Nehalem (microarchitecture) explained

Nehalem
Cores:	1-6 (1-8 Xeon)
Cores1:	6 (12) (Extreme)
Cores2:	4-8 (4-16) (Xeon) -->
Transistors:	731M to 2300M 45 nm
Transistors1:	774M 45 nm (B0, B1)
Transistors2:	2,300M 45 nm (D0) -->
Clock:	1.06 GHz to 3.33 GHz
L1cache:	64 KB per core (32 KB data + 32 KB instructions)
L2cache:	256 KB per core
L3cache:	2 MB to 24 MB shared
Dmi-Slowest:	2
Qpi-Slowest:	4.80
Qpi-Fastest:	6.40
Arch:	x86-16, IA-32, x86-64
Microarch:	Nehalem
Instructions:	MMX, SSE, SSE2, SSE3, SSSE3, SSE4, SSE4.1, SSE4.2
Extensions:	VT-x, VT-d
Sock1:	LGA 1156
Sock2:	LGA 1366
Sock3:	LGA 1567
Sock4:	μPGA 988
Predecessor:	Core (tock) Penryn (tick)
Successor:	Westmere (tick) Sandy Bridge (tock)
Size-From:	45 nm
Support Status:	Unsupported
Soldby:	Intel
Designfirm:	Intel
Manuf1:	Intel
Model1:	Core i5-7xx
Model2:	Core i7-8xx
Model3:	Core i7-9xx Extreme
Model4:	Xeon 3000/5000/7000
Model5:	Mobile Core i7-7xxQM
Model6:	Mobile Core i7-8xxQM
Core1:	Clarksfield
Core2:	Lynnfield
Core3:	Bloomfield
Core4:	Jasper Forest
Core5:	Gainestown (Nehalem-EP)
Core6:	Beckton (Nehalem-EX)
Core7:	~~Aubrurndale~~
Core8:	~~Havendale~~
Model7:	Mobile Core i7-9xxXM Extreme
Pcode1:	NHM

Nehalem is the codename for Intel's 45 nm microarchitecture released in November 2008. It was used in the first generation of the Intel Core i5 and i7 processors, and succeeds the older Core microarchitecture used on Core 2 processors. The term "Nehalem" comes from the Nehalem River.

Nehalem is built on the 45 nm process, is able to run at higher clock speeds without sacrificing efficiency, and is more energy-efficient than Penryn microprocessors. Hyper-threading is reintroduced, along with a reduction in L2 cache size, as well as an enlarged L3 cache that is shared among all cores. Nehalem is an architecture that differs radically from NetBurst, while retaining some of the latter's minor features.

Nehalem later received a die-shrink to 32 nm with Westmere, and was fully succeeded by "second-generation" Sandy Bridge in January 2011.

Technology

Cache line block on L2/L3 cache was reduced from 128 bytes in NetBurst & Merom/Penryn to 64 bytes per line in this generation (same size as Yonah and Pentium M).
Hyper-threading reintroduced.
Intel Turbo Boost 1.0.^[1]
2–24 MiB L3 cache with Smart Cache in some models.
Instruction Fetch Unit (IFU) containing second-level branch predictor with two level Branch Target Buffer (BTB) and Return Stack Buffer (RSB). Nehalem also supports all predictor types previously used in Intel's processors like Indirect Predictor and Loop Detector.^[2]
sTLB (second level unified translation lookaside buffer) (i.e. both instructions and data) that contains 512 entries for small pages only, and is again 4 way associative.^[3]
3 integer ALU, 2 vector ALU and 2 AGU per core.^[4]
Native (all processor cores on a single die) quad-, hex-, and octa-core processors
Intel QuickPath Interconnect in HEDT, server, and workstation models and Direct Media Interface on other models replacing the legacy front side bus.
64 KB L1 cache per core (32 KB L1 data and 32 KB L1 instruction), and 256 KB L2 cache per core.
Integration of PCI Express and DMI into the processor in mid-range models, replacing the northbridge.
Integrated memory controller supporting two or three memory channels of DDR3 SDRAM or four FB-DIMM2 channels.
Second-generation Intel Virtualization Technology, which introduced Extended Page Table support, virtual processor identifiers (VPIDs), and non-maskable interrupt-window exiting. ^[5]
SSE4.2 and POPCNT instructions.
Macro-op fusion now works in 64-bit mode.
20 to 24 pipeline stages.^[6]

