Intel Nehalem (microarchitecture)

From Wikipedia, the free encyclopedia

Nehalem (pronounced /nəˈheɪləm/^[1]) is the codename for an Intel processor microarchitecture,^[2] successor to the Core microarchitecture.^{[citation needed]} The first processor released with the Nehalem architecture is the desktop Core i7,^[3] which was released on November 15, 2008 in Tokyo and November 17, 2008 in the USA.^[4]

Initial Nehalem processors use the same 45 nm manufacturing methods as Penryn. A working system with two Nehalem processors was shown at Intel Developer Forum Fall 2007,^[5] and a large number of Nehalem systems were shown at Computex in June 2008.

The microarchitecture is named after the Nehalem Native American nation in Oregon.^{[citation needed]} At that stage it was supposed to be the latest evolution of the NetBurst microarchitecture. Since the abandonment of NetBurst, the codename has been recycled and refers to a completely different project, although Nehalem still has some things in common with NetBurst. Nehalem-based microprocessors utilize higher clock speeds and are more energy-efficient than Penryn microprocessors. Hyper-Threading is reintroduced along with an L3 Cache missing from most Core-based microprocessors.

The first computer to use Nehalem-based Xeon processors was the Apple Inc. Mac Pro workstation announced on March 3, 2009.^[6] Nehalem-based Xeon EX processors for larger servers are expected in Q4 2009.^[7] Mobile Nehalem-based processors were introduced in September 2009.

Contents 1 Technology 2 Performance and power improvements 3 Code names 4 Variants 4.1 45 nm processors 4.2 32 nm processors 4.2.1 Westmere 5 Successor 6 See also 7 References 8 External links

[edit] Technology

Various sources have stated the specifications of processors in the Nehalem family:

• Two, four, six, or eight cores
  • 731 million transistors for the quad core variant
• 45 nm manufacturing process
• Integrated memory controller supporting two or three memory channels of DDR3 SDRAM or four FB-DIMM channels
• Integrated graphics processor (IGP) located off-die, but in the same CPU package^[8]
• A new point-to-point processor interconnect, the Intel QuickPath Interconnect, in high-end models, replacing the legacy front side bus
• Integration of PCI Express and Direct Media Interface into the processor in mid-range models, replacing the northbridge
• Simultaneous multithreading (SMT) by multiple cores which enables two threads per core. Intel calls this hyper-threading. Simultaneous multithreading has not been present on a consumer desktop Intel processor since 2006 with the Pentium 4 and Pentium XE. Intel reintroduced SMT with their Atom Architecture.
• Native (monolithic, i.e. all processor cores on a single die) quad- and octa-core processors^[9]
• The following caches:
  • 32 KB L1 instruction and 32 KB L1 data cache per core
  • 256 KB L2 cache per core
  • 4–8 MB L3 cache shared by all cores
• 33% more in-flight micro-ops than Conroe^[10]
• Second-level branch predictor and second-level translation lookaside buffer^[10]
• Modular blocks of components such as cores that can be added and subtracted for varying market segments^[11]

[edit] Performance and power improvements

It has been reported that Nehalem will have a focus on performance, which accounts for the increased core size.^[12] Compared to Penryn, Nehalem will have:

• 1.1x to 1.25x the single-threaded performance or 1.2x to 2x the multithreaded performance at the same power level
• 30% lower power usage for the same performance
• According to a preview from AnandTech "expect a 20–30% overall advantage over Penryn with a 10% increase in power usage. It looks like Intel is on track to delivering just that in Q4."^[13]
• Per Core, clock-for-clock, Nehalem will provide a 15–20% increase in performance compared to Penryn.^[14]

PC Watch found that a Nehalem "Gainestown" processor has 1.6x the SPECint_rate2006 integer performance and 2.4x the SPECfp_rate_2006 floating-point performance of a 3.0 GHz Xeon X5365 "Clovertown" quad-core processor.^[12]

A 2.93 GHz Nehalem "Bloomfield" system has been used to run a 3DMark Vantage benchmark and gave a CPU score of 17,966.^[15] The 2.66 GHz variant scores 16,294. A 2.4 GHz Core 2 Duo E6600 scores 4,300.^[16]

AnandTech tested the Intel QuickPath Interconnect ("QPI", 4.8 GT/s version) and found the copy bandwidth using triple-channel 1066 MHz DDR3 was 12.0 GB/s. A 3.0 GHz Core 2 Quad system using dual-channel 1066 MHz DDR3 achieved 6.9 GB/s.^[17]

Overclocking will be possible with Bloomfield processors and the X58 chipset. The Lynnfield processor will use a PCH removing the need for a northbridge chipset.^[18]

The Nehalem processors are the first to incorporate the SSE 4.2 SIMD instructions, adding 7 new instructions to the SSE 4.1 set available in the Core 2 series.

