Hubbry Logo
Cascade LakeCascade LakeMain
Open search
Cascade Lake
Community hub
Cascade Lake
logo
8 pages, 0 posts
0 subscribers
Be the first to start a discussion here.
Be the first to start a discussion here.
Cascade Lake
Cascade Lake
Cascade Lake
View on Wikipedia
from Wikipedia
For other uses, see Cascade Lake (disambiguation).
Cascade Lake
Xeon W-3235 with 12 cores
General information
LaunchedApril 2, 2019; 6 years ago (April 2, 2019)
Marketed byIntel
Designed byIntel
Common manufacturer
  • Intel
Product code80695
Performance
Max. CPU clock rateUp to 4.8 GHz
QPI speeds9.6 GT/s to 10.4 GT/s
DMI speeds8 GT/s
Cache
L1 cache64 KB per core (32 instructions + 32 data)
L2 cache1 MB per core
L3 cacheUp to 77 MB (1.375 MiB/core)
Architecture and classification
Technology node14 nm (Tri-Gate) transistors
MicroarchitectureSkylake
Instruction setx86-64
InstructionsMMX, SSE, SSE2, SSE3, SSSE3, SSE4, SSE4.1, SSE4.2, AVX, AVX2, FMA3, AVX-512,
Extensions
Physical specifications
Cores
  • 4-56
Sockets
  • LGA 2066 (Cascade Lake-X)
  • LGA 3647 (Cascade Lake-SP and Cascade Lake-W)
  • BGA 5903 (Cascade Lake-AP)
Products, models, variants
Product code name
  • CSL
Models
  • Cascade Lake-X (enthusiast)
  • Cascade Lake-SP (Scalable Performance)
  • Cascade Lake-W (workstation)
  • Cascade Lake-AP (Advanced Performance)
Brand name
History
PredecessorSkylake
Successors1S & 2S Systems

4S & 8S Systems

Support status
Xeon Scalable: Discontinued as of October 2, 2023[1]
Xeon W-2200 and Core X-Series: Discontinued after April 26, 2024[2]

Cascade Lake is an Intel codename for a 14 nm server, workstation and enthusiast processor generation, launched in April 2019.[3][4] In Intel's process–architecture–optimization model, Cascade Lake is an optimization of Skylake.[5][6][7][8][9] Intel states that this will be their first generation to support 3D XPoint-based memory modules.[10] It also features Deep Learning Boost (DPL) instructions and mitigations for Meltdown and Spectre.[4][11] Intel officially launched new Xeon Scalable SKUs on February 24, 2020.[12]

Variants

[edit]
  • Server: Cascade Lake-SP, Cascade Lake-AP
  • Workstation: Cascade Lake-W
  • Enthusiast: Cascade Lake-X

List of Cascade Lake processors

[edit]

[13]

Cascade Lake-X (HEDT)

[edit]
Name sSpec number
(stepping) 		Cores
(threads) 		Base
clock 		Turbo Boost 2.0 Turbo
Boost
3.0 		Memory
support 		Socket Optane
memory
support 		Cache PCIe
3.0
lanes 		TDP Release
date 		Part number(s) Price
(USD)
Single core All cores L2 L3

Core i9-10980XE SRGSG (L1) 18 (36) 3.0 GHz 4.6 GHz 3.8 GHz 4.8 GHz 4 × DDR4-2933
up to 256 GiB 		LGA2066 Yes 1 MB
per core 		24.75 MB 48 165 W 25 November 2019[14] CD8069504381800
BXC8069510980XE 		$979-$1000
Core i9-10940X SRGSH (L1) 14 (28) 3.3 GHz 4.1 GHz 19.25 MB CD8069504381900
BX8069510940X 		$784-$797
Core i9-10920X SRGSJ (L1) 12 (24) 3.5 GHz 4.3 GHz CD8069504382000
BX8069510920X 		$689-$700
Core i9-10900X SRGV7 (L1) 10 (20) 3.7 GHz 4.5 GHz 4.7 GHz CD8069504382100
BX8069510900X 		$590-$599

Cascade Lake-AP (Advanced Performance)

[edit]

Cascade Lake-AP is branded as Xeon Platinum 9200 series and all SKUs are soldered to the motherboard. These CPUs will not work with Optane Memory.

Xeon Platinum 9200 series

[edit]
Model number sSpec number Cores (threads) Frequency Turbo Boost L2 Cache L3 Cache TDP Socket I/O Bus Memory Release date Part number(s) Release Price
9282 56 (112) 2.60 GHz 3.80 GHz 56 x 1 MB 77 MB 400W BGA5903 4 UPI 12 x DDR4-2933 April 2, 2019
9242 48 (96) 2.30 GHz 3.80 GHz 48 x 1 MB 71.5 MB 350W
9222[15] 32 (64) 2.30 GHz 3.70 GHz 32 x 1 MB 71.5 MB 250W
9221[15] 32 (64) 2.10 GHz 3.70 GHz 32 x 1 MB 71.5 MB 250W

Cascade Lake-SP (Scalable Performance)

[edit]

Xeon Platinum 8200 series

[edit]
Model number sSpec number Cores (threads) Frequency Turbo Boost L2 Cache L3 Cache TDP Socket I/O Bus Memory Optane memory support Release date Part number(s) Release Price

8280M SRF9Q (B1) 28 (56) 2.70 GHz 4.00 GHz 28 x 1 MB 39 MB 205W LGA3647 3 UPI 6 x DDR4-2933 Yes April 2, 2019 CD8069504228101 $13,012
8280L SRF9R (B1) CD8069504228201 $17,906
8280 SRF9P (B1) CD8069504228001 $10,009
8276M SRF98 (B1) 2.20 GHz 4.00 GHz 39 MB 165W CD8069504195401 $11,722
8276L SRF97 (B1) CD8069504195301 $16,616
8276 SRF99 (B1) CD8069504195501 $8,719
8270 SRF96 (B1) 26 (52) 2.70 GHz 4.00 GHz 26 x 1 MB 36 MB 205W CD8069504195201 $7,405
8268 SRF95 (B1) 24 (48) 2.90 GHz 3.90 GHz 24 x 1 MB CD8069504195101 $6,302
8260Y SRF9F (B1) 2.40 GHz 3.90 GHz 165W CD8069504200902 $5,320
8260M SRF9J (B1) CD8069504201201 $7,705
8260L SRF9G (B1) CD8069504201001 $12,599
8260 SRF9H (B1) CD8069504201101 $4,702
8256 SRF94 (B1) 4 (8) 3.80 GHz 3.90 GHz 4 x 1 MB 17 MB 105W CD8069504194701

BX806958256

$7,007
8253 SRF93 (B1) 16 (32) 2.20 GHz 3.00 GHz 16 x 1 MB 22 MB 125W CD8069504194601 $3,115

Xeon Gold 6200 series

[edit]