Translation lookaside buffer sizes^[7]
Cache		Page Size
Name	Level	4 KB	2 MB
DTLB	1st	64	32
ITLB	1st	128	7 / logical core
STLB	2nd	512	none

Performance and power improvements

It has been reported that Nehalem has a focus on performance, thus the increased core size.Compared to Penryn, Nehalem has:

10 - 25% better single-threaded performance / 20 - 100% better multithreaded performance at the same power level
30% lower power consumption for the same performance
On average, Nehalem provides a 15 - 20% clock-for-clock increase in performance per core.

Overclocking is possible with Bloomfield processors and the X58 chipset. Lynnfield processors use a PCH removing the need for a northbridge.

Nehalem processors incorporate SSE4.2 SIMD instructions, adding seven new instructions to the SSE 4.1 set in the Core 2 series. The Nehalem architecture reduces atomic operation latency by 50% in an attempt to eliminate overhead on atomic operations such as the LOCK CMPXCHG compare-and-swap instruction.

Variants

Processing Cores (interface)	Process	Die Size	million transistors	CPUID	Model	Stepping	Mobile	Desktop, UP Server	DP Server	MP Server
Eight-Core (Quad-Channel)	45 nm	684 mm²	2.300^[8]	206E6	46	D0				Beckton (80604)
Quad-Core (Triple-Channel)		263 mm²	731	106A4 106A5	26	C0/C1 D0		Bloomfield (80601)	Gainestown (80602)
Quad-Core (Dual-Channel, PCIe)		296 mm²	774	106E4 106E5	30	B0 B1	Clarksfield (80607)	Lynnfield (80605)	Jasper Forest (80612)
Dual-Core (Dual-Channel, PCIe, Graphics Core)			?				Auburndale (80608) (canceled)	Havendale (80606) (canceled)

Lynnfield processors feature 16 PCIe lanes, which can be used in 1x16 or 2x8 configuration.
¹ 6500 series scalable up to 2 sockets, 7500 series scalable up to 4/8 sockets.^[9]

Server, workstation, and desktop processors

See main article: List of Intel Core i5 processors, List of Intel Core i7 processors and List of Intel Xeon processors (Nehalem-based).

