Jump to content

SPARC

From Wikipedia, the free encyclopedia
(Redirected from SPARC V9)

SPARC
DesignerSun Microsystems (acquired by Oracle Corporation)[1][2]
Bits64-bit (32 → 64)
Introduced1986; 38 years ago (1986) (production)
1987; 37 years ago (1987) (shipments)
VersionV9 (1993) / OSA2017
DesignRISC
TypeLoad–store
EncodingFixed
BranchingCondition code
EndiannessBi (Big → Bi)
Page size8 KB (4 KB → 8 KB)
ExtensionsVIS 1.0, 2.0, 3.0, 4.0
OpenYes, and royalty free
Registers
General-purpose31 (G0 = 0; non-global registers use register windows)
Floating point32 (usable as 32 single-precision, 32 double-precision, or 16 quad-precision)
A Sun UltraSPARC II microprocessor (1997)

SPARC (Scalable Processor ARChitecture) is a reduced instruction set computer (RISC) instruction set architecture originally developed by Sun Microsystems.[1][2] Its design was strongly influenced by the experimental Berkeley RISC system developed in the early 1980s. First developed in 1986 and released in 1987,[3][2] SPARC was one of the most successful early commercial RISC systems, and its success led to the introduction of similar RISC designs from many vendors through the 1980s and 1990s.

The first implementation of the original 32-bit architecture (SPARC V7) was used in Sun's Sun-4 computer workstation and server systems, replacing their earlier Sun-3 systems based on the Motorola 68000 series of processors. SPARC V8 added a number of improvements that were part of the SuperSPARC series of processors released in 1992. SPARC V9, released in 1993, introduced a 64-bit architecture and was first released in Sun's UltraSPARC processors in 1995. Later, SPARC processors were used in symmetric multiprocessing (SMP) and non-uniform memory access (CC-NUMA) servers produced by Sun, Solbourne, and Fujitsu, among others.

The design was turned over to the SPARC International trade group in 1989, and since then its architecture has been developed by its members. SPARC International is also responsible for licensing and promoting the SPARC architecture, managing SPARC trademarks (including SPARC, which it owns), and providing conformance testing. SPARC International was intended to grow the SPARC architecture to create a larger ecosystem; SPARC has been licensed to several manufacturers, including Atmel, Bipolar Integrated Technology, Cypress Semiconductor, Fujitsu, Matsushita and Texas Instruments. Due to SPARC International, SPARC is fully open, non-proprietary and royalty-free.

As of 2024, the latest commercial high-end SPARC processors are Fujitsu's SPARC64 XII (introduced in September 2017 for its SPARC M12 server) and Oracle's SPARC M8 introduced in September 2017 for its high-end servers.

On September 1, 2017, after a round of layoffs that started in Oracle Labs in November 2016, Oracle terminated SPARC design after completing the M8. Much of the processor core development group in Austin, Texas, was dismissed, as were the teams in Santa Clara, California, and Burlington, Massachusetts.[4][5]

Fujitsu will also discontinue their SPARC production (has already shifted to producing their own ARM-based CPUs), after two "enhanced" versions of Fujitsu's older SPARC M12 server in 2020–22 (formerly planned for 2021) and again in 2026–27, end-of-sale in 2029, of UNIX servers and a year later for their mainframe and end-of-support in 2034 "to promote customer modernization".[6]

Features

[edit]

The SPARC architecture was heavily influenced by the earlier RISC designs, including the RISC I and II from the University of California, Berkeley and the IBM 801. These original RISC designs were minimalist, including as few features or op-codes as possible and aiming to execute instructions at a rate of almost one instruction per clock cycle. This made them similar to the MIPS architecture in many ways, including the lack of instructions such as multiply or divide. Another feature of SPARC influenced by this early RISC movement is the branch delay slot.

The SPARC processor usually contains as many as 160 general-purpose registers. According to the "Oracle SPARC Architecture 2015" specification an "implementation may contain from 72 to 640 general-purpose 64-bit" registers.[7] At any point, only 32 of them are immediately visible to software — 8 are a set of global registers (one of which, g0, is hard-wired to zero, so only seven of them are usable as registers) and the other 24 are from the stack of registers. These 24 registers form what is called a register window, and at function call/return, this window is moved up and down the register stack. Each window has eight local registers and shares eight registers with each of the adjacent windows. The shared registers are used for passing function parameters and returning values, and the local registers are used for retaining local values across function calls.

