Movatterモバイル変換


[0]ホーム

URL:


Jump to content
WikipediaThe Free Encyclopedia
Search

Goldmont

From Wikipedia, the free encyclopedia
CPU microarchitecture used in Intel SoCs
Goldmont
General information
Product code
  • 80668 (Apollo Lake)
  • 80765 (Denverton)
Architecture and classification
Technology node14 nm
Instructionsx86-64,Intel 64
Extensions
Physical specifications
Cores
  • 2–4
Products, models, variants
Brand names
History
PredecessorAirmont (die shrink)
SuccessorGoldmont Plus (optimization)

Goldmont is amicroarchitecture for low-powerAtom,Celeron andPentium branded processors used insystems on a chip (SoCs) made byIntel. They allow only onethread percore.

TheApollo Lake platform with 14 nm Goldmont core was unveiled at theIntel Developer Forum (IDF) inShenzhen, China, April 2016.[1] The Goldmont architecture borrows heavily from theSkylake Core processors, so it offers a more than 30 percent performance boost compared to the previousBraswell platform, and it can be used to implement power-efficient low-end devices includingCloudbooks, 2-in-1netbooks, small PCs,IP cameras, andin-car entertainment systems.[2][3]

Design

[edit]

Goldmont is the 2nd generationout-of-order low-powerAtom microarchitecture designed for the entry level desktop and notebook computers.[4] Goldmont is built on the 14 nm manufacturing process and supports up to four cores for the consumer devices. It includes theIntel Gen9 graphics architecture introduced with theSkylake.

The Goldmont microarchitecture builds on the success of the Silvermont microarchitecture, and provides the following enhancements:

  • An out-of-order execution engine with a 3-widesuperscalarpipeline.[5] Specifically:
    • The decoder can decode 3 instructions per cycle.
    • The microcode sequencer can send 3μops per cycle for allocation into the reservation stations.
    • Retirement supports a peak rate of 3 per cycle.
  • Enhancement inbranch prediction which de-couples the fetch pipeline from the instruction decoder.
  • Larger out-of-order execution window and buffers that enable deeper out-of-order execution acrossinteger,FP/SIMD, and memory instruction types.
  • Fully out-of-order memory execution and disambiguation. The Goldmont microarchitecture can execute one load and one store per cycle (compared to one load or one store per cycle in the Silvermont microarchitecture). The memory execution pipeline also includes a second levelTLB enhancement with 512 entries for 4KBpages.
  • Integer execution cluster in the Goldmont microarchitecture provides three pipelines and can execute up to three simple integerALU operations per cycle.
  • SIMD integer and floating-point instructions execute in a 128-bit wide engine. Throughput and latency of many instructions have improved, including PSHUFB with 1-cycle throughput (versus 5 cycles for Silvermont microarchitecture) and many other SIMD instructions with doubled throughput.
  • Throughput and latency of instructions for accelerating encryption/decryption (AES) andcarry-less multiplication (PCLMULQDQ) have been improved significantly in the Goldmont microarchitecture.
  • The Goldmont microarchitecture provides new instructions with hardware accelerated secure hashing algorithm,SHA1 andSHA256.
  • The Goldmont microarchitecture also adds support for theRDSEED instruction for random number generation meeting theNIST SP800-90C standard.
  • PAUSE instruction latency is optimized to enable better power efficiency.

Technology

[edit]
See also:Intel HD and Iris Graphics

Erratum

[edit]

Similar to the previous Silvermont generation, design flaws were found in processor circuitry, resulting in cease of operation when processors are actively used for several years. An Erratum named APL46 "System May Experience Inability to Boot or May Cease Operation"[8] was added to documentation in June 2017, stating that low pin count (LPC), real time clock (RTC), SD card and GPIO interfaces may stop functioning.

Mitigations[9] were found to limit impact on systems. A firmware update for the LPC bus called LPC_CLKRUN# reduces the utilization of the LPC interface, which in turn decreases (but does not eliminate) LPC bus degradation – some systems are however not compatible with this new firmware. It is recommended not to use SD cards as boot devices, and to remove the card from the system when not in use; other possible solutions being using only UHS-I cards and operating them at 1.8 V.

Congatec also states the issues impact USB buses and eMMC, although those are not mentioned in Intel's public documentation. USB should have a maximum of 12% active time and there is a 60TB transmit traffic life expectancy over the lifetime of the port. eMMC should have a maximum of 33% active time and should be set to D3 device low power state by the operating system when not in use. Newer designs such as Atom C3000 Denverton do not seem to be affected.[10]

List of Goldmont processors

[edit]

Desktop processors (Apollo Lake)

[edit]

List of desktop processors as follows:[3][11]

Target
segment
Cores
(threads = cores)
Processor
branding & model
GPU modelTDP

(W)

CPU freq.

