0.007 μm
7.0e-6 mm
12 January 2021Ryzen 9Cezanne81635 W
35,000 mW
0.0469 hp
0.035 kW
16 MiB
16,384 KiB
16,777,216 B
0.0156 GiB
3 GHz
3,000 MHz
3,000,000 kHz
4.6 GHz
4,600 MHz
4,600,000 kHz
5900HX7 nm
0.007 μm
7.0e-6 mm
12 January 2021Ryzen 9Cezanne81645 W
45,000 mW
0.0603 hp
0.045 kW
16 MiB
16,384 KiB
16,777,216 B
0.0156 GiB
3.3 GHz
3,300 MHz
3,300,000 kHz
4.6 GHz
4,600 MHz
4,600,000 kHz
5900X$ 549.00
€ 494.10
£ 444.69
¥ 56,728.17
7 nm
0.007 μm
7.0e-6 mm
, 12 nm
0.012 μm
1.2e-5 mm
5 November 2020Ryzen 9Vermeer1224105 W
105,000 mW
0.141 hp
0.105 kW
64 MiB
65,536 KiB
67,108,864 B
0.0625 GiB
3.7 GHz
3,700 MHz
3,700,000 kHz
4.8 GHz
4,800 MHz
4,800,000 kHz
128 GiB
131,072 MiB
134,217,728 KiB
137,438,953,472 B
0.125 TiB
5950X$ 799.00
€ 719.10
£ 647.19
¥ 82,560.67
7 nm
0.007 μm
7.0e-6 mm
, 12 nm
0.012 μm
1.2e-5 mm
5 November 2020Ryzen 9Vermeer1632105 W
105,000 mW
0.141 hp
0.105 kW
64 MiB
65,536 KiB
67,108,864 B
0.0625 GiB
3.4 GHz
3,400 MHz
3,400,000 kHz
4.9 GHz
4,900 MHz
4,900,000 kHz
128 GiB
131,072 MiB
134,217,728 KiB
137,438,953,472 B
0.125 TiB
5980HS7 nm
0.007 μm
7.0e-6 mm
12 January 2021Ryzen 9Cezanne81635 W
35,000 mW
0.0469 hp
0.035 kW
16 MiB
16,384 KiB
16,777,216 B
0.0156 GiB
3 GHz
3,000 MHz
3,000,000 kHz
4.8 GHz
4,800 MHz
4,800,000 kHz
5980HX7 nm
0.007 μm
7.0e-6 mm
12 January 2021Ryzen 9Cezanne81645 W
45,000 mW
0.0603 hp
0.045 kW
16 MiB
16,384 KiB
16,777,216 B
0.0156 GiB
3.3 GHz
3,300 MHz
3,300,000 kHz
4.8 GHz
4,800 MHz
4,800,000 kHz
PRO 5945WX7 nm
0.007 μm
7.0e-6 mm
8 March 2022Ryzen ThreadripperChagall1224280 W
280,000 mW
0.375 hp
0.28 kW
64 MiB
65,536 KiB
67,108,864 B
0.0625 GiB
4.1 GHz
4,100 MHz
4,100,000 kHz
4.5 GHz
4,500 MHz
4,500,000 kHz
2,048 GiB
2,097,152 MiB
2,147,483,648 KiB
2,199,023,255,552 B
2 TiB
PRO 5955WX7 nm
0.007 μm
7.0e-6 mm
8 March 2022Ryzen ThreadripperChagall1632280 W
280,000 mW
0.375 hp
0.28 kW
64 MiB
65,536 KiB
67,108,864 B
0.0625 GiB
4 GHz
4,000 MHz
4,000,000 kHz
4.5 GHz
4,500 MHz
4,500,000 kHz
2,048 GiB
2,097,152 MiB
2,147,483,648 KiB
2,199,023,255,552 B
2 TiB
PRO 5965WX7 nm
0.007 μm
7.0e-6 mm
8 March 2022Ryzen ThreadripperChagall2448280 W
280,000 mW
0.375 hp
0.28 kW
128 MiB
131,072 KiB
134,217,728 B
0.125 GiB
3.8 GHz
3,800 MHz
3,800,000 kHz
4.5 GHz
4,500 MHz
4,500,000 kHz
2,048 GiB
2,097,152 MiB
2,147,483,648 KiB
2,199,023,255,552 B
2 TiB
PRO 5975WX7 nm
0.007 μm
7.0e-6 mm
8 March 2022Ryzen ThreadripperChagall3264280 W
280,000 mW
0.375 hp
0.28 kW
128 MiB
131,072 KiB
134,217,728 B
0.125 GiB
3.6 GHz
3,600 MHz
3,600,000 kHz
4.5 GHz
4,500 MHz
4,500,000 kHz
2,048 GiB
2,097,152 MiB
2,147,483,648 KiB
2,199,023,255,552 B
2 TiB
PRO 5995WX7 nm
0.007 μm
7.0e-6 mm
8 March 2022Ryzen ThreadripperChagall64128280 W
280,000 mW
0.375 hp
0.28 kW
256 MiB
262,144 KiB
268,435,456 B
0.25 GiB
2.7 GHz
2,700 MHz
2,700,000 kHz
4.5 GHz
4,500 MHz
4,500,000 kHz
2,048 GiB
2,097,152 MiB
2,147,483,648 KiB
2,199,023,255,552 B
2 TiB
 Multiprocessors (dual-socket)
72F3$ 2,468.00
€ 2,221.20
£ 1,999.08
¥ 255,018.44
15 March 2021EPYCMilan816180 W
180,000 mW
0.241 hp
0.18 kW
256 MiB
262,144 KiB
268,435,456 B
0.25 GiB
3.7 GHz
3,700 MHz
3,700,000 kHz
4.1 GHz
4,100 MHz
4,100,000 kHz
4,096 GiB
4,194,304 MiB
4,294,967,296 KiB
4,398,046,511,104 B
4 TiB
7313$ 1,083.00
€ 974.70
£ 877.23
¥ 111,906.39
15 March 2021EPYCMilan1632155 W
155,000 mW
0.208 hp
0.155 kW
128 MiB
131,072 KiB
134,217,728 B
0.125 GiB
3 GHz
3,000 MHz
3,000,000 kHz
3.7 GHz
3,700 MHz
3,700,000 kHz
4,096 GiB
4,194,304 MiB
4,294,967,296 KiB
4,398,046,511,104 B
4 TiB
7343$ 1,565.00
€ 1,408.50
£ 1,267.65
¥ 161,711.45
15 March 2021EPYCMilan1632190 W
190,000 mW
0.255 hp
0.19 kW
128 MiB