The "scalable" in SPARC comes from the fact that the SPARC specification allows implementations to scale from embedded processors up through large server processors, all sharing the same core (non-privileged) instruction set. One of the architectural parameters that can scale is the number of implemented register windows; the specification allows from three to 32 windows to be implemented, so the implementation can choose to implement all 32 to provide maximum call stack efficiency, or to implement only three to reduce cost and complexity of the design, or to implement some number between them. Other architectures that include similar register file features include Intel i960, IA-64, and AMD 29000.

The architecture has gone through several revisions. It gained hardware multiply and divide functionality in version 8.[8][9] 64-bit (addressing and data) were added to the version 9 SPARC specification published in 1994.[10]

In SPARC version 8, the floating-point register file has 16 double-precision registers. Each of them can be used as two single-precision registers, providing a total of 32 single-precision registers. An odd–even number pair of double-precision registers can be used as a quad-precision register, thus allowing 8 quad-precision registers. SPARC Version 9 added 16 more double-precision registers (which can also be accessed as 8 quad-precision registers), but these additional registers can not be accessed as single-precision registers. No SPARC CPU implements quad-precision operations in hardware as of 2024.[11]

Tagged add and subtract instructions perform adds and subtracts on values checking that the bottom two bits of both operands are 0 and reporting overflow if they are not. This can be useful in the implementation of the run time for ML, Lisp, and similar languages that might use a tagged integer format.

The endianness of the 32-bit SPARC V8 architecture is purely big-endian. The 64-bit SPARC V9 architecture uses big-endian instructions, but can access data in either big-endian or little-endian byte order, chosen either at the application instruction (load–store) level or at the memory page level (via an MMU setting). The latter is often used for accessing data from inherently little-endian devices, such as those on PCI buses.

History

[edit]

There have been three major revisions of the architecture. The first published version was the 32-bit SPARC version 7 (V7) in 1986. SPARC version 8 (V8), an enhanced SPARC architecture definition, was released in 1990. The main differences between V7 and V8 were the addition of integer multiply and divide instructions, and an upgrade from 80-bit "extended-precision" floating-point arithmetic to 128-bit "quad-precision" arithmetic. SPARC V8 served as the basis for IEEE Standard 1754-1994, an IEEE standard for a 32-bit microprocessor architecture.

SPARC version 9, the 64-bit SPARC architecture, was released by SPARC International in 1993. It was developed by the SPARC Architecture Committee consisting of Amdahl Corporation, Fujitsu, ICL, LSI Logic, Matsushita, Philips, Ross Technology, Sun Microsystems, and Texas Instruments. Newer specifications always remain compliant with the full SPARC V9 Level 1 specification.

In 2002, the SPARC Joint Programming Specification 1 (JPS1) was released by Fujitsu and Sun, describing processor functions which were identically implemented in the CPUs of both companies ("Commonality"). The first CPUs conforming to JPS1 were the UltraSPARC III by Sun and the SPARC64 V by Fujitsu. Functionalities which are not covered by JPS1 are documented for each processor in "Implementation Supplements".

At the end of 2003, JPS2 was released to support multicore CPUs. The first CPUs conforming to JPS2 were the UltraSPARC IV by Sun and the SPARC64 VI by Fujitsu.

In early 2006, Sun released an extended architecture specification, UltraSPARC Architecture 2005. This includes not only the non-privileged and most of the privileged portions of SPARC V9, but also all the architectural extensions developed through the processor generations of UltraSPARC III, IV IV+ as well as CMT extensions starting with the UltraSPARC T1 implementation:

  • the VIS 1 and VIS 2 instruction set extensions and the associated GSR register
  • multiple levels of global registers, controlled by the GL register
  • Sun's 64-bit MMU architecture
  • privileged instructions ALLCLEAN, OTHERW, NORMALW, and INVALW
  • access to the VER register is now hyperprivileged
  • the SIR instruction is now hyperprivileged

In 2007, Sun released an updated specification, UltraSPARC Architecture 2007, to which the UltraSPARC T2 implementation complied.

In August 2012, Oracle Corporation made available a new specification, Oracle SPARC Architecture 2011, which besides the overall update of the reference, adds the VIS 3 instruction set extensions and hyperprivileged mode to the 2007 specification.[12]

In October 2015, Oracle released SPARC M7, the first processor based on the new Oracle SPARC Architecture 2015 specification.[7][13] This revision includes VIS 4 instruction set extensions and hardware-assisted encryption and silicon secured memory (SSM).[14]

SPARC architecture has provided continuous application binary compatibility from the first SPARC V7 implementation in 1987 through the Sun UltraSPARC Architecture implementations.