(GHz)

GPU freq.

(MHz)

L2
cache
Release
date
Price
(USD)
Brand name &
model number
EUBaseTurboBaseTurbo
Desktop4 (4)PentiumJ4205HD Graphics 50518101.52.62508002 MBQ3 2016$161
CeleronJ3455HD Graphics 500122.3750$107
2 (2)J33552.02.5700

Server processors (Denverton)

[edit]
Target
segment
Cores

(threads)

Processor
branding & model
TDP

(W)

CPU freq.

(GHz)

L2
cache
DDR4 speedRelease
date
Price
(USD)
BaseTurbo
Server16 (16)AtomC3958[12][13]312.016 MB2400Q3 2017$449
16 (16)C3955[12]322.12.4$434
12 (12)C3858[12]252.012 MB$332
12 (12)C3850[12]2.12.4$323
12 (12)C3830[12]211.92.32133$289
12 (12)C3808[12]252.0$369
8 (8)C3758[12]2.216 MB2400$193
8 (8)C3750[12]212.22.4$171
8 (8)C3708[12]171.72133$209
4 (4)C3558[12]162.28 MB$86
4 (4)C3538[12]152.1$75
4 (4)C3508[12]11.251.61866$86
2 (2)C3338[12]91.52.24 MBQ1 2017$27
2 (2)C3308[12]9.51.62.1Q3 2017$32

Mobile processors (Apollo Lake)

[edit]

List of mobile processors as follows:[3][11]

Target
segment
Cores
(Threads)
Processor
branding & model
GPU modelTDP

(W)

CPU freq.

(GHz)

GPU freq.

(MHz)

L2
cache
Release
date
Price
(USD)
Brand name &
model number
EUBaseTurboBaseTurbo
Mobile4 (4)PentiumN4200HD Graphics 5051861.12.52007502 MBQ3 2016$161
CeleronN3450HD Graphics 500122.2700$107
2 (2)N33502.4650

Embedded processors (Apollo Lake)

[edit]

List of embedded processors as follows:

Target
segment
Cores
(threads)
Processor
branding & model
GPU modelTDP

(W)

CPU freq.

(GHz)

GPU freq.

(MHz)

L2
cache
Release
date
Price
(USD)
Brand name &
model number
EUBaseTurboBaseTurbo
Embedded4 (4)Atom x7E3950HD Graphics 50518121.62.05006502 MBQ3 2016?
Atom x5E3940HD Graphics 500129.51.8400600
2 (2)E39306.51.3550

Automotive processors (Apollo Lake)

[edit]

There is also an Atom A3900 series exclusively for automotive customers withAEC-Q100 qualification:[14]

Target
segment
Cores
(threads)
Processor
branding & model
GPU modelTDP

(W)

CPU freq.

(GHz)

GPU freq.

(MHz)

L2
cache
Release
date
Price
(USD)
Brand name &
model number
EUBaseTurboNormalTurbo
Automotive4 (4)Atom x7A3960HD Graphics 5051812.51.92.46007502 MB?
A39509.51.62.0500650
Atom x5A3940HD Graphics 5001281.8400600
2 (2)A393061.3550

Tablet processors (Willow Trail)

[edit]

Willow Trail platform was canceled. Apollo Lake will be offered instead.[15]

See also

[edit]