Bolded denotes new SKUs released February 24, 2020.[16]

Model number sSpec number Cores (threads) Frequency Turbo Boost L2 Cache L3 Cache TDP Socket I/O Bus Memory Optane memory support Release date Part number(s) Release price

6262V SRFQ4 (B1) 24 (48) 1.90 GHz 3.60 GHz 24 x 1 MB 33 MB 135W LGA3647 3 UPI 6 x DDR4-2400 Yes Q2, 2019 CD8069504285004 $2,900
6262 2 UPI
6258R SRGZF (B1) 28 (56) 2.70 GHz 4.00 GHz 39 MB 205W 2 UPI 6 x DDR4-2933 February 24, 2020 CD8069504449301 $3,950
6256 SRGTQ (B1) 12 (24) 3.60 GHz 4.50 GHz 33 MB 205W 3 UPI February 24, 2020 CD8069504425301 $3,900
6254 SRF92 (B1) 18 (36) 3.10 GHz 4.00 GHz 18 x 1 MB 25 MB 200W April 2, 2019 CD8069504194501 $3,803
6252N SRFPQ (B1) 24 (48) 2.30 GHz 3.60 GHz 24 x 1 MB 36 MB 150W Q2, 2019 CD8069504283503 $3,984
6252 SRF91 (B1) 2.10 GHz 3.70 GHz 24 x 1 MB April 2, 2019 CD8069504194401

BX806956252

$3,655
6250L SRH5D (B1) 8 (16) 3.90 GHz 4.50 GHz 36 MB 185W February 24, 2020 CD8069504497400 $6,404
6250 SRGTR (B1) 8 (16) February 24, 2020 CD8069504425402 $3,400
6248R SRGZG (B1) 24 (48) 3.00 GHz 4.00 GHz 36 MB 205W 2 UPI February 24, 2020 CD8069504449401 $2,700
6248 SRF90 (B1) 20 (40) 2.50 GHz 3.90 GHz 20 x 1 MB 28 MB 150W 3 UPI April 2, 2019 CD8069504194301

BX806956248

$3,072
6246R SRGZL (B1) 16 (32) 3.40 GHz 4.10 GHz 36 MB 205W 2 UPI February 24, 2020 CD8069504449801 $3,286
6246 SRFPJ (B1) 12 (24) 3.30 GHz 4.20 GHz 12 x 1 MB 25 MB 165W 3 UPI Q2, 2019 CD8069504282905 $3,286
6244 SRF8Z (B1) 8 (16) 3.60 GHz 4.40 GHz 8 x 1 MB 25 MB 150W April 2, 2019 CD8069504194202 $2,925
6242R SRGZJ (B1) 20 (40) 3.10 GHz 4.10 GHz 36 MB 205W 2 UPI February 24, 2020 CD8069504449601 $2,529
6242 SRF8Y (B1) 16 (32) 2.80 GHz 3.90 GHz 16 x 1 MB 22 MB 150W 3 UPI April 2, 2019 CD8069504194101

BX806956242

$2,529
6240Y SRF9D (B1) 18 (36) 2.60 GHz 3.90 GHz 18 x 1 MB 25 MB 150W CD8069504200501 $2,726
6240R SRGZ8 (B1) 24 (48) 2.40 GHz 4.00 GHz 36 MB 165W 2 UPI February 24, 2020 CD8069504448600

BX806956240R

$2,200
6240M SRFPZ (B1) 18 (36) 2.60 GHz 3.90 GHz 18 x 1 MB 25 MB 150W 3 UPI Q2, 2019 CD8069504284403 $5,448
6240L SRFQ0 (B1) CD8069504284503 $10,342
6240 SRF8W (B1) April 2, 2019 CD8069504194001

BX806956240

$2,445
6238R SRGZ9 (B1) 28 (56) 2.20 GHz 4.00 GHz 39 MB 165W 2 UPI February 24, 2020 CD8069504448701

BX806956238R

$2,612
6238T SRF9C (B1) 22 (44) 1.90 GHz 3.70 GHz 22 x 1 MB 30 MB 125W 3 UPI April 2, 2019 CD8069504200401 $2,742
6238M SRFQ1 (B1) 2.10 GHz 3.70 GHz 140W Q2, 2019 CD8069504284604 $5,615
6238L SRFQ2 (B1) CD8069504284704 $10,510
6238 SRFPL (B1) CD8069504283104

BX806956238

$2,612
6234 SRFPN (B1) 8 (16) 3.30 GHz 4.00 GHz 8 x 1 MB 25 MB 130W CD8069504283304

BX806956234

$2,214
6230T SRFPS (B1) 20 (40) 2.10 GHz 3.90 GHz 20 x 1 MB 28 MB 125W CD8069504283704 $1,988
6230R SRGZA (B1) 26 (52) 2.10 GHz 4.00 GHz 36 MB 150W 2 UPI February 24, 2020 CD8069504448800

BX806956230R

$1,894
6230N SRFPR (B1) 20 (40) 2.30 GHz 3.90 GHz 20 x 1 MB 28 MB 125W 3 UPI Q2, 2019 CD8069504202700 $2,046
6230 SRF8W (B1) 2.10 GHz April 2, 2019 CD8069504193701

BX806956230

$1,894
6226R SRGZC (B1) 16 (32) 2.90 GHz 3.90 GHz 22 MB 150W 2 UPI February 24, 2020 CD8069504449000

BX806956226R

$1,300
6226 SRFPP (B1) 12 (24) 2.70 GHz 3.70 GHz 12 x 1 MB 19 MB 125W 3 UPI Q2, 2019 CD8069504283404 $1,776
6222V SRFQ5 (B1) 20 (40) 1.80 GHz 3.60 GHz 20 x 1 MB 28 MB 115W 6 x DDR4-2400 CD8069504285204 $1,600
6222 2 UPI Q2, 2019
6210U SRF9B 20 (40) 2.90 GHz 3.90 GHz 22 MB 150W N/A 6 x DDR4-2933 February 24, 2020 CD8069504198101 $989
6208U SRGZD (B1) 16 (32) 2.90 GHz 3.90 GHz 22 MB 150W N/A February 24, 2020 CD8069504449101 $989

Xeon Gold 5200 Series

[edit]

Bolded denotes new SKUs released February 24, 2020.[16]

Model number sSpec number Cores (threads) Frequency Turbo Boost L2 Cache L3 Cache TDP Socket I/O Bus Memory Optane Memory supported Release date Part number(s) Release price

5222 SRF8V (B1) 4 (8) 3.80 GHz 3.90 GHz 4 x 1 MB 17 MB 105W LGA3647 2 UPI 6 x DDR4-2933 Yes April 2, 2019 CD8069504193501 $1,221
5220T SRFPK (B1) 18 (36) 1.90 GHz 3.90 GHz 18 x 1 MB 25 MB 105W 6 x DDR4-2666 Q2, 2019 CD8069504283006 $1,727
5220R SRGZP (B1) 24 (48) 2.20 GHz 4.00 GHz 36 MB 150W February 24, 2020 CD8069504451301