Among various implementations of SPARC, Sun's SuperSPARC and UltraSPARC-I were very popular, and were used as reference systems for SPEC CPU95 and CPU2000 benchmarks. The 296 MHz UltraSPARC-II is the reference system for the SPEC CPU2006 benchmark.

Architecture

[edit]

SPARC is a load–store architecture (also known as a register–register architecture); except for the load/store instructions used to access memory, all instructions operate on the registers, in accordance with the RISC design principles.

A SPARC processor includes an integer unit (IU) that performs integer load, store, and arithmetic operations.[15]: 9 [10]: 15–16  It may include a floating-point unit (FPU) that performs floating-point operations[15]: 9 [10]: 15–16  and, for SPARC V8, may include a co-processor (CP) that performs co-processor-specific operations; the architecture does not specify what functions a co-processor would perform, other than load and store operations.[15]: 9 

Registers

[edit]

The SPARC architecture has an overlapping register window scheme. At any instant, 32 general-purpose registers are visible. A Current Window Pointer (CWP) variable in the hardware points to the current set. The total size of the register file is not part of the architecture, allowing more registers to be added as the technology improves, up to a maximum of 32 windows in SPARC V7 and V8 as CWP is 5 bits and is part of the PSR register.

In SPARC V7 and V8 CWP will usually be decremented by the SAVE instruction (used by the SAVE instruction during the procedure call to open a new stack frame and switch the register window), or incremented by the RESTORE instruction (switching back to the call before returning from the procedure). Trap events (interrupts, exceptions or TRAP instructions) and RETT instructions (returning from traps) also change the CWP. For SPARC V9, CWP register is decremented during a RESTORE instruction, and incremented during a SAVE instruction. This is the opposite of PSR.CWP's behavior in SPARC V8. This change has no effect on nonprivileged instructions.

Window Addressing
Register group Mnemonic Register address Availability
global G0...G7 R[00]...R[07] always the same ones, G0 being zero always
out O0...O7 R[08]...R[15] to be handed over to, and returned from, the called subroutine, as its "in"
local L0...L7 R[16]...R[23] truly local to the current subroutine
in I0...I7 R[24]...R[31] handed over from the caller, and returned to the caller, as its "out"

SPARC registers are shown in the figure above.

There is also a non-windowed Y register, used by the multiply-step, integer multiply, and integer divide instructions.[15]: 32 

A SPARC V8 processor with an FPU includes 32 32-bit floating-point registers, each of which can hold one single-precision IEEE 754 floating-point number. An even–odd pair of floating-point registers can hold one double-precision IEEE 754 floating-point number, and a quad-aligned group of four floating-point registers can hold one quad-precision IEEE 754 floating-point number.[15]: 10 

A SPARC V9 processor with an FPU includes:[10]: 36–40 

  • 32 32-bit floating-point registers, each of which can hold one single-precision IEEE 754 floating-point number;
  • 32 64-bit floating-point registers, each of which can hold one double-precision IEEE 754 floating-point number;
  • 16 128-bit floating-point registers, each of which can hold one quad-precision IEEE 754 floating-point number.

The registers are organized as a set of 64 32-bit registers, with the first 32 being used as the 32-bit floating-point registers, even–odd pairs of all 64 registers being used as the 64-bit floating-point registers, and quad-aligned groups of four floating-point registers being used as the 128-bit floating-point registers.

Floating-point registers are not windowed; they are all global registers.[10]: 36–40 

Instruction formats

[edit]

All SPARC instructions occupy a full 32-bit word and start on a word boundary. Four formats are used, distinguished by the first two bits. All arithmetic and logical instructions have 2 source operands and 1 destination operand.[16] RD is the "destination register", where the output of the operation is deposited. The majority of SPARC instructions have at least this register, so it is placed near the "front" of the instruction format. RS1 and RS2 are the "source registers", which may or may not be present, or replaced by a constant.