References

[edit]
  1. ^Eric Brown (2016-04-18).""Apollo Lake" Atoms to offer graphics-rich Goldmont cores". Hackerboards.com. Archived fromthe original on 2016-06-20. Retrieved2016-06-11.
  2. ^Anton Shilov (2016-04-15)."Intel Unveils New Low-Cost PC Platform: Apollo Lake with 14nm Goldmont Cores". Anandtech.com. Retrieved2016-06-11.
  3. ^abcAlexander Fagot/ Allen Ngo (2016-06-07)."Intel claims Apollo Lake will be 30 percent faster than Braswell". Notebookcheck.net. Retrieved2016-06-11.
  4. ^Anton Shilov (2015-06-10)."Intel preps 'Apollo Lake' CPUs with 'Goldmont' cores, Gen9 graphics". Retrieved2016-06-11.
  5. ^Kanter, David."Goldmont Takes Atom to 14nm".The Linley Group. The Linley Group. Retrieved17 August 2017.
  6. ^"Release Notes : DRIVER VERSION: 31.0.101.2115"(PDF).Downloadmirror.intel.com. Retrieved29 December 2022.
  7. ^"Mesa 13.0 Released With Intel OpenGL 4.5, RADV Radeon Vulkan Driver - Phoronix".Phoronix.com.
  8. ^"Intel Pentium and Celeron Processor N- and J- Series Specification Update"(PDF). Retrieved2018-04-13.
  9. ^"Errata sheet - congatec Apollo Lake designs"(PDF).congatec. 2017-07-26. Archived fromthe original(PDF) on 2018-04-15. Retrieved2018-04-15.
  10. ^"Intel Atom Processor C3000 Product Family Specification Update"(PDF). Retrieved2018-04-13.
  11. ^abGennadiy Shvets (2016-06-09)."Model numbers of Apollo Lake processors revealed". cpu-world.com. Retrieved2016-06-11.
  12. ^abcdefghijklmnKennedy, Patrick (15 August 2017)."Intel Atom C3000 Series Launch SKUs and Differentiation". Serve the Home. Retrieved15 August 2017.
  13. ^Cutress, Ian (15 August 2017)."More Denverton Noise: Gigabyte's MA1-ST0 Features Unannounced 16-Core C3958". Anandtech. Retrieved15 August 2017.
  14. ^"Intel Atom Processor E3900 and A3900 Series Datasheet Addendum"(PDF). Intel. Retrieved12 January 2018.
  15. ^Ryan Smith & Ian Cutress (2016-04-29)."Intel's Changing Future: Smartphone SoCs Broxton & SoFIA Officially Cancelled". Anandtech.com.
Lists
Microarchitectures
IA-32 (32-bit x86)
x86-64 (64-bit)
x86ULV
Current products
x86-64 (64-bit)
Discontinued
BCD oriented (4-bit)
pre-x86 (8-bit)
Earlyx86 (16-bit)
x87 (externalFPUs)
8/16-bit databus
8087 (1980)
16-bit databus
80C187
80287
80387SX
32-bit databus
80387DX
80487
IA-32 (32-bit x86)
x86-64 (64-bit)
Other
Related
Intel CPU core roadmaps fromP6 to Panther Lake
Atom (ULV)Node namePentium/Core
Microarch.StepMicroarch.Step
600 nmP6Pentium Pro
(133 MHz)
500 nmPentium Pro
(150 MHz)
350 nmPentium Pro
(166–200 MHz)
Klamath
250 nmDeschutes
KatmaiNetBurst
180 nmCoppermineWillamette
130 nmTualatinNorthwood
Pentium MBaniasNetBurst(HT)NetBurst(×2)
90 nmDothanPrescottPrescott‑2MSmithfield
TejasCedarmill (Tejas)
65 nmYonahNehalem (NetBurst)Cedar MillPresler
CoreMerom4 cores on mainstream desktop,DDR3 introduced
BonnellBonnell45 nmPenryn
NehalemNehalemHT reintroduced, integratedMC, PCH
L3-cache introduced, 256KB L2-cache/core
Saltwell32 nmWestmereIntroduced GPU on same package andAES-NI
Sandy BridgeSandy BridgeOn-die ring bus, no more non-UEFI motherboards
SilvermontSilvermont22 nmIvy Bridge
HaswellHaswellFully integrated voltage regulator
Airmont14 nmBroadwell
SkylakeSkylakeDDR4 introduced on mainstream desktop
GoldmontGoldmontKaby Lake
Coffee Lake6 cores on mainstream desktop
Amber LakeMobile-only
Goldmont PlusGoldmont PlusWhiskey LakeMobile-only
Coffee Lake Refresh8 cores on mainstream desktop
Comet Lake10 cores on mainstream desktop
Sunny CoveCypress Cove (Rocket Lake)Backported Sunny Cove microarchitecture for 14nm
TremontTremont10 nmSkylakePalm Cove (Cannon Lake)Mobile-only
Sunny CoveSunny Cove (Ice Lake)512 KB L2-cache/core
Willow Cove (Tiger Lake)Xe graphics engine
GracemontGracemontIntel 7
(10nm ESF)
Golden CoveGolden Cove (Alder Lake)Hybrid, DDR5, PCIe 5.0
Raptor Cove (Raptor Lake)
CrestmontCrestmontIntel 4Redwood CoveMeteor LakeMobile-only
NPU,chiplet architecture
SkymontSkymontN3B (TSMC)Lion CoveLunar LakeLow power mobile only (9-30W)
Arrow Lake
DarkmontDarkmont18ACougar CovePanther Lake
  • Strike-through indicates cancelled processors
  • Bold names are microarchitectures
  • Italic names are future processors
Retrieved from "https://en.wikipedia.org/w/index.php?title=Goldmont&oldid=1254433308"
Categories:
Hidden categories:

[8]ページ先頭

©2009-2025 Movatter.jp