[edit] Code names

Each combination of a Nehalem/Westmere processor die and package has both a separate codename and a product code. Typically, the same dies are used for uniprocessor (UP) and dual-processor (DP) servers, but using an extra QuickPath link for the inter-processor communication in the DP server variant. Where the Core microarchitecture used four different processor sockets, one for each market segment, Nehalem now uses Socket 1366 for the high-end of both UP and DP machines, and Socket 1156 for the low end UP machines. The name for the UP version of Gulftown is not yet known, its product code is 80613 and can be found in Intel's product database^[19].

	Mobile	Desktop UP Server	DP Server	MP Server
Dual-Core 45 nm Dual-Channel, PCIe, Graphics Core	Auburndale canceled	Havendale canceled
Dual-Core 32 nm Dual-Channel, PCIe, Graphics Core	Arrandale 80617	Clarkdale 80616
Quad-Core 45 nm Dual-Channel, PCIe	Clarksfield 80607	Lynnfield 80605	Jasper Forest 80612
Quad-Core 45 nm Triple-Channel		Bloomfield 80601	Gainestown 80602
Six-Core 32 nm Triple-Channel		Gulftown 80613	Gulftown 80614
Eight-Core 45 nm Triple-Channel				Beckton 80604

[edit] Variants

These tables list all the processors of Nehalem microarchitecture to have been leaked so far. The table is ordered roughly by descending performance, which usually means descending price and power too. Released processors are set in bold.

Notes:

• "Extreme" processors have an unlocked clock multiplier. Thermal design power (TDP) values for CPUs with integrated GPUs include the GPU.
• All variants have 64 KiB L1 cache per core, and 256 KiB L2 cache per core.

[edit] 45 nm processors

Codename	Market	Cores (Threads)	Socket	Brand	Processor No.	Clock rate			Turbo	TDP	Interfaces			L3 cache	Release	1k Unit Price
						Base	Core	Uncore			Chipset	Memory	PCIe
Beckton	MP Server	8 (16)	LGA-1567							130/105/90 W	4x QPI	4x [DDR3 with SMB motherboard]	n/a	24 MB	Q1 2010[20]
Gainestown	DP Server [21]	4 (8)	LGA-1366	Xeon[22]	W5580	133 MHz	3.2 GHz		Yes	130 W	2x QPI 6.4 GT/s	3x DDR3-13331	n/a	8 MB	March 29, 2009[23]	$1500
					X5570		2.93 GHz			95 W						$1286
					X5560		2.8 GHz									$1072
					X5550		2.66 GHz									$858
					E5540		2.53 GHz			80 W	2x 5.86 GT/s	3x DDR3-10661				$744
					E5530		2.4 GHz									$530
					E5520		2.26 GHz									$373
					L5520		2.26 GHz			60 W						$530
		4 (4)			E5506		2.13 GHz		No	80 W	2x 4.8 GT/s	3x DDR3-8001		4 MB		$266
					E5504		2.0 GHz									$224
					L5506		2.13 GHz			60 W						$423
		2 (2)			E5502		1.86 GHz			80 W						$188
Bloomfield	UP Server [24]	4 (8)	LGA-1366	Xeon [25]	W3580	133 MHz	3.33 GHz		Yes	130 W	1x QPI 6.4 GT/s	3x DDR3-1333	n/a	8 MB	August 9, 2009	$999
					W3570		3.2 GHz								March 29, 2009 [25]	$999
					W3550		3.06 GHz				1x QPI 4.8 GT/s	3x DDR3-1066			August 9, 2009	$562
					W3540		2.93 GHz								March 29, 2009 [25]	$562
					W3520		2.66 GHz									$284
Lynnfield	UP Server	4 (8)	LGA 1156	Xeon	X3470	133 MHz	2.93 GHz		Yes	95 W	DMI	2x DDR3-1333		8 MB	September 8, 2009	$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	Extreme/Performance Desktop [26]	4 (8)	LGA-1366	Core i7 Extreme	975 [27]	133 MHz	3.33 GHz	2.66 GHz	Yes	130 W	1x QPI 6.4 GT/s	3x DDR3-1066	n/a	8 MB	May 31, 2009	$999
					965		3.2 GHz								November 17, 2008	$999
				Core i7	960 [2]		3.2 GHz				1x QPI 4.8 GT/s				October 20, 2009	$562
					950 [27]		3.06 GHz	2.13 GHz							May 31, 2009	$562
					940		2.93 GHz								November 17, 2008	$562
					930		2.8 GHz								Q1 2010	$284
					920		2.66 GHz								November 17, 2008	$284
Lynnfield	Performance/Mainstream Desktop	4 (8)	LGA 1156	Core i7	870 [28]	133 MHz	2.93 GHz	2.4 GHz	Yes	95 W	DMI	2x DDR3-1333	1x16 / 2x8	8 MB	September 8, 2009	$562
					860		2.8 GHz									$284
					860s		2.53 GHz			82 W					Q1 2010	$337
		4 (4)		Core i5	750 [29]		2.66 GHz	2.13 GHz		95 W					September 8, 2009	$196
					750s		2.4 GHz			82 W					Q1 2010	$259
Clarksfield	Extreme/Performance Mobile	4 (8)	mPGA-989	Core i7 Extreme	920XM	133 MHz	2.0 GHz		Yes	55 W	DMI	2x DDR3-1333	1x16 / 2x8	8 MB	September 23, 2009	$1054
				Core i7	820QM		1.73 GHz			45 W						$546
					720QM		1.6 GHz							6 MB		$364