Codename	Market	Cores (threads)	Socket	Processor Branding & Model	CPU Clock rate	Turbo	TDP	Interfaces		L3 cache	Release Date	Price for 1k Unit
Codename	Market	Cores (threads)	Socket	Chipset	CPU Clock rate	Turbo	TDP	Memory		L3 cache	Release Date	Price for 1k Unit
Beckton¹	MP Server / DP Server	8 (16)	LGA 1567	Xeon7000	X7560	2.26 GHz	130 W	4× QPI 6.4 GT/s	DDR3-800 / 1066 (Up to 4x with SMB-Ready motherboard)	24 MB	2010-03-30	$3692
					X7550	2.0 GHz	18 MB			$2837
					X6550	2.0 GHz	18 MB			$2461
					L7555	1.86 GHz	95 W	4× QPI 5.86 GT/s		24 MB		$3157
		6 (12)			E7540	2.0 GHz	105 W	4× QPI 6.4 GT/s		18 MB		$1980
					E6540	2.0 GHz		4× QPI 6.4 GT/s		12 MB		$1712
					E7530	1.86 GHz		4× QPI 5.86 GT/s		12 MB		$1391
					L7545	1.86 GHz	95 W			18 MB		$2087
		6 (6)			X7542	2.66 GHz	130 W					$1980
		4 (8)			E7520	1.86 GHz	105 W	4× QPI 4.8 GT/s				$856
		4 (8)			E6510	1.73 GHz	12 MB	4× QPI 4.8 GT/s		$744
Gainestown	DP Server	4 (8)	LGA 1366	Xeon5000	W5590	3.33 GHz	130 W	2× QPI 6.4 GT/s	3× DDR3-1333¹	8 MB	2009-08-09	$1600
					W5580	3.2 GHz	2009-03-29				$1500
					X5570	2.93 GHz					95 W	$1286
					X5560	2.8 GHz						$1072
					X5550	2.66 GHz						$858
					E5540	2.53 GHz		80 W	2× 5.86 GT/s		3× DDR3-1066¹	$744
					E5530	2.4 GHz						$530
					E5520	2.26 GHz						$373
					L5530	2.4 GHz	60 W	2009-08-09				$744
					L5520	2.26 GHz		2009-03-30				$530
					L5518	2.13 GHz		2009-03-30				$
		4 (4)			E5507	2.26 GHz	80 W	2× 4.8 GT/s	3× DDR3-800¹	4 MB	2010-03-16	$266
					E5506	2.13 GHz	2009-03-29					$266
					L5506	2.13 GHz					60 W	$423
					E5504	2.0 GHz					80 W	$224
		2 (4)			L5508	2.0 GHz			38 W	2× 5.86 GT/s	3× DDR3-1066	8 MB	$
		2 (2)			E5503	2.0 GHz	80 W	2× 4.8 GT/s	3× DDR3-800	4 MB	2010-03-16	$224
		2 (2)			E5502	1.86 GHz	2009-03-29	2× 4.8 GT/s	3× DDR3-800	4 MB	$188
JasperForest		4 (8)			EC5549	2.53 GHz	85 W	1× 5.86 GT/s	3× DDR3-1333	8 MB	2010-02-11	$530
					LC5528	2.13 GHz	60 W	1× 4.8 GT/s	3× DDR3-1066			$519
					LC5518	1.73 GHz	48 W					$519
		4 (4)			EC5509	2 GHz	85 W					$265
		2 (4)			EC5539	2.27 GHz	65 W	1× 5.86 GT/s	3× DDR3-1333	4 MB		$387
Bloomfield	UP Server	4 (8)		Xeon3000	W3580	3.33 GHz	130 W	1× QPI 6.4 GT/s	3× DDR3-1333	8 MB	2009-08-09	$999
					W3570	3.2 GHz	2009-03-29	1× QPI 6.4 GT/s	3× DDR3-1333			$999
					W3565	3.2 GHz	1× QPI 4.8 GT/s	3× DDR3-1066	2009-11-01		$562
					W3550	3.06 GHz			2009-08-09
					W3540	2.93 GHz			2009-03-29
					W3530	2.8 GHz			2010-03-16		$294
					W3520	2.66 GHz			2009-03-29		$284
		2 (2)			W3505	2.53 GHz				4 MB	$
		2 (2)			W3503	2.4 GHz				$
JasperForest		4 (4)			EC3539	2.13 GHz	65 W		DMI	8 MB	2010-02-11	$302
		2 (4)			LC3528	1.73 GHz		35 W		3× DDR3-800		$302	4 MB
		1 (1)			LC3518	1.73 GHz		23 W		3× DDR3-800		2 MB	$192
Lynnfield		4 (8)	LGA 1156		X3480	3.06 GHz	95 W	DMI	2× DDR3-1333	8 MB	2010-05-30	$612
					X3470	2.93 GHz	2009-09-08				$589
					X3460	2.8 GHz					$316
					X3450	2.66 GHz					$241
					X3440	2.53 GHz					$215
					L3426	1.86 GHz					45 W	$284
		4 (4)			X3430	2.4 GHz					95 W	$189
Bloomfield	Enthusiast Desktop^[10]	4 (8)	LGA 1366	Core i7 Extreme	975X	3.33 GHz	130 W	1× QPI 6.4 GT/s	3× DDR3-1066		2009-05-31	$999
				Core i7 Extreme	965X	3.2 GHz	2008-11-17	1× QPI 6.4 GT/s				$999
				Core i7	960	3.2 GHz	1× QPI 4.8 GT/s	2009-10-20			$562
					950	3.06 GHz		2009-05-31
					940	2.93 GHz		2008-11-17
					930	2.8 GHz		2010-02-28			$294
					920	2.66 GHz		2008-11-17			$284
Lynnfield	Performance Desktop		LGA 1156		880	3.06 GHz	95 W	DMI	2× DDR3-1333		2010-05-30	$583
					875K	2.93 GHz	$342				2010-05-30
					870	2.93 GHz	2009-09-08				$562
					870S	2.66 GHz	82 W				2010-07-19	$351
					860	2.8 GHz	95 W				2009-09-08	$284
					860S	2.53 GHz	82 W				2010-01-07	$337
		4 (4)		Core i5	760	2.8 GHz	95 W				2010-07-17	$209
					750	2.66 GHz	95 W				2009-09-08	$196
					750S	2.4 GHz	82 W				2010-01-07	$259
JasperForest	Embedded Desktop	1 (2)	LGA 1366	Celeron	P1053	1.33 GHz			30 W	3× DDR3-800	2 MB	2010-12-02	$160

