Cell microprocessor implementations - Revision history

Cewbot: Fixing broken anchor: #Cell based Blades→most alike anchor IBM BladeCenter#Cell based

2023-08-17T10:00:07Z

Fixing broken anchor: #Cell based Blades→most alike anchor IBM BladeCenter#Cell based

← Previous revision		Revision as of 10:00, 17 August 2023
Line 134:		Line 134:
	On 12 March 2007, IBM announced that it started producing 65 nm Cells in its East Fishkill fab. The chips produced there are apparently only for IBMs own Cell [[Computing blade\|blade]] servers, which were the first to get the 65 nm Cells. Sony introduced the third generation of the PS3 in November 2007, the 40GB model without PS2-compatibility which was [https://www.engadget.com/2007/10/30/40gb-ps3-features-65nm-chips-lower-power-consumption/ confirmed] to use the 65 nm Cell. Thanks to the shrunk Cell, power consumption was reduced from 200{{nbsp}}W to 135{{nbsp}}W.		On 12 March 2007, IBM announced that it started producing 65 nm Cells in its East Fishkill fab. The chips produced there are apparently only for IBMs own Cell [[Computing blade\|blade]] servers, which were the first to get the 65 nm Cells. Sony introduced the third generation of the PS3 in November 2007, the 40GB model without PS2-compatibility which was [https://www.engadget.com/2007/10/30/40gb-ps3-features-65nm-chips-lower-power-consumption/ confirmed] to use the 65 nm Cell. Thanks to the shrunk Cell, power consumption was reduced from 200{{nbsp}}W to 135{{nbsp}}W.

	At first it was only known that the 65 nm-Cells clock up to 6 GHz and run on 1.3{{nbsp}}V core voltage, as [http://news.spong.com/article/11413?cb=936 demonstrated] on the [[ISSCC]] 2007. This would have given the chip a theoretical peak performance of 384{{nbsp}}GFLOPS in FP8 quarter precision (48{{nbsp}}GFLOPs in FP64 dual precision), a significant improvement to the 204.8{{nbsp}}GFLOPS peak (25.6{{nbsp}}GFLOPs FP64 dual precision) that a 90 nm 3.2 GHz Cell could provide with 8 active SPUs. IBM further announced it implemented new power-saving features and a dual power supply for the SRAM array. This version was not yet the long-rumoured "Cell+" with enhanced Double Precision floating point performance, which first saw the light of day mid-2008 in the [[IBM Roadrunner\|Roadrunner supercomputer]] in the form of [[QS22#Cell based ~~Blades~~\|QS22]] PowerXCell blades. Although IBM talked about and even showed higher-clocked Cells before, clock speed has remained constant at 3.2 GHz, even for the double precision enabled "Cell+" of the Roadrunner. By keeping clockspeed constant, IBM has instead opted to reduce power consumption. PowerXCell clusters even best IBMs [[Blue Gene]] clusters (371{{nbsp}}MFLOPS/watt), which are far more power-efficient already than clusters made up of conventional CPUs (265{{nbsp}}MFLOPS/watt and lower).		At first it was only known that the 65 nm-Cells clock up to 6 GHz and run on 1.3{{nbsp}}V core voltage, as [http://news.spong.com/article/11413?cb=936 demonstrated] on the [[ISSCC]] 2007. This would have given the chip a theoretical peak performance of 384{{nbsp}}GFLOPS in FP8 quarter precision (48{{nbsp}}GFLOPs in FP64 dual precision), a significant improvement to the 204.8{{nbsp}}GFLOPS peak (25.6{{nbsp}}GFLOPs FP64 dual precision) that a 90 nm 3.2 GHz Cell could provide with 8 active SPUs. IBM further announced it implemented new power-saving features and a dual power supply for the SRAM array. This version was not yet the long-rumoured "Cell+" with enhanced Double Precision floating point performance, which first saw the light of day mid-2008 in the [[IBM Roadrunner\|Roadrunner supercomputer]] in the form of [[QS22#Cell based\|QS22]] PowerXCell blades. Although IBM talked about and even showed higher-clocked Cells before, clock speed has remained constant at 3.2 GHz, even for the double precision enabled "Cell+" of the Roadrunner. By keeping clockspeed constant, IBM has instead opted to reduce power consumption. PowerXCell clusters even best IBMs [[Blue Gene]] clusters (371{{nbsp}}MFLOPS/watt), which are far more power-efficient already than clusters made up of conventional CPUs (265{{nbsp}}MFLOPS/watt and lower).