BX806955220R

$1,555
5220S SRFPT (B1) 18 (36) 2.70 GHz 3.90 GHz 18 x 1 MB 25 MB 125W Q2, 2019 CD8069504283804 $2,000
5220 SRFBJ (L1) 2.20 GHz 125W April 2, 2019 CD8069504214601

BX806955220

$1,555
5218T SRFPM (B1) 16 (32) 2.10 GHz 3.80 GHz 16 x 1 MB 22 MB 105W Q2, 2019 CD8069504283204 $1,349
5218R SRGZ7 (B1) 20 (40) 2.10 GHz 4.00 GHz 28 MB 125W February 24, 2020 CD8069504446300

BX806955218R

$1,273
5218N SRFD9 (L1) 16 (32) 2.30 GHz 3.70 GHz 16 x 1 MB 22 MB 105W April 2, 2019 CD8069504289900 $1,375
5218B SRFDJ (L1) 3.90 GHz 125W CD8069504295701 $1,273
5218 SRF8T (B1) 3.90 GHz 125W CD8069504193301

BX806955218

$1,273
5217 SRFBF (L1) 8 (16) 3.00 GHz 3.70 GHz 8 x 1 MB 11 MB 115W CD8069504214302 $1,522
5215M SRFBD (L1) 10 (20) 2.50 GHz 3.40 GHz 10 x 1 MB 14 MB 85W CD8069504214102 $4,224
5215L SRFBE (L1) 14 MB CD8069504214202 $9,119
5215 SRFBC (L1) 13.75 MB CD8069504214002 $1,221

Xeon Silver series

[edit]

Bolded denotes new SKUs released February 24, 2020.[16]

Model number sSpec number Cores (threads) Frequency Turbo Boost L2 Cache L3 Cache TDP Socket I/O Bus Memory Optane Memory supported Release date Part number(s) Release price

4216 SRFBB (L1) 16 (32) 2.10 GHz 3.20 GHz 16 x 1 MB 22 MB 100W LGA3647 2 UPI 6 x DDR4-2400 No April 2, 2019 CD8069504213901

BX806954216

$1,002
4215R SRGZE (B1) 8 (16) 3.20 GHz 4.00 GHz 11 MB 130W Yes February 24, 2020 CD8069504449200 $794
4215 SRFBA (L1) 2.50 GHz 3.50 GHz 8 x 1 MB 85W April 2, 2019 CD8069504212701 $794
4214Y SRFDG (L1) 12 (24) 2.20 GHz 3.20 GHz 12 x 1 MB 17 MB 85W No CD8069504294401 $768
4214R SRG1W (L1) 2.40 GHz 3.50 GHz 100W February 24, 2020 CD8069504343701

BX806954214R

$694
4214 SRFB9 (L1) 2.20 GHz 3.20 GHz 12 x 1 MB 85W April 2, 2019 CD8069504212601

BX806954214

$694
4210T SRGYH (R1) 10 (20) 2.30 GHz 3.20 GHz 14 MB 95W February 24, 2020 CD8069504444900 $555
4210R SRG24 (R1) 2.40 GHz 100W CD8069504344500

BX806954210R

$501
4210 SRFBL (R1) 2.20 GHz 10 x 1 MB 85W April 2, 2019 CD8069503956302

BX806954210

$501
4209T SRFBQ (R1) 8 (16) 2.20 GHz 3.20 GHz 8 x 1 MB 11 MB 70W CD8069503956900 $501
4208 SRFBM (R1) 8 (16) 2.10 GHz 3.20 GHz 8 x 1 MB 11 MB 85W CD8069503956401

BX806954208

$417

Xeon Bronze series

[edit]

Bolded denotes new SKUs released February 24, 2020.[16]

Model number sSpec number Cores (threads) Frequency Turbo Boost L2 Cache L3 Cache TDP Socket I/O Bus Memory Optane Memory supported Release date Part number(s) Release price
3206R SRG25 (R1) 8 (8) 1.90 GHz N/A 11 MB 85W LGA3647 2 UPI 6 x DDR4-2133 No February 24, 2020 CD8069504344600

BX806953206R

$306
3204 SRFBP (R1) 6 (6) 6 x 1 MB 8.25 MB April 2, 2019 CD8069503956700

BX806953204

$213

Cascade Lake-W (Workstation)

[edit]

Xeon W-3200 series

[edit]
Model number sSpec number Cores (threads) Frequency Turbo Boost 2.0 Turbo Boost Max 3.0 L3 cache TDP Socket Memory Release date Part number(s) Release price

3275M SRFFK (B1) 28 (56) 2.50 GHz 4.40 GHz 4.60 GHz 38.5 MB 205W LGA3647 6 x DDR4-2933 Q2 2019 CD8069504248702 $7,453
3275 SRFFF (B1) CD8069504153101 $4,449
3265M SRFFJ (B1) 24 (48) 2.70 GHz 4.40 GHz 4.60 GHz 33 MB CD8069504248601 $6,353
3265 SRFFE (B1) CD8069504153002 $3,349
3245M SRFFH (B1) 16 (32) 3.20 GHz 4.40 GHz 4.60 GHz 22 MB CD8069504248501 $5,002
3245 SRFFD (B1) CD8069504152900 $1,999
3235 SRFFC (B1) 12 (24) 3.30 GHz 4.40 GHz 4.50 GHz 19.25 MB 180W CD8069504152802 $1,398
3225 SRFFB (B1) 8 (16) 3.70 GHz 4.30 GHz 4.40 GHz 16.5 MB 160W 6 x DDR4-2666 CD8069504152705 $1,199
3223 SRFFG (B1) 3.50 GHz 4.00 GHz 4.20 GHz CD8069504248402 $749

Xeon W-2200 series

[edit]
Model number sSpec number Cores

(threads)

Frequency Turbo Boost 2.0 Turbo Boost Max 3.0 L2 cache L3 cache I/O Bus Memory TDP Socket Release

date

Part number(s) Price

(USD)


Xeon W-2295 SRGSL (L1) 18 (36) 3.0 GHz 4.6 GHz 4.8 GHz 1 MB per core 24.75 MB N/A 4 x DDR4-2933 165 W LGA2066 Q4'19 CD8069504393000 $1333
Xeon W-2275 SRGSP (L1) 14 (28) 3.3 GHz 19.25 MB CD8069504393300 $1112
Xeon W-2265 SRGSQ (L1) 12 (24) 3.5 GHz CD8069504393400 $944
Xeon W-2255 SRGV8 (L1) 10 (20) 3.7 GHz 4.5 GHz 4.7 GHz CD8069504393600 $778
Xeon W-2245 SRH02 (L1) 8 (16) 3.9 GHz 16.5 MB 155 W CD8069504393801 $667
Xeon W-2235 SRGVA (L1) 6 (12) 3.8 GHz 4.6 GHz N/A 8.25 MB 130 W CD8069504439102