Intel states the Gainestown processors have six memory channels. Gainestown processors have dual QPI links and have a separate set of memory registers for each link in effect, a multiplexed six-channel system.

Mobile processors

See main article: List of Intel Core i7 processors.

Codename	Market	Cores / Threads	Socket	Processor Branding & Model		Core Clock rate	Turbo	TDP	L3 cache	Interface	Release Date
Clarksfield	Extrememobile	4 (8)	Socket G1rPGA 988	Core i7 Extreme	940XM	2.13 GHz	55 W	8 MB	DMI * 2x DDR3-1333 * PCIe 1 x16 / 2 x8	2010-06-21	$1096
	Extrememobile			Core i7 Extreme	920XM	2.0 GHz	2009-09-23			$1054
	Quad-coremobilePerformandemobile			Core i7	840QM	1.86 GHz	45 W			2010-06-21	$568
					820QM	1.73 GHz				2009-09-23	$546
					740QM	1.73 GHz		6 MB		2010-06-21	$378
					720QM	1.6 GHz		6 MB		2009-09-23	$364

External links

Nehalem processor at Intel.com

Notes and References

Web site: Archived copy . November 30, 2015 . November 29, 2015 . https://web.archive.org/web/20151129231923/http://files.shareholder.com/downloads/INTC/0x0x348508/C9259E98-BE06-42C8-A433-E28F64CB8EF2/TurboBoostWhitePaper.pdf . dead .
Web site: Inside Nehalem: Intel's Future Processor and System. David. Kanter.
Web site: Inside Nehalem: Intel's Future Processor and System. David. Kanter.
Web site: Intel's Sandy Bridge Microarchitecture. David. Kanter.
Web site: Inside Nehalem: Intel's Future Processor and System. December 29, 2014.
Web site: Feature - Intel Core i7 - Nehalem Architecture Dive. bit-tech. December 29, 2014.
Web site: Intel Developer Zone. Intel.
Web site: Nehalem - Microarchitectures - Intel - WikiChip. 2021-02-09. en.wikichip.org. en.
Web site: AnandTech - High-End x86: The Nehalem EX Xeon 7500 and Dell R810. Johan De Gelas. December 29, 2014.
Web site: Intel Core i7 Processor Extreme Datasheet, Volume 2, (Document Number: 320835-002) . November 1, 2008 . March 24, 2009 . Intel.