131,072 KiB
134,217,728 B
0.125 GiB
3.2 GHz
3,200 MHz
3,200,000 kHz
3.9 GHz
3,900 MHz
3,900,000 kHz
4,096 GiB
4,194,304 MiB
4,294,967,296 KiB
4,398,046,511,104 B
4 TiB
73F3$ 3,521.00
€ 3,168.90
£ 2,852.01
¥ 363,824.93
15 March 2021EPYCMilan1632240 W
240,000 mW
0.322 hp
0.24 kW
256 MiB
262,144 KiB
268,435,456 B
0.25 GiB
3.5 GHz
3,500 MHz
3,500,000 kHz
4 GHz
4,000 MHz
4,000,000 kHz
4,096 GiB
4,194,304 MiB
4,294,967,296 KiB
4,398,046,511,104 B
4 TiB
7413$ 1,825.00
€ 1,642.50
£ 1,478.25
¥ 188,577.25
15 March 2021EPYCMilan2448180 W
180,000 mW
0.241 hp
0.18 kW
128 MiB
131,072 KiB
134,217,728 B
0.125 GiB
2.65 GHz
2,650 MHz
2,650,000 kHz
3.6 GHz
3,600 MHz
3,600,000 kHz
4,096 GiB
4,194,304 MiB
4,294,967,296 KiB
4,398,046,511,104 B
4 TiB
7443$ 2,010.00
€ 1,809.00
£ 1,628.10
¥ 207,693.30
15 March 2021EPYCMilan2448200 W
200,000 mW
0.268 hp
0.2 kW
128 MiB
131,072 KiB
134,217,728 B
0.125 GiB
2.85 GHz
2,850 MHz
2,850,000 kHz
4 GHz
4,000 MHz
4,000,000 kHz
4,096 GiB
4,194,304 MiB
4,294,967,296 KiB
4,398,046,511,104 B
4 TiB
7453$ 1,570.00
€ 1,413.00
£ 1,271.70
¥ 162,228.10
15 March 2021EPYCMilan2856225 W
225,000 mW
0.302 hp
0.225 kW
64 MiB
65,536 KiB
67,108,864 B
0.0625 GiB
2.75 GHz
2,750 MHz
2,750,000 kHz
3.45 GHz
3,450 MHz
3,450,000 kHz
4,096 GiB
4,194,304 MiB
4,294,967,296 KiB
4,398,046,511,104 B
4 TiB
74F3$ 2,900.00
€ 2,610.00
£ 2,349.00
¥ 299,657.00
15 March 2021EPYCMilan2448240 W
240,000 mW
0.322 hp
0.24 kW
256 MiB
262,144 KiB
268,435,456 B
0.25 GiB
3.2 GHz
3,200 MHz
3,200,000 kHz
4 GHz
4,000 MHz
4,000,000 kHz
4,096 GiB
4,194,304 MiB
4,294,967,296 KiB
4,398,046,511,104 B
4 TiB
7513$ 2,840.00
€ 2,556.00
£ 2,300.40
¥ 293,457.20
15 March 2021EPYCMilan3264200 W
200,000 mW
0.268 hp
0.2 kW
128 MiB
131,072 KiB
134,217,728 B
0.125 GiB
2.6 GHz
2,600 MHz
2,600,000 kHz
3.65 GHz
3,650 MHz
3,650,000 kHz
4,096 GiB
4,194,304 MiB
4,294,967,296 KiB
4,398,046,511,104 B
4 TiB
7543$ 3,761.00
€ 3,384.90
£ 3,046.41
¥ 388,624.13
15 March 2021EPYCMilan3264225 W
225,000 mW
0.302 hp
0.225 kW
256 MiB
262,144 KiB
268,435,456 B
0.25 GiB
2.8 GHz
2,800 MHz
2,800,000 kHz
3.7 GHz
3,700 MHz
3,700,000 kHz
4,096 GiB
4,194,304 MiB
4,294,967,296 KiB
4,398,046,511,104 B
4 TiB
75F3$ 4,860.00
€ 4,374.00
£ 3,936.60
¥ 502,183.80
15 March 2021EPYCMilan3264280 W
280,000 mW
0.375 hp
0.28 kW
256 MiB
262,144 KiB
268,435,456 B
0.25 GiB
2.95 GHz
2,950 MHz
2,950,000 kHz
4 GHz
4,000 MHz
4,000,000 kHz
4,096 GiB
4,194,304 MiB
4,294,967,296 KiB
4,398,046,511,104 B
4 TiB
7643$ 4,995.00
€ 4,495.50
£ 4,045.95
¥ 516,133.35
15 March 2021EPYCMilan4896225 W
225,000 mW
0.302 hp
0.225 kW
256 MiB
262,144 KiB
268,435,456 B
0.25 GiB
2.3 GHz
2,300 MHz
2,300,000 kHz
3.6 GHz
3,600 MHz
3,600,000 kHz
4,096 GiB
4,194,304 MiB
4,294,967,296 KiB
4,398,046,511,104 B
4 TiB
7663$ 6,366.00
€ 5,729.40
£ 5,156.46
¥ 657,798.78
15 March 2021EPYCMilan56112240 W
240,000 mW
0.322 hp
0.24 kW
256 MiB
262,144 KiB
268,435,456 B
0.25 GiB
2 GHz
2,000 MHz
2,000,000 kHz
3.5 GHz
3,500 MHz
3,500,000 kHz
4,096 GiB
4,194,304 MiB
4,294,967,296 KiB
4,398,046,511,104 B
4 TiB
7713$ 7,060.00
€ 6,354.00
£ 5,718.60
¥ 729,509.80
15 March 2021EPYCMilan64128225 W
225,000 mW
0.302 hp
0.225 kW
256 MiB
262,144 KiB
268,435,456 B
0.25 GiB
2 GHz
2,000 MHz
2,000,000 kHz
3.675 GHz
3,675 MHz
3,675,000 kHz
4,096 GiB
4,194,304 MiB
4,294,967,296 KiB
4,398,046,511,104 B
4 TiB
7763$ 7,890.00
€ 7,101.00
£ 6,390.90
¥ 815,273.70
15 March 2021EPYCMilan64128280 W
280,000 mW
0.375 hp
0.28 kW
256 MiB
262,144 KiB
268,435,456 B
0.25 GiB
2.45 GHz
2,450 MHz
2,450,000 kHz
3.5 GHz
3,500 MHz
3,500,000 kHz
4,096 GiB
4,194,304 MiB
4,294,967,296 KiB
4,398,046,511,104 B
4 TiB
Count: 57