BX80695W2235

$555
Xeon W-2225 SRH03 (L1) 4 (8) 4.1 GHz 105 W CD8069504394102 $444
Xeon W-2223 SRGSX (L1) 3.6 GHz 3.9 GHz 4 x DDR4-2666 120 W CD8069504394701

BX80695W2223

$297

See also

[edit]

References

[edit]
Revisions and contributorsEdit on WikipediaRead on Wikipedia

Cascade Lake

View on Grokipedia
from Grokipedia
Cascade Lake is the codename for Intel's second-generation Scalable processors, a family of server, , and high-end desktop microprocessors launched starting April 2, , that succeed the Skylake-SP and are fabricated on an enhanced node. These processors emphasize scalability for and AI workloads, supporting configurations from 2 to 56 cores per socket with up to 112 threads, higher clock speeds reaching up to 4.4 GHz in turbo boost, and multi-socket systems up to 8-way via (UPI). Key enhancements include six channels of DDR4-2933 support per socket for up to 1.5 TB of capacity, integration with Intel Optane DC persistent for up to 4.5 TB per socket, and AVX-512 Vector Neural Network Instructions (VNNI) to accelerate inference by up to 30x compared to previous generations. A defining feature of Cascade Lake is its built-in hardware mitigations against side-channel vulnerabilities such as Meltdown and Spectre Variant 2, providing enhanced security without relying solely on software patches and minimizing performance overhead. The lineup includes variants like Cascade Lake-SP for scalable servers, Cascade Lake-AP for advanced performance with up to 56 cores in a , Cascade Lake-W for workstations, and Cascade Lake-X for enthusiast desktops, all compatible with and sockets respectively. Overall, Cascade Lake processors deliver up to 1.7x better performance in technical computing and AI tasks over Skylake predecessors, while supporting PCIe 3.0 with up to 48 lanes per socket for expanded I/O connectivity.

Overview

Introduction

Cascade Lake is Intel's 14 nm microarchitecture refresh of the Skylake design, targeted at server and data center applications within the Xeon Scalable processor family. Launched in April 2019, it serves as the direct successor to the first-generation Skylake-SP processors, introducing optimizations for enterprise workloads including built-in AI acceleration via Intel Deep Learning Boost (DL Boost), which enhances vector neural network instructions (VNNI) for deep learning inference. In its SP (Scalable Performance) variants, Cascade Lake supports up to 28 cores per socket, with each processor featuring six integrated DDR4 memory controllers operating at speeds up to 2933 MHz, enabling system configurations with up to 3 TB of DDR4 memory in dual-socket setups using high-capacity LRDIMMs. The architecture maintains core similarities to Skylake, such as the engine and support, while adding refinements for higher performance density. At a high level, the Cascade Lake consists of multiple core tiles—each housing one core (supporting ) with a private L2 cache—interconnected with distributed L3 cache slices via a 2D mesh network for low-latency . Integrated I/O includes 48 lanes of PCIe 3.0 and support for Optane DC , positioning it as a foundational platform for scalable enterprise servers handling compute-intensive tasks like , databases, and emerging AI applications.

Release Timeline

Intel first announced the Cascade Lake microarchitecture at its Data-Centric Innovation Summit on August 8, 2018, where the company outlined its roadmap for data-centric computing, including enhancements for AI and security in the upcoming server processors. The release timeline was influenced by the need to incorporate hardware-based mitigations for Spectre and Meltdown vulnerabilities, which delayed shipments from an initial target of late 2018 to early 2019. Cascade Lake-SP processors, the scalable performance variant for data centers, began shipping on April 2, 2019, marking the official launch of the family. The Cascade Lake-AP variant, featuring advanced performance with multi-chip modules for high-end computing, followed with initial system shipments in the first half of 2019 and broader availability ramping in the second half. Cascade Lake-W processors for workstations were introduced in 2019 with the W-3200 series. The enthusiast-oriented Cascade Lake-X series was announced in October 2019 and began shipping in November 2019, completing the major variant rollouts. Production of Cascade Lake processors phased out starting in 2021 with the introduction of Ice Lake-SP in April 2021, and continued into 2022 as preparations advanced, with final orders for many models ending by 2023-2024.

Development

Background and Predecessors

Cascade Lake emerged as part of Intel's transition from the traditional Tick-Tock development model, which alternated between process shrinks ("Tick") and architectural overhauls ("Tock"), to a more flexible Process-Architecture-Optimization (PAO) approach amid delays in advancing beyond the 14 nm process node. Under this evolved strategy, Cascade Lake served as an optimization phase for the Skylake server microarchitecture, refining its 14 nm fabrication to improve power efficiency without a full architectural redesign. This positioned Cascade Lake as a direct evolutionary step from Skylake-SP, Intel's first-generation Scalable processors launched in 2017, which themselves built upon the Broadwell-EP architecture introduced in 2014 for prior E5 v4 series. Broadwell had marked Intel's initial adoption of the 14 nm process for high-end servers, establishing a foundation for scalable multi-socket designs, while Skylake introduced the mesh interconnect and expanded core counts to meet growing demands. The development of Cascade Lake was significantly influenced by intensifying in the server market, particularly following AMD's launch of the processor family in mid-2017, which challenged Intel's dominance with higher core counts and competitive pricing in multi-socket configurations. AMD's "Naples" (EPYC 7001 series) offered up to 32 cores per socket and superior , eroding Intel's in hyperscale and enterprise environments and prompting Intel to accelerate optimizations for the Xeon Scalable lineup. In response, Cascade Lake emphasized enhancements in multi-threaded performance and scalability for multi-socket servers, aiming to reclaim leadership in workloads like , databases, and . Internally, Intel's for Cascade Lake began around 2016, shortly after the initial planning for Skylake-SP, with a focus on addressing power efficiency limitations inherent in the 14 nm process through process refinements that enabled higher clock speeds and better . These efforts also incorporated early considerations for vulnerabilities, culminating in hardware-based mitigations integrated during the final design phases to close gaps exposed by emerging threats in server environments. By 2018, prototypes and key tweaks were showcased at industry events like Hot Chips, reflecting a streamlined timeline that leveraged Skylake's established framework to deliver timely updates.