	===Future editions in CMOS===		===Future editions in CMOS===

Beland: convert special characters found by Wikipedia:Typo Team/moss (via WP:JWB)

2023-08-10T00:20:03Z

convert special characters found by Wikipedia:Typo Team/moss (via WP:JWB)

← Previous revision		Revision as of 00:20, 10 August 2023
Line 106:		Line 106:
	! Voltage !! Frequency !! Power !! Die Temp.		! Voltage !! Frequency !! Power !! Die Temp.
	\|-		\|-
	\| 0.9 V \|\| 2.0 GHz \|\| {{0}}1 W \|\| 25&~~#8239~~;°C		\| 0.9 V \|\| 2.0 GHz \|\| {{0}}1 W \|\| 25 °C
	\|-		\|-
	\| 0.9 V \|\| 3.0 GHz \|\| {{0}}2 W \|\| 27&~~#8239~~;°C		\| 0.9 V \|\| 3.0 GHz \|\| {{0}}2 W \|\| 27 °C
	\|-		\|-
	\| 1.0 V \|\| 3.8 GHz \|\| {{0}}3 W \|\| 31&~~#8239~~;°C		\| 1.0 V \|\| 3.8 GHz \|\| {{0}}3 W \|\| 31 °C
	\|-		\|-
	\| 1.1 V \|\| 4.0 GHz \|\| {{0}}4 W \|\| 38&~~#8239~~;°C		\| 1.1 V \|\| 4.0 GHz \|\| {{0}}4 W \|\| 38 °C
	\|-		\|-
	\| 1.2 V \|\| 4.4 GHz \|\| {{0}}7 W \|\| 47&~~#8239~~;°C		\| 1.2 V \|\| 4.4 GHz \|\| {{0}}7 W \|\| 47 °C
	\|-		\|-
	\| 1.3 V \|\| 5.0 GHz \|\| 11 W \|\| 63&~~#8239~~;°C		\| 1.3 V \|\| 5.0 GHz \|\| 11 W \|\| 63 °C
	\|}		\|}
	As tested by IBM under a heavy transformation and lighting workload [average IPC of 1.4], the performance profile of this implementation for a single SPU processor is qualified as follows:		As tested by IBM under a heavy transformation and lighting workload [average IPC of 1.4], the performance profile of this implementation for a single SPU processor is qualified as follows:

79.245.165.147: /* Implementation */

2023-07-31T15:30:38Z

Implementation

← Previous revision		Revision as of 15:30, 31 July 2023
Line 8:		Line 8:

	===First edition Cell on 90 nm CMOS===		===First edition Cell on 90 nm CMOS===
		⚫	{\| class="wikitable" align="right" style="font-size:90%"

⚫	IBM has published information concerning two different versions of Cell in this process, an early engineering sample designated ''DD1'', and an enhanced version designated ''DD2'' intended for production.
⚫	{\| class="wikitable" align="~~left~~"
	\|+ Known Cell variants in 90 nm process		\|+ Known Cell variants in 90 nm process
	! Designation !! Die area !! First disclosed !! Enhancement		! Designation !! Die area !! First disclosed !! Enhancement
Line 18:		Line 16:
	\| DD2 \|\| 235 mm<sup>2</sup> \|\| Cool Chips April 2005 \|\| Enhanced PPE core		\| DD2 \|\| 235 mm<sup>2</sup> \|\| Cool Chips April 2005 \|\| Enhanced PPE core
	\|}		\|}
		⚫	IBM has published information concerning two different versions of Cell in this process, an early engineering sample designated ''DD1'', and an enhanced version designated ''DD2'' intended for production.
⚫	{{clear}}

	The main enhancement in DD2 was a small lengthening of the die to accommodate a larger PPE core, which is reported to "contain more SIMD/vector execution resources"{{ref\|dtwang3}}.		The main enhancement in DD2 was a small lengthening of the die to accommodate a larger PPE core, which is reported to "contain more SIMD/vector execution resources"{{ref\|dtwang3}}.
	Some preliminary information released by IBM references the DD1 variant. As a result, some early journalistic accounts of the Cell's capabilities now differ from production hardware.		Some preliminary information released by IBM references the DD1 variant. As a result, some early journalistic accounts of the Cell's capabilities now differ from production hardware.