• ¹ Though there is only one memory controller and it has only three channels, 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^[30] so thereby, in effect, a multiplexed six channel system.

The Havendale and Auburndale variants (which contained Gilo and Ironlake) have been cancelled.^[31]

[edit] 32 nm processors

Codename	Market	Cores (Threads)	Socket	Brand	Processor No.	Clock rate			Turbo	TDP	Interfaces			L3 cache	Release	1k Unit Price
						Base	Core	GPU			Chipset	Memory	PCIe
Gulftown[32]	Extreme/Performance Desktop	6 (12)	LGA-1366	Core i7	980X	133 MHz	3.06 GHz	n/a	Yes	130 W	2x QPI	3x DDR3-1066	n/a	12 MB	March, 2010	$999
Clarkdale[33]	Mainstream/Value Desktop	2 (4)	LGA-1156	Core i5	670	133 MHz	3.46 GHz	733 MHz	Yes	73 W	DMI	2x DDR3-1333	1 x16	4 MB	January 2010	$284
					661		3.33 GHz	900 MHz		87 W
					660			733 MHz		73 W						$196
					650		3.2 GHz									$176
				Core i3	540		3.06 GHz		No							$143
					530		2.93 GHz									$123
		2 (2)		Pentium	G6950		2.8 GHz	533 MHz				2x DDR3-1066		3 MB		$87
Arrandale[34]	Mainstream/Value Mobile	2 (4)	mPGA-989	Core i7	620M	133 MHz	2.66 GHz	766 MHz	Yes	35 W	DMI	2x DDR3-1066 [3]	1 x16	4 MB	Q3 2010[35]	$332
					640LM		2.13 GHz	566 MHz		25 W						$332
					620LM		2.0 GHz									$305
					640UM		1.2 GHz	500 MHz		18 W						$305
					620UM		1.06 GHz									$278
				Core i5	540M		2.53 GHz	766 MHz		35 W				3 MB		$257
					520M		2.4 GHz								January 2010[36]	$225
					430M		2.26 GHz
				Core i3	350M		2.26 GHz	667 MHz	No	35W
					330M		2.13 GHz			35W

For the desktop, Gulftown is to be an "Extreme Edition" CPU and so will coexist with Bloomfield. [4] It will have Turbo Boost and similar clock speeds to Bloomfield. [5]

Lynnfield and Clarksfield may make the 32 nm transition in the middle of 2010^[34], sometime after Q2 [6], while Beckton will move to 32 nm at the end of 2010.^[37] The 32 nm CPUs will not have significantly different clock speeds compared to 45 nm CPUs.^[38]