Design Objectives

Intel's Cascade Lake microarchitecture was designed with key objectives centered on advancing server processor capabilities for data center environments, particularly by accelerating artificial intelligence and machine learning workloads through the introduction of Vector Neural Network Instructions (VNNI) as part of Intel Deep Learning Boost (DL Boost). VNNI enables fused multiply-add operations for low-precision integer formats (INT8 and INT16), targeting the intensive matrix multiplications common in deep neural network inference and reducing the instruction count needed for these computations compared to prior generations. Additionally, a primary goal was to incorporate hardware-based mitigations for emerging microarchitectural vulnerabilities, including Microarchitectural Data Sampling (MDS), where later production steppings set the MDS_NO bit in the IA32_ARCH_CAPABILITIES MSR to indicate immunity without software overhead. To balance innovation with practicality, Cascade Lake maintained the 14 nm process node from its Skylake predecessor, opting for an optimized variant to avoid the costs and risks of a full node transition while enabling feature additions like VNNI and security enhancements. This approach prioritized socket compatibility with the interface and the existing Purley platform, allowing drop-in upgrades without requiring motherboard replacements or ecosystem disruptions. The trade-off involved forgoing aggressive density scaling but achieving higher per-core frequencies (up to 100-300 MHz boosts) and refined power delivery to support denser core configurations in multi-socket systems. Performance targets for Cascade Lake emphasized substantial uplifts in AI-specific tasks, aiming for up to 11x improvement in inference throughput over Skylake in benchmarks like Caffe ResNet-50, driven by VNNI's ability to process 256 INT8 elements per cycle per core. These gains were realized through software optimizations in frameworks like Intel's oneAPI Deep Neural Network Library, focusing on rather than to address the growing demand for real-time AI deployment in data centers. From a sustainability perspective, Cascade Lake sought to enhance power efficiency in operations, leveraging the refined for better in targeted workloads, with reports indicating up to 9% overall improvement in server configurations compared to Skylake equivalents. In AI inference scenarios, the efficiency gains were more pronounced due to reduced demands from VNNI.

Microarchitecture

Core Design

The Cascade Lake microarchitecture retains the fundamental core structure of its Skylake predecessor, with targeted refinements for data center efficiency and scalability. Each core employs an pipeline featuring a 6-wide decode that can process up to six x86 instructions per cycle, primarily through a combination of simple and complex decoders to handle variable-length instructions effectively. Following decode, a 4-wide issue allocates up to four micro-operations per cycle to the reservation station and execution ports, supporting a unified scheduler with 97 entries for and floating-point operations. The reorder buffer holds 224 entries, enabling a substantial out-of-order window to maximize while maintaining retirement at up to four instructions per cycle. To address power and thermal constraints in multi-core environments, Cascade Lake incorporates via Technology 2.0, allowing eligible processors to reach turbo frequencies of up to 4.5 GHz in single-core workloads, depending on the specific model and cooling configuration. This boost provides a uplift for latency-sensitive tasks without compromising all-core stability, where frequencies typically range from 2.0 to 3.0 GHz under full utilization. A prominent enhancement in the core's vector processing capabilities is the integration of instructions with Vector Neural Network Instructions (VNNI) as part of Deep Learning Boost. VNNI fuses three operations—specifically for INT8 dot products—into a single instruction (e.g., VPDPBUSD), accelerating inference by up to eight times compared to prior generations using AVX2 on compatible workloads. This extension targets INT8 quantized neural networks, while FP16 operations leverage the existing FMA units for half-precision floating-point computations in AI training and scenarios.

Cache and Memory Hierarchy

The cache hierarchy in Cascade Lake processors features a re-architected compared to predecessors, with each core equipped with a 32 KiB L1 instruction cache and a 32 KiB L1 cache, both 8-way set associative with 64-byte line sizes, a private 1 MB L2 cache that serves as the primary cache, eliminating the need for L1 cache snooping in many cases. The shared L3 cache, known as the last-level cache (LLC), is non-inclusive and provides up to 1.375 MB per core, resulting in a total of up to 38.5 MB per socket for 28-core SP variants. This reconfiguration enhances prefetching efficiency by reducing conflicts between private L2 and shared L3 contents, allowing hardware s—such as the DCU streamer and IP —to better anticipate and load without duplicating entries across levels. The on-die interconnect shifts from the ring topology of prior generations to a 2D architecture, which scales more effectively for higher core counts by distributing cores, caches, and I/O tiles across a grid of routers and links. This mesh supports multi-socket configurations of up to eight processors via (UPI) links, with each socket featuring up to three UPI channels operating at 10.4 GT/s for inter-socket communication. Cascade Lake supports DDR4 memory at speeds up to 2933 MT/s across six channels per socket in standard SP configurations, with certain variants extending to 12 channels. This setup delivers aggregate memory bandwidth of approximately 141 GB/s per socket when fully populated. Additionally, integration with Intel Optane DC persistent memory enables hybrid memory modes, combining DRAM with larger-capacity persistent DIMMs for capacities up to 4.5 TB per socket while maintaining byte-addressable access.

Key Features

Security Enhancements

Cascade Lake processors incorporate hardware mitigations for Microarchitectural Data Sampling (MDS) vulnerabilities, which encompass attacks such as ZombieLoad, Fallout, and RIDL that exploit transient execution to leak data across security boundaries. In specific steppings (6 and 7 of family model 06_55H), these processors eliminate vulnerability to key MDS variants like MFBDS, MSBDS, and MLPDS through architectural improvements that prevent data leakage from microarchitectural buffers. Earlier steppings (5) rely on microcode updates combined with software mechanisms, such as the MD_CLEAR feature (enumerated via CPUID leaf 7), which flushes affected structures like load ports and store buffers using instructions like VERW or L1D_FLUSH. Building on prior generations, Cascade Lake integrates first-generation (SGX) for secure computation within isolated enclaves, enabling applications to protect sensitive data and code from higher-privilege software or physical attacks. This support allows up to 256 MB of Enclave Page Cache (EPC) per processor. Enclaves in Cascade Lake benefit from dynamic memory management and remote attestation, ensuring encrypted execution and verification of trusted environments. To address side-channel risks akin to Spectre, Cascade Lake introduces enhanced hardware partitioning that strengthens isolation between hyper-threaded logical cores, reducing the potential for cross-thread data leakage through . These "protective walls" in the microarchitecture limit the sharing of prediction resources and buffers, providing built-in defenses against Variant 2 ( target injection) exploits without solely depending on software barriers like Indirect Branch Restricted Speculation (IBRS). Post-launch, Intel issued microcode updates for Cascade Lake to patch residual MDS-related issues, including ZombieLoad (a Fallout variant exploiting store buffer leaks) and broader Fallout attacks on load ports. These firmware patches, deployed via BIOS or OS loaders, enable full enumeration of MDS immunity (via IA32_ARCH_CAPABILITIES MSR bit 5) and activate selective clearing of transient states, ensuring ongoing protection against evolving side-channel threats.