		⚫	{{clear}}
⚫	====Cell floorplan====
⚫	Powerpoint material accompanying an STI presentation given by Dr Peter Hofstee], includes a photograph of the DD2 Cell die overdrawn with functional unit boundaries which are also captioned by name, which reveals the breakdown of silicon area by function unit as follows:

		⚫	====Cell floorplan====
	<!-- I wasn't able to square-up the image when measuring dimensions off screen, since I only had area, but not WH so these are a little more approx. than they might have been; I had WH for the included SPU but the top four SPU measure differently than the bottom four in the way they overdrew the boundaries, leaving me with the same problem -->		<!-- I wasn't able to square-up the image when measuring dimensions off screen, since I only had area, but not WH so these are a little more approx. than they might have been; I had WH for the included SPU but the top four SPU measure differently than the bottom four in the way they overdrew the boundaries, leaving me with the same problem -->
	{\| class="wikitable"		{\| class="wikitable" align="right" style="font-size:90%"
	\|+ Cell function units and footprint		\|+ Cell function units and footprint
	! Cell function unit !! Area !! Description		! Cell function unit !! Area !! Description
Line 49:		Line 47:
	\| Rambus FlexIO \|\| {{0}}5.7% \|\| External signalling for I/O pins		\| Rambus FlexIO \|\| {{0}}5.7% \|\| External signalling for I/O pins
	\|}		\|}
		⚫	Powerpoint material accompanying an STI presentation given by Dr Peter Hofstee], includes a photograph of the DD2 Cell die overdrawn with functional unit boundaries which are also captioned by name, which reveals the breakdown of silicon area by function unit as follows:

	{{clear}}		{{clear}}

	====SPE floorplan====		====SPE floorplan====
⚫	Additional details concerning the internal SPE implementation have been disclosed by IBM engineers, including [[Peter Hofstee]], IBM's chief architect of the synergistic processing element, in a scholarly IEEE publication.{{ref\|90nmsoi}}

⚫	This document includes a photograph of the 2.54 × 5.81 mm SPE, as implemented in 90-nm [[Silicon on Insulator\|SOI]]. In this technology, the SPE contains 21 million transistors of which 14 million are contained in arrays (a term presumably designating register files and the local store) and 7 million transistors are logic. This photograph is overdrawn with functional unit boundaries, which are also captioned by name, which reveals the breakdown of silicon area by function unit as follows:

	<!-- if I had found a way I would have made the table smaller than the surrounding text, esp. if the surrounding text is largish, but that's asking a lot -->		<!-- if I had found a way I would have made the table smaller than the surrounding text, esp. if the surrounding text is largish, but that's asking a lot -->
	{\| class="wikitable" <!-- bad effects at certain Firefox sizings with style="text-align:left" --> align="right" style="~~margin~~: ~~1em auto 1em auto~~"		{\| class="wikitable" <!-- bad effects at certain Firefox sizings with style="text-align:left" --> align="right" style="font-size:90%"
	\|+ SPU function units and footprint		\|+ SPU function units and footprint
	! SPU function<br>unit !! Area !! Description !! Pipe		! SPU function<br>unit !! Area !! Description !! Pipe
Line 95:		Line 89:
	\|}		\|}
	<!-- OK, I see the method for aligning columns by decimal points in the table help. Not for this chicken. Some diehard can suffer or this can wait until MediaWiki is fixed properly. -->		<!-- OK, I see the method for aligning columns by decimal points in the table help. Not for this chicken. Some diehard can suffer or this can wait until MediaWiki is fixed properly. -->

		⚫	Additional details concerning the internal SPE implementation have been disclosed by IBM engineers, including [[Peter Hofstee]], IBM's chief architect of the synergistic processing element, in a scholarly IEEE publication.{{ref\|90nmsoi}}