Clarkdale and Arrandale contain the 32 nm dual core processor Hillel and the 45 nm integrated graphics device Ironlake, and support switchable graphics.^[31][38][39] The lowest-power variant of Arrandale may have a 10 W CPU TDP, and a maximum clock speed of 1.6 GHz. [7]

A successor to Bloomfield and entry level server chips are also expected in Q2 2010.^[31][34][37]

[edit] Westmere

Westmere (formerly Nehalem-C) is the name given to the 32 nm die shrink of Nehalem. Westmere should be ready for a Q4 2009 release provided that Intel stays on target with its roadmap. However, it appears that the bulk of Westmere's versions, excluding mobile versions, will be released sometime in Q1, 2010.^[40][41] Westmere's features and improvements from Nehalem have been reported as follows:

• Native six-core, and possibly dual-die hex-core (12-cores), processors.^[42]
  • The successor to Bloomfield and Gainestown is six-core.
• A new set of instructions that gives over 3x the encryption and decryption rate of Advanced Encryption Standard (AES) processes compared to before.^[43]
  • Delivers seven new instructions that will be used by the AES algorithm. Also an instruction called PCLMULQDQ that will perform carry-less multiplication.^[44] These instructions will allow the processor to perform hardware-accelerated encryption, not only resulting in faster execution but also protecting against software targeted attacks.
  • AES-NI may be included in the integrated graphics of Westmere.
• Integrated graphics, released at the same time as the processor.
• Improved virtualization latency.^[45]
• New virtualization capability: "VMX Unrestricted mode support" -- which allows 16-bit guests to run. (real mode and big real mode).

[edit] Successor

The successor to Nehalem and Westmere will be Sandy Bridge, scheduled for release in 2011, according to Intel roadmaps.^[46] The successor to Sandy Bridge will be Haswell, scheduled for release in 2012. It will come with a new cache subsystem, a FMA (fused multiply-add) unit, and a vector coprocessor.^[14]

[edit] See also

v • d • e Intel processors Discontinued pre-x86 4004 · 4040 · 8008 · 8080 · 8085 x86-16 (16 bit) 8086 · 8088 · 80186 · 80188 · 80286 x86-32/IA-32 (32 bit) 80386 · 80486 · Pentium · Pentium Pro · Pentium II · Pentium III · Pentium 4 · Pentium M · Core · Celeron M · Celeron D · A100 x86-64/EM64T (64 bit) Pentium 4 (Some) · Pentium D · Pentium Extreme Edition · Celeron D (Some) Other iAPX 432 — RISC: i860 · i960 · StrongARM · XScale Current EP80579 · Atom · Celeron · Pentium Dual-Core · Core (2 · i5 · i7) · Xeon · Itanium Upcoming Tukwila · Moorestown · Core i3 Lists CPU Sockets · Chipsets · Microarchitectures · Processors · Future Processors · Codenames · GMA Atom · Celeron · Core (2 · i5 · i7) · Itanium · Pentium (Pro · II · III · 4 · D · Dual-Core · M) · Xeon Microarchitectures P5   P5 based cores 0.90 μm P5 0.60 μm P54C 0.35 μm P54CS · P55CS 0.25 μm Tillamook P6   P6 based cores 0.50 μm P6 0.35 μm P6 · Klamath 0.25 μm Mendocino · Dixon · Tonga · Covington · Deschutes · Katmai · Drake · Tanner 180 nm Coppermine · Cascades 130 nm Tualatin · Banias 90 nm Dothan · Stealey 65 nm Yonah · Sossaman NetBurst   NetBurst based cores 180 nm Willamette · Foster 130 nm Northwood · Gallatin · Prestonia 90 nm Tejas and Jayhawk · Prescott · Smithfield · Nocona · Irwindale · Cranford · Potomac · Paxville 65 nm Cedar Mill · Presler · Dempsey · Tulsa Core   Core based cores 65 nm Merom-L · Merom · Conroe-L · Allendale · Conroe · Kentsfield · Woodcrest · Clovertown · Tigerton 45 nm Penryn · Penryn-QC · Wolfdale · Yorkfield · Wolfdale-DP · Harpertown · Dunnington Atom   Atom based cores 45 nm Silverthorne · Diamondville Nehalem   Nehalem based cores 45 nm Clarksfield · Lynnfield · Jasper Forest · Bloomfield · Gainestown · Beckton 32 nm Arrandale · Clarkdale · Gulftown Future Larrabee · Sandy Bridge · Haswell

[edit] References

[edit] External links