Performance Optimizations

Cascade Lake processors incorporate the Deep Learning Boost (DL Boost) suite, which includes Vector Neural Network Instructions (VNNI) to accelerate low-precision operations in . VNNI enables up to 2x higher throughput for INT8 compared to prior architectures without dedicated support, by doubling the performance of SIMD integer multiply-accumulate operations essential for workloads. For (HPC) applications, Cascade Lake builds on vector extensions to deliver enhanced vector processing capabilities, supporting wider in scientific simulations and data analytics. Successor platforms, such as Cooper Lake, introduce bfloat16 (BF16) support within , doubling theoretical compute throughput over FP32 for AI training tasks while maintaining through shared exponent ranges with single-precision floats. Power efficiency is optimized through via Technology 2.0, which adjusts core frequencies based on demands and constraints to balance and use. High-end SKUs, like the Platinum 8280, support a maximum TDP of 205W, enabling sustained high frequencies in multi-socket configurations for demanding server environments. In standardized benchmarks, Cascade Lake achieves 10-15% gains in SPEC CPU 2017 integer rate (SPECint_rate2017) over equivalent Skylake configurations in integer-heavy workloads, driven by higher base/turbo clocks, increased from DDR4-2933 support, and architectural refinements.

Processor Variants

Cascade Lake-X

Cascade Lake-X processors represent Intel's high-end desktop (HEDT) offering within the Cascade Lake family, targeted at enthusiast and workstations. These processors utilize the socket and support single-socket configurations only, with no dual-socket capability. They share the core detailed in the broader Cascade Lake design but are optimized for consumer-oriented tasks such as , , and gaming. The lineup includes models from the 10th Generation X-series, scaling up to 18 cores and 36 threads, as exemplified by the flagship i9-10980XE with a base clock of 3.0 GHz, turbo boost up to 4.6 GHz, and a (TDP) of 165 W. Other variants include the 14-core i9-10940XE, 12-core i9-10920X, and 10-core i9-10900X, all featuring unlocked multipliers for support across the series to enable by enthusiasts. Memory configuration is quad-channel DDR4-2933, supporting up to 256 GB total capacity, which provides ample bandwidth for memory-intensive workloads. Announced on October 7, 2019, and available starting November 2019, Cascade Lake-X processors were positioned to compete directly with AMD's Threadripper 3000 series in the HEDT market, emphasizing price reductions of up to 50% compared to prior generations alongside incremental improvements in clock speeds and platform features like up to 48 PCIe 3.0 lanes. This release marked Intel's effort to regain competitiveness in multi-threaded performance for desktop users while maintaining compatibility with existing X299 chipsets.

Cascade Lake-AP

Cascade Lake-AP represents Intel's advanced performance variant within the Cascade Lake family, designed specifically for extreme in multi-socket server environments supporting up to eight sockets. This configuration enables massive parallel processing capabilities, with processors featuring high core densities to handle demanding workloads requiring extensive computational resources. The builds on the base Cascade Lake design by integrating dual dies per package, which enhances scalability while maintaining compatibility with the socket and UPI interconnects for multi-socket coherence. The Platinum 9200 series constitutes the primary lineup for Cascade Lake-AP, offering models such as the Platinum 9248 with 48 cores and 96 threads at a 350 W TDP, and the flagship Platinum 9282 delivering 56 cores and 112 threads with a base frequency of 2.60 GHz, turbo up to 3.80 GHz, and a 400 W TDP. These processors support 12 channels of DDR4-2933 memory per socket, providing up to 1.5 TB capacity and significantly higher bandwidth compared to standard variants, which is crucial for memory-intensive tasks. Released in April 2019, the series was produced in limited volumes tailored for custom deployments in specialized systems. Engineered primarily for (HPC) and applications, Cascade Lake-AP processors excel in scenarios involving large-scale simulations, risk analysis, and where ultra-high core counts and throughput directly impact outcomes. For instance, the 12-channel subsystem facilitates faster access in bandwidth-bound workloads, enabling up to 2x the memory channels of conventional Cascade Lake-SP models for superior handling of terabyte-scale datasets in and scientific modeling. This focus on extreme density and I/O capacity positions Cascade Lake-AP as a specialized solution for enterprise-scale servers optimized for throughput rather than broad-market versatility.

Cascade Lake-SP

Cascade Lake-SP is the primary variant of the Cascade Lake processor family designed for scalable performance in mainstream data center servers, targeting configurations from single-socket to eight-socket systems. It offers a broad lineup across multiple tiers to address diverse enterprise and cloud workloads, including the Xeon Platinum 8200 and 6200 series for high-end performance, Gold 5200 series for balanced compute, Silver 4200 series for cost-effective scalability, and Bronze 3100 series for entry-level density. For instance, the Platinum 8280 model features 28 cores, a base frequency of 2.7 GHz, and a thermal design power (TDP) of 205 W, exemplifying the series' focus on high-core-count processing for demanding applications. A key enabler for multi-socket in Cascade Lake-SP is the Ultra Path Interconnect (UPI), which provides up to three 10.4 GT/s links per processor to support coherent communication in 2-, 4-, or 8-socket configurations. This interconnect facilitates efficient data sharing across sockets, making it suitable for large-scale server deployments. Additionally, the architecture supports expansive memory configurations, with up to 6 TB of capacity per socket when utilizing Optane DC alongside DDR4, enabling handling of memory-intensive tasks such as and databases. Released on April 2, 2019, Cascade Lake-SP quickly became dominant in and enterprise environments, with major providers like and adopting it for compute- and -optimized virtual machines. Its combination of core density, interconnect efficiency, and memory expandability positioned it as a cornerstone for hyperscale data centers and enterprise infrastructure during its lifecycle.

Cascade Lake-W

Cascade Lake-W processors represent Intel's workstation-oriented implementation of the Cascade Lake , tailored for single-socket configurations to handle intensive professional workloads in creative and engineering fields, such as (CAD) and . These processors utilize the socket for the W-2200 series and the socket for the W-3200 series, enabling up to 28 cores per processor to deliver robust multi-threaded performance for specialized applications. The W-3200 series, launched in the second quarter of , features high-core-count models optimized for demanding environments, exemplified by the W-3275M with 28 cores, 56 threads, a base frequency of 2.50 GHz, and a (TDP) of 205 W. Complementing this, the W-2200 series, released in the fourth quarter of , targets similar professional use cases with mid-range core configurations, such as the W-2295 offering 18 cores, 36 threads, a base frequency of 3.00 GHz, and a 165 W TDP. The W-3200 series supports up to six channels of DDR4-2933 with a maximum capacity of 2 TB, while the W-2200 series supports four channels up to 1 TB. A key differentiator for Cascade Lake-W is its robust support for error-correcting code (ECC) DDR4 memory, which enhances for mission-critical professional tasks, alongside compatibility with up to 64 PCIe 3.0 lanes for accelerated graphics and storage. These processors are also validated through (ISV) certifications, ensuring optimized performance with applications from (e.g., Premiere Pro) and (e.g., ), as integrated in certified workstation platforms from vendors like and series. Introduced as the successor to the Skylake-W series, Cascade Lake-W builds on prior designs by incorporating Cascade Lake's architectural refinements, including improved and features, while maintaining a focus on single-socket reliability for professional users. It shares some high-end desktop (HEDT) overlaps with the Cascade Lake-X lineup in target market and socket compatibility for the W-2200 series.