		⚫	This document includes a photograph of the 2.54 mm × 5.81 mm SPE, as implemented in 90-nm [[Silicon on Insulator\|SOI]]. In this technology, the SPE contains 21 million transistors of which 14 million are contained in arrays (a term presumably designating register files and the local store) and 7 million transistors are logic. This photograph is overdrawn with functional unit boundaries, which are also captioned by name, which reveals the breakdown of silicon area by function unit as follows:

	Understanding the dispatch pipes is important to write efficient code. In the SPU architecture, two instructions can be dispatched (started) in each clock cycle using dispatch pipes designated ''even'' and ''odd''. The two pipes provide different execution units, as shown in the table above. As IBM partitioned this, most of the arithmetic instructions execute on the ''even'' pipe, while most of the memory instructions execute on the ''odd'' pipe. The permute unit is closely associated with memory instructions as it serves to pack and unpack data structures located in memory into the SIMD multiple operand format that the SPU computes on most efficiently.		Understanding the dispatch pipes is important to write efficient code. In the SPU architecture, two instructions can be dispatched (started) in each clock cycle using dispatch pipes designated ''even'' and ''odd''. The two pipes provide different execution units, as shown in the table above. As IBM partitioned this, most of the arithmetic instructions execute on the ''even'' pipe, while most of the memory instructions execute on the ''odd'' pipe. The permute unit is closely associated with memory instructions as it serves to pack and unpack data structures located in memory into the SIMD multiple operand format that the SPU computes on most efficiently.
Line 101:		Line 99:

	Understanding the limitations of the restrictive two pipeline design is one of the key concepts a programmer must grasp to write efficient SPU code at the lowest level of abstraction. For programmers working at higher levels of abstraction, a good compiler will automatically balance pipeline concurrency where possible.		Understanding the limitations of the restrictive two pipeline design is one of the key concepts a programmer must grasp to write efficient SPU code at the lowest level of abstraction. For programmers working at higher levels of abstraction, a good compiler will automatically balance pipeline concurrency where possible.
			{{clear}}

	====SPE power and performance====		====SPE power and performance====
	{\| class="wikitable" align="right"		{\| class="wikitable" align="right" style="font-size:90%"
	\|+ Relationship of speed to temperature		\|+ Relationship of speed to temperature
	! Voltage !! Frequency !! Power !! Die Temp.		! Voltage !! Frequency !! Power !! Die Temp.
Line 128:		Line 127:

	IBM has publicly announced their intention to implement Cell on a future technology below the 90 nm node to improve power consumption. Reduced power consumption could ''potentially'' allow the existing design to be boosted to 5 GHz or above without exceeding the thermal constraints of existing products.		IBM has publicly announced their intention to implement Cell on a future technology below the 90 nm node to improve power consumption. Reduced power consumption could ''potentially'' allow the existing design to be boosted to 5 GHz or above without exceeding the thermal constraints of existing products.
			{{clear}}