Designers[edit]

  • Mark Evers, Chief Architect

Bibliography[edit]

  • AMD 'Tech Day', February 22, 2017
  • AMD 2017 Financial Analyst Day, May 16, 2017

References[edit]

  1. "White Paper: Security Analysis of AMD Predictive Store Forwarding", AMD, March 2021
  2. "AMD64 Architecture Programmer’s Manual Volume 2: System Programming", AMD Publ. #24593, Rev. 3.37, March 2021
  3. "AMD64 Architecture Programmer’s Manual Volume 3: General-Purpose and System Instructions", AMD Publ. #24594, Rev. 3.32, March 2021
  4. "AMD64 Architecture Programmer’s Manual Volume 4: 128-Bit and 256-Bit Media Instructions", AMD Publ. #26568, Rev. 3.24, May 2020
  5. 5.0 5.1 "AMD64 Technology Platform Quality of Service Extensions", AMD Publ. #56375, Rev. 1.02, October 2020
  6. "Software Optimization Guide for AMD Family 19h Processors (PUB)", AMD Publ. #56665, Rev. 3.00, November 2020
  7. "Preliminary Processor Programming Reference (PPR) for AMD Family 19h Model 01h, Revision B1 Processors", AMD Publ. #55898, Rev. 0.35, February 5, 2021
  8. "Preliminary Processor Programming Reference (PPR) for AMD Family 19h Model 01h, Revision B1 Processors", AMD Publ. #55898, Rev. 0.50, May 27, 2021
  9. "FP6 Processor Motherboard Design Guide", AMD Publ. #56178, Rev. 1.03, January 2020

See Also[edit]

codenameZen 3 +
core count64 +, 56 +, 48 +, 32 +, 28 +, 24 +, 16 +, 12 +, 8 + and 6 +
designerAMD +
first launchedOctober 8, 2020 +
full page nameamd/microarchitectures/zen 3 +
instance ofmicroarchitecture +
instruction set architecturex86-64 +
manufacturerTSMC + and GlobalFoundries +
microarchitecture typeCPU +
nameZen 3 +
pipeline stages19 +