Applications and Impact

Server and Data Center Use

Cascade Lake processors, part of Intel's second-generation Scalable family, were widely adopted in enterprise server environments for their enhanced core counts, memory support, and optimizations for workloads. In , (AWS) powered its EC2 C5 instances with custom Cascade Lake-based Intel Platinum processors, delivering sustained all-core turbo frequencies up to 3.6 GHz for compute-intensive applications. Similarly, utilized Cascade Lake in its C2 compute-optimized machine series, offering high single-threaded performance for workloads requiring low latency and high throughput, and in N2 general-purpose instances for configurations up to 80 vCPUs. On-premises deployments frequently incorporated Cascade Lake into servers, such as the R740 and R640 models, which supported dual-socket configurations for scalable rack environments. In server and data center settings, Cascade Lake excelled in and database workloads, providing improved resource utilization over prior generations. For virtualization platforms like , Cascade Lake enabled higher densities, with studies showing up to 62% greater user density per compute node in virtual desktop infrastructure compared to Skylake processors, benefiting mixed enterprise applications. In database scenarios, such as , Cascade Lake delivered scalable performance in virtualized setups; for instance, on a dual-socket system with 48 cores, it supported up to six 8-vCPU VMs running benchmarks, achieving steady throughput scaling without significant overhead. Overall, deployments reported 20-30% better performance efficiency in mixed loads, including and databases, due to architectural enhancements like larger caches and higher memory speeds. A notable of Cascade Lake's application in high-scale data processing occurred at , where it powered nodes in the DEEP-EST project for High Energy Physics simulations. Specifically, 16 dual-socket nodes equipped with Platinum 8260M Cascade Lake processors and 384 GB RAM per node handled reconstruction tasks for the CMS experiment's calorimeters, achieving approximately 50% speed-up in High-Level Trigger configurations for the High Luminosity LHC data pipeline during 2019-2022 operations. By 2021, many data centers began migrating from Cascade Lake to 's third-generation Scalable processors (Ice Lake) for further performance and efficiency gains, though legacy support persisted through extended servicing updates. Intel committed to fulfilling Cascade Lake orders until October 2026, ensuring ongoing compatibility in enterprise environments.

Workstation and HEDT Adoption

Cascade Lake processors, particularly the Xeon W-series variants, were widely integrated into professional workstations such as the HP Z8 G4 and Lenovo ThinkStation P520, P720, and P920 models, enabling dual-socket configurations for demanding single-user workloads. These systems supported Cascade Lake's enhanced core counts and memory capacities, making them suitable for resource-intensive tasks like 3D modeling in applications such as Autodesk Maya and video editing in Adobe Premiere Pro. Engineers and content creators adopted these workstations for their balance of multi-threaded performance and reliability in creative pipelines. In rendering benchmarks, Cascade Lake processors delivered notable improvements over prior generations, appealing to the professional user base. For instance, the W-3200 series (Cascade Lake-W) achieved up to 43% faster single-threaded and 67% faster multi-threaded performance in rendering tests compared to Skylake-W equivalents, accelerating scene finalization for 3D artists. Similarly, in workloads, the Core i9-10980XE (Cascade Lake-X) edged out its Skylake-X predecessor with modest gains of around 5-10%, though it prioritized stability for prolonged creative sessions over raw speed. These enhancements, combined with features like vector extensions, positioned Cascade Lake as a reliable choice for content creators handling complex visualizations and edits. Cascade Lake's HEDT offerings, including the X-series, competed directly with AMD's Threadripper lineup, prompting Intel to reduce pricing by up to 50% in late 2019 to maintain competitiveness in the high-end desktop market. By 2020, retained a dominant position in HEDT sales through workstation integrators like Puget Systems, where AMD's gains were notable but Intel's ecosystem advantages—such as broader ISV certifications—sustained its lead. The platform's longevity extended into 2025, with cost-sensitive upgrades still viable in legacy workstations due to ongoing support until mid-2026 and availability of refurbished components.