	====Cell at 65 nm====		====Cell at 65 nm====

79.245.165.147: /* SPE floorplan */

2023-07-31T15:25:23Z

SPE floorplan

← Previous revision		Revision as of 15:25, 31 July 2023
Line 60:		Line 60:
	{\| class="wikitable" <!-- bad effects at certain Firefox sizings with style="text-align:left" --> align="right" style="margin: 1em auto 1em auto"		{\| class="wikitable" <!-- bad effects at certain Firefox sizings with style="text-align:left" --> align="right" style="margin: 1em auto 1em auto"
	\|+ SPU function units and footprint		\|+ SPU function units and footprint
	! SPU function unit !! Area ~~(%)~~ !! Description !! Pipe		! SPU function<br>unit !! Area !! Description !! Pipe
	\|-		\|-
	\| single precision \|\| 10.0 \|\| single precision FP execution unit \|\| even		\| single precision \|\| 10.0% \|\| single precision FP execution unit \|\| even
	\|-		\|-
	\| double precision \|\| 4.4 \|\| double precision FP execution unit \|\| even		\| double precision \|\| {{0}}4.4% \|\| double precision FP execution unit \|\| even
	\|-		\|-
	\| simple fixed \|\| 3.25 \|\| fixed point execution unit \|\| even		\| simple fixed \|\| {{0}}3.25% \|\| fixed point execution unit \|\| even
	\|-		\|-
	\| issue control \|\| 2.5 \|\| feeds execution units		\| issue control \|\| {{0}}2.5% \|\| feeds execution units
	\|-		\|-
	\| forward macro \|\| 3.75 \|\| feeds execution units		\| forward macro \|\| {{0}}3.75% \|\| feeds execution units
	\|-		\|-
	\| GPR \|\| 6.25 \|\| general purpose register file		\| GPR \|\| {{0}}6.25% \|\| general purpose register file
	\|-		\|-
	\| permute \|\| 3.25 \|\| permute execution unit \|\| odd		\| permute \|\| {{0}}3.25% \|\| permute execution unit \|\| odd
	\|-		\|-
	\| branch \|\| 2.5 \|\| branch execution unit \|\| odd		\| branch \|\| {{0}}2.5% \|\| branch execution unit \|\| odd
	\|-		\|-
	\| channel \|\| 6.75 \|\| channel interface (three discrete blocks) \|\| odd		\| channel \|\| {{0}}6.75% \|\| channel interface (three discrete blocks) \|\| odd
	\|-		\|-
	\| LS0–LS3 \|\| 30.0 \|\| four 64 KiB blocks of local store \|\| odd		\| LS0–LS3 \|\| 30.0% \|\| four 64 KiB blocks of local store \|\| odd
	\|-		\|-
	\| MMU \|\| 4.75 \|\| memory management unit		\| MMU \|\| {{0}}4.75% \|\| memory management unit
	\|-		\|-
	\| DMA \|\| 7.5 \|\| direct memory access unit		\| DMA \|\| {{0}}7.5% \|\| direct memory access unit
	\|-		\|-
	\| BIU \|\| 9.0 \|\| bus interface unit		\| BIU \|\| {{0}}9.0% \|\| bus interface unit
	\|-		\|-
	\| RTB \|\| 2.5 \|\| array built-in test block (ABIST)		\| RTB \|\| {{0}}2.5% \|\| array built-in test block (ABIST)
	\|-		\|-
	\| ATO \|\| 1.6 \|\| atomic unit for atomic DMA updates		\| ATO \|\| {{0}}1.6% \|\| atomic unit for atomic DMA updates
	\|-		\|-
	\| HB \|\| 0.5 \|\| obscure		\| HB \|\| {{0}}0.5% \|\| obscure
	\|}		\|}
	<!-- OK, I see the method for aligning columns by decimal points in the table help. Not for this chicken. Some diehard can suffer or this can wait until MediaWiki is fixed properly. -->		<!-- OK, I see the method for aligning columns by decimal points in the table help. Not for this chicken. Some diehard can suffer or this can wait until MediaWiki is fixed properly. -->

79.245.165.147: /* SPE power and performance */

2023-07-31T15:22:42Z

SPE power and performance

← Previous revision		Revision as of 15:22, 31 July 2023
Line 103:		Line 103:

	====SPE power and performance====		====SPE power and performance====
		⚫	{\| class="wikitable" align="right"
⚫	As tested by IBM under a heavy transformation and lighting workload [average IPC of 1.4], the performance profile of this implementation for a single SPU processor is qualified as follows:

⚫	{\| class="wikitable" align="right~~" style="margin: 1em auto 1em auto~~"
	\|+ Relationship of speed to temperature		\|+ Relationship of speed to temperature
	! Voltage ~~(V)~~ !! Frequency ~~(GHz)~~ !! Power ~~(W)~~ !! Die ~~temp~~. ~~(°C)~~		! Voltage !! Frequency !! Power !! Die Temp.
	\|-		\|-
	\| 0.9 \|\| 2.0 \|\| 1 \|\| 25		\| 0.9 V \|\| 2.0 GHz \|\| {{0}}1 W \|\| 25 °C
	\|-		\|-
	\| 0.9 \|\| 3.0 \|\| 2 \|\| 27		\| 0.9 V \|\| 3.0 GHz \|\| {{0}}2 W \|\| 27 °C
	\|-		\|-
	\| 1.0 \|\| 3.8 \|\| 3 \|\| 31		\| 1.0 V \|\| 3.8 GHz \|\| {{0}}3 W \|\| 31 °C
	\|-		\|-
	\| 1.1 \|\| 4.0 \|\| 4 \|\| 38		\| 1.1 V \|\| 4.0 GHz \|\| {{0}}4 W \|\| 38 °C
	\|-		\|-
	\| 1.2 \|\| 4.4 \|\| 7 \|\| 47		\| 1.2 V \|\| 4.4 GHz \|\| {{0}}7 W \|\| 47 °C
	\|-		\|-
	\| 1.3 \|\| 5.0 \|\| 11 \|\| 63		\| 1.3 V \|\| 5.0 GHz \|\| 11 W \|\| 63 °C
	\|}		\|}
		⚫	As tested by IBM under a heavy transformation and lighting workload [average IPC of 1.4], the performance profile of this implementation for a single SPU processor is qualified as follows:

	The entry for 2.0 GHz operation at 0.9 V represents a low power configuration. Other entries show the peak stable operating frequency achieved with each voltage increment. As a general rule in CMOS circuits, power dissipation rises in a rough relationship to V{{sup\|2}}F, the square of the voltage times the operating frequency.		The entry for 2.0 GHz operation at 0.9 V represents a low power configuration. Other entries show the peak stable operating frequency achieved with each voltage increment. As a general rule in CMOS circuits, power dissipation rises in a rough relationship to V{{sup\|2}}F, the square of the voltage times the operating frequency.

79.245.165.147: /* Cell floorplan */

2023-07-31T15:20:03Z

Cell floorplan

← Previous revision		Revision as of 15:20, 31 July 2023
Line 27:		Line 27:

	<!-- I wasn't able to square-up the image when measuring dimensions off screen, since I only had area, but not WH so these are a little more approx. than they might have been; I had WH for the included SPU but the top four SPU measure differently than the bottom four in the way they overdrew the boundaries, leaving me with the same problem -->		<!-- I wasn't able to square-up the image when measuring dimensions off screen, since I only had area, but not WH so these are a little more approx. than they might have been; I had WH for the included SPU but the top four SPU measure differently than the bottom four in the way they overdrew the boundaries, leaving me with the same problem -->
			{\| class="wikitable"

	{\| class="wikitable" style="margin: 1em auto 1em auto"
	\|+ Cell function units and footprint		\|+ Cell function units and footprint
	! Cell function unit !! Area ~~(%)~~ !! Description		! Cell function unit !! Area !! Description
	\|-		\|-
	\| XDR interface \|\| 5.7 \|\| Interface to Rambus system memory		\| XDR interface \|\| {{0}}5.7% \|\| Interface to Rambus system memory
	\|-		\|-
	\| memory controller \|\| 4.4 \|\| Manages external memory and L2 cache		\| memory controller \|\| {{0}}4.4% \|\| Manages external memory and L2 cache
	\|-		\|-
	\| 512 KiB L2 cache \|\| 10.3 \|\| Cache memory for the PPE		\| 512 KiB L2 cache \|\| 10.3% \|\| Cache memory for the PPE
	\|-		\|-
	\| PPE core \|\| 11.1 \|\| PowerPC processor		\| PPE core \|\| 11.1% \|\| PowerPC processor
	\|-		\|-
	\| test \|\| 2.0 \|\| Unspecified "test and decode logic"		\| test \|\| {{0}}2.0% \|\| Unspecified "test and decode logic"
	\|-		\|-
	\| EIB \|\| 3.1 \|\| Element interconnect bus linking processors		\| EIB \|\| {{0}}3.1% \|\| Element interconnect bus linking processors
	\|-		\|-
	\| SPE (each) × 8 \|\| 6.2 \|\| Synergistic coprocessing element		\| SPE (each) × 8 \|\| {{0}}6.2% \|\| Synergistic coprocessing element
	\|-		\|-
	\| I/O controller \|\| 6.6 \|\| External I/O logic		\| I/O controller \|\| {{0}}6.6% \|\| External I/O logic
	\|-		\|-
	\| Rambus FlexIO \|\| 5.7 \|\| External signalling for I/O pins		\| Rambus FlexIO \|\| {{0}}5.7% \|\| External signalling for I/O pins
	\|}		\|}