References

  1. https://www.hpcwire.com/2019/04/02/intel-launches-second-gen-scalable-xeons-with-up-to-56-cores/
  2. https://www.intel.com/content/dam/www/public/us/en/documents/product-briefs/2nd-gen-xeon-scalable-processors-brief.pdf
  3. https://en.wikichip.org/wiki/intel/microarchitectures/cascade_lake
  4. https://community.intel.com/t5/Processors/Meltdown-and-Spectre-Variant-2-silicon-fix/m-p/558366
  5. https://www.intel.com/content/www/us/en/products/platforms/details/cascade-lake.html
  6. https://www.microway.com/knowledge-center-articles/detailed-specifications-of-the-cascade-lake-sp-intel-xeon-processor-scalable-family-cpus/
  7. https://www.intel.com/content/www/us/en/products/sku/198652/intel-xeon-gold-6250-processor-35-75m-cache-3-90-ghz/specifications.html
  8. https://www.supermicro.com/en/products/motherboard/x11dpu
  9. https://www.nas.nasa.gov/hecc/support/kb/cascade-lake-processors_579.html
  10. https://www.cisco.com/c/en/us/products/collateral/servers-unified-computing/whitepaper-c11-742358.html
  11. https://www.servethehome.com/intel-data-centric-innovation-summit-2018-cascade-lake-optane-and-more/
  12. https://www.nextplatform.com/2018/08/22/the-skinny-on-future-cascade-lake-xeons/
  13. https://www.nextplatform.com/2019/04/02/intel-pushes-xeon-sps-to-the-next-level-with-cascade-lake/
  14. https://www.intc.com/news-events/press-releases/detail/81/intel-announces-broadest-product-portfolio-for-moving
  15. https://www.servethehome.com/intel-xeon-w-3200-sku-list-and-value-analysis/
  16. https://www.techradar.com/news/intel-to-launch-cascade-lake-x-processors-in-november-cheaper-than-ever-before
  17. https://www.tomshardware.com/news/intels-xeon-cascade-lake-cpus-have-reached-the-end-of-the-road
  18. https://www.servethehome.com/intel-discontinues-68-intel-xeon-platinum-gold-silver-and-bronze-processors/
  19. https://www.nextplatform.com/2017/07/11/x86-battle-lines-drawn-intels-skylake-launch/
  20. https://www.pogolinux.com/blog/intel-xeon-servers-cascade-lake-refresh-overview/
  21. https://www.networkworld.com/article/966604/intel-responds-to-the-epyc-server-threat-from-amd.html
  22. https://old.hotchips.org/hc30/2conf/2.15_Intel_Cascade_Lake_HC30.Intel.Akhilesh.CLXCPU.Final.pdf
  23. https://www.intel.com/content/www/us/en/developer/articles/technical/software-security-guidance/resources/processors-affected-mds.html
  24. https://dl.dell.com/manuals/common/refreshed_poweredge_server_performance_and_energy_efficiency.pdf
  25. https://cdrdv2-public.intel.com/821612/248966-Optimization-Reference-Manual-V1-050.pdf
  26. https://old.hotchips.org/wp-content/uploads/hc_archives/hc29/HC29.22-Tuesday-Pub/HC29.22.90-Server-Pub/HC29.22.930-Xeon-Skylake-sp-Kumar-Intel.pdf
  27. https://www.intel.com/content/dam/www/public/us/en/documents/sales-briefs/2nd-gen-xeon-scalable-processors-sales-brief.pdf
  28. https://www.intel.com/content/www/us/en/products/docs/processors/xeon/2nd-gen-xeon-scalable-processors-brief.html
  29. https://www.dell.com/support/kbdoc/en-us/000176921/bios-characterization-for-hpc-with-intel-cascade-lake-processors
  30. https://dspace.mit.edu/bitstream/handle/1721.1/143928/Dai-milesdai-meng-eecs-2021-thesis.pdf
  31. https://arstechnica.com/gadgets/2018/11/intel-announces-cascade-lake-xeons-48-cores-and-12-channel-memory-per-socket/
  32. https://www.intel.com/content/www/us/en/support/articles/000059657/processors/intel-xeon-processors.html
  33. https://www.intel.com/content/www/us/en/support/articles/000059614/processors/intel-xeon-processors.html
  34. https://download.intel.com/newsroom/2021/archive/2018-03-15-editorials-advancing-security-silicon-level.pdf
  35. https://www.intel.com/content/dam/develop/public/us/en/include/software-security/affected-processors-2018tab-april2025.html
  36. https://cdrdv2-public.intel.com/821613/355308-Optimization-Reference-Manual-050-Changes-Doc.pdf
  37. https://download.intel.com/newsroom/2021/archive/2019-08-06-news-next-generation-intel-xeon-scalable-processors-deliver-breakthrough-platform-performance-56-processor-cores.pdf
  38. https://www.intel.com/content/www/us/en/products/sku/192478/intel-xeon-platinum-8280-processor-38-5m-cache-2-70-ghz/specifications.html
  39. https://www.intel.com/content/www/us/en/products/sku/198017/intel-core-i910980xe-extreme-edition-processor-24-75m-cache-3-00-ghz/specifications.html
  40. https://www.techpowerup.com/cpu-specs/core-i9-10980xe.c2238
  41. https://www.techpowerup.com/259724/intel-10th-gen-core-x-cascade-lake-hedt-processors-launch-on-october-7
  42. https://arstechnica.com/gadgets/2019/10/intels-cascade-x-hedt-squares-off-with-amds-threadripper/
  43. https://www.servethehome.com/intel-xeon-platinum-9200-formerly-cascade-lake-ap-launched/
  44. https://www.intel.com/content/www/us/en/products/sku/194146/intel-xeon-platinum-9282-processor-77m-cache-2-60-ghz/specifications.html
  45. https://www.intel.com/content/dam/www/public/us/en/documents/technology-briefs/performance-for-hpc-platforms-brief.pdf
  46. https://wccftech.com/intel-cascade-lake-ap-xeon-cpu-vs-amd-eypc-performance-numbers/
  47. https://www.channelfutures.com/cloud/google-ibm-adopt-new-intel-xeon-chips-for-the-cloud
  48. https://www.mitaccomputing.com/en/newsletter/Cascade_Lake-SP_Products
  49. https://www.intel.com/content/www/us/en/ark/products/series/125035/intel-xeon-w-processor.html
  50. https://www.intel.com/content/www/us/en/products/sku/193754/intel-xeon-w3275m-processor-38-5m-cache-2-50-ghz/specifications.html
  51. https://www.intel.com/content/www/us/en/products/sku/198011/intel-xeon-w2295-processor-24-75m-cache-3-00-ghz/specifications.html
  52. https://arstechnica.com/gadgets/2019/10/intels-w-2200-series-xeon-joins-x-series-cascade-lake/
  53. https://i.dell.com/sites/doccontent/shared-content/data-sheets/en/Documents/Precision-7820-Tower-Spec-Sheet.pdf
  54. https://h20195.www2.hp.com/v2/getpdf.aspx/c05527757.pdf
  55. https://aws.amazon.com/ec2/instance-types/c5/
  56. https://www.cloudbolt.io/gcp-cost-optimization/gcp-instance-types/
  57. https://ubuntu.com/certified/201904-26975
  58. https://infohub.delltechnologies.com/p/improved-performance-and-user-densities-with-dell-emc-ready-solutions-for-vdi-powered-by-2nd-generation-intel/
  59. https://www.vmware.com/content/dam/digitalmarketing/vmware/en/pdf/techpaper/performance/vsphere7-sql-server-perf.pdf
  60. https://cloud.google.com/blog/products/compute/choose-the-right-google-compute-engine-machine-type-for-you
  61. https://cds.cern.ch/record/2813816/files/document.pdf
  62. https://www.servethehome.com/intel-xeon-ice-lake-edition-marks-the-start-and-end-of-an-era/
  63. https://blog.itcreations.com/upgrading-the-hp-z8-g4-workstation-to-support-intel-cascade-lake-cpus/
  64. https://www.bargainhardware.co.uk/knowledge-base/hp-z8-g4-cpu-support-list
  65. https://forums.lenovo.com/t5/ThinkStation-Workstations/Upgrade-CPU-in-P520C/m-p/5299611
  66. https://download.lenovo.com/pccbbs/thinkcentre_pdf/p920_p720_power_configurator_v1.7.pdf
  67. https://www.hp.com/us-en/workstations/z8-g4.html
  68. https://www.lenovo.com/us/en/p/workstations/thinkstation-p-series/lenovo-thinkstation-px-intel-tower-workstation/len102s0013
  69. https://www.pugetsystems.com/labs/articles/cinema-4d-cpu-rendering-performance-intel-xeon-w-3200-series-processors-1557/
  70. https://hothardware.com/reviews/intel-core-i9-10980xe-cascade-lake-x-review?page=4
  71. https://techgage.com/article/intel-core-i9-10980xe-review/5/
  72. https://www.tomshardware.com/news/intel-cascade-lake-x-pricing-availability-launch-specifications-10th-generation%2C40526.html
  73. https://www.tomshardware.com/news/ryzen-dominates-puget-systems-sales
  74. https://www.intel.com/content/www/us/en/support/articles/000022396/processors.html
Add your contribution
Related Hubs
User Avatar
No comments yet.