Harshdaank: Added info

2023-05-08T06:43:04Z

Added info

← Previous revision		Revision as of 06:43, 8 May 2023
Line 1:		Line 1:
	{{context\|date=January 2020}}		{{context\|date=January 2020}}
	The first commercial [[Cell microprocessor]], the Cell BE, was designed for the Sony PlayStation 3. IBM designed the PowerXCell 8i for use in the [[Roadrunner supercomputer]].<ref>		'''Cell''' microprocessors are multi-core processors that use cellular architecture for high performance distributed computing. The first commercial [[Cell microprocessor]], the Cell BE, was designed for the Sony PlayStation 3. IBM designed the PowerXCell 8i for use in the [[Roadrunner supercomputer]].<ref>
	Kevin J. Barker, Kei Davis, Adolfy Hoisie, Darren J. Kerbyson, Mike Lang, Scott Pakin, Jose C. Sancho.		Kevin J. Barker, Kei Davis, Adolfy Hoisie, Darren J. Kerbyson, Mike Lang, Scott Pakin, Jose C. Sancho.
	[https://www.academia.edu/4460100/Entering_the_petaflop_era_the_architecture_and_performance_of_Roadrunner "Entering the Petaflop Era:The Architecture and Performance of Roadrunner"].		[https://www.academia.edu/4460100/Entering_the_petaflop_era_the_architecture_and_performance_of_Roadrunner "Entering the Petaflop Era:The Architecture and Performance of Roadrunner"].

CoolingGibbon at 20:59, 30 December 2022

2022-12-30T20:59:27Z

← Previous revision		Revision as of 20:59, 30 December 2022
Line 152:		Line 152:

	[[Category:Cell BE architecture]]		[[Category:Cell BE architecture]]
	[[Category:Microarchitectures]]

Liruqi: Fix en wikitable style syntax, src: https://github.com/liruqi/wikipedia/blob/master/wikitable.py

2022-12-28T04:48:43Z

Fix en wikitable style syntax, src: https://github.com/liruqi/wikipedia/blob/master/wikitable.py

← Previous revision		Revision as of 04:48, 28 December 2022
Line 59:		Line 59:

	<!-- if I had found a way I would have made the table smaller than the surrounding text, esp. if the surrounding text is largish, but that's asking a lot -->		<!-- if I had found a way I would have made the table smaller than the surrounding text, esp. if the surrounding text is largish, but that's asking a lot -->
	{\| class="wikitable" <!-- bad effects at certain Firefox sizings with style="text-align:left" --> style="margin: 1em auto 1em auto~~" align="right~~"		{\| class="wikitable" <!-- bad effects at certain Firefox sizings with style="text-align:left" --> align="right" style="margin: 1em auto 1em auto"
	\|+ SPU function units and footprint		\|+ SPU function units and footprint
	! SPU function unit !! Area (%) !! Description !! Pipe		! SPU function unit !! Area (%) !! Description !! Pipe
Line 106:		Line 106:
	As tested by IBM under a heavy transformation and lighting workload [average IPC of 1.4], the performance profile of this implementation for a single SPU processor is qualified as follows:		As tested by IBM under a heavy transformation and lighting workload [average IPC of 1.4], the performance profile of this implementation for a single SPU processor is qualified as follows:

	{\| class="wikitable" style="margin: 1em auto 1em auto~~" align="right~~"		{\| class="wikitable" align="right" style="margin: 1em auto 1em auto"
	\|+ Relationship of speed to temperature		\|+ Relationship of speed to temperature
	! Voltage (V) !! Frequency (GHz) !! Power (W) !! Die temp. (°C)		! Voltage (V) !! Frequency (GHz) !! Power (W) !! Die temp. (°C)

89.107.6.24: /* SPE floorplan */

2020-05-11T02:31:18Z

SPE floorplan

← Previous revision		Revision as of 02:31, 11 May 2020
Line 81:		Line 81:
	\| channel \|\| 6.75 \|\| channel interface (three discrete blocks) \|\| odd		\| channel \|\| 6.75 \|\| channel interface (three discrete blocks) \|\| odd
	\|-		\|-
	\| ~~LS0-LS3~~ \|\| 30.0 \|\| four 64 KiB blocks of local store \|\| odd		\| LS0–LS3 \|\| 30.0 \|\| four 64 KiB blocks of local store \|\| odd
	\|-		\|-
	\| MMU \|\| 4.75 \|\| memory management unit		\| MMU \|\| 4.75 \|\| memory management unit