SPARC instruction formats
Type Bit
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
SETHI format 00 RD 100 Immediate constant 22 bits
I Branch format 00 A icc 010 Displacement constant 22 bits
F Branch format 00 A fcc 110 Displacement constant 22 bits
C Branch format 00 A ccc 111 Displacement constant 22 bits
CALL disp 01 PC-relative displacement
Arithmetic register 10 RD opcode RS1 0 0 RS2
Arithmetic immediate 10 RD opcode RS1 1 Immediate constant 13 bits
FPU operation 10 FD 110100/110101 FS1 opf FS2
CP operation 10 RD 110110/110111 RS1 opc RS2
JMPL register 10 RD 111000 RS1 0 0 RS2
JMPL immediate 10 RD 111000 RS1 1 Immediate constant 13 bits
LD/ST register 11 RD opcode RS1 0 0 RS2
LD/ST immediate 11 RD opcode RS1 1 Immediate constant 13 bits

Instructions

[edit]

Loads and stores

[edit]

Load and store instructions have a three-operand format, in that they have two operands representing values for the address and one operand for the register to read or write to. The address is created by adding the two address operands to produce an address. The second address operand may be a constant or a register. Loads take the value at the address and place it in the register specified by the third operand, whereas stores take the value in the register specified by the first operand and place it at the address. To make this more obvious, the assembler language indicates address operands using square brackets with a plus sign separating the operands, instead of using a comma-separated list. Examples:[16]

ld [%L1+%L2],%L3  !load the 32-bit value at address %L1+%L2 and put the value into %L3
ld [%L1+8],%L2    !load the value at %L1+8 into %L2
ld [%L1],%L2      !as above, but no offset, which is the same as +%G0
st %L1,[%I2]      !store the value in %L1 into the location stored in %I2
st %G0,[%I1+8]    !clear the memory at %I1+8

Due to the widespread use of non-32-bit data, such as 16-bit or 8-bit integral data or 8-bit bytes in strings, there are instructions that load and store 16-bit half-words and 8-bit bytes, as well as instructions that load 32-bit words. During a load, those instructions will read only the byte or half-word at the indicated location and then either fill the rest of the target register with zeros (unsigned load or with the value of the uppermost bit of the byte or half-word (signed load). During a store, those instructions discard the upper bits in the register and store only the lower bits. There are also instructions for loading double-precision values used for floating-point arithmetic, reading or writing eight bytes from the indicated register and the "next" one, so if the destination of a load is L1, L1 and L2 will be set. The complete list of load and store instructions for the general-purpose registers in 32-bit SPARC is LD, ST, LDUB (unsigned byte), LDSB (signed byte), LDUH (unsigned half-word), LDSH (signed half-word), LDD (load double), STB (store byte), STH (store half-word), STD (store double).[16]

In SPARC V9, registers are 64-bit, and the LD instruction, renamed LDUW, clears the upper 32 bits in the register and loads the 32-bit value into the lower 32 bits, and the ST instruction, renamed STW, discards the upper 32 bits of the register and stores only the lower 32 bits. The new LDSW instruction sets the upper bits in the register to the value of the uppermost bit of the word and loads the 32-bit value into the lower bits. The new LDX instruction loads a 64-bit value into the register, and the STX instruction stores all 64 bits of the register.

The LDF, LDDF, and LDQF instructions load a single-precision, double-precision, or quad-precision value from memory into a floating-point register; the STF, STDF, and STQF instructions store a single-precision, double-precision, or quad-precision floating-point register into memory.

The memory barrier instruction, MEMBAR, serves two interrelated purposes: it articulates order constraints among memory references and facilitates explicit control over the completion of memory references. For example, all effects of the stores that appear prior to the MEMBAR instruction must be made visible to all processors before any loads following the MEMBAR can be executed.[17]

ALU operations

[edit]

Arithmetic and logical instructions also use a three-operand format, with the first two being the operands and the last being the location to store the result. The middle operand can be a register or a 13-bit signed integer constant; the other operands are registers. Any of the register operands may point to G0; pointing the result to G0 discards the results, which can be used for tests. Examples include:[16]

add %L1,%L2,%L3   !add the values in %L1 and %L2 and put the result in %L3
add %L1,1,%L1     !increment %L1
add %G0,%G0,%L4   !clear any value in %L4

The list of mathematical instructions is ADD, SUB, AND, OR, XOR, and negated versions ANDN, ORN, and XNOR. One quirk of the SPARC design is that most arithmetic instructions come in pairs, with one version setting the NZVC condition code bits in the status register, and the other not setting them, with the default being not to set the codes. This is so that the compiler has a way to move instructions around when trying to fill delay slots. If one wants the condition codes to be set, this is indicated by adding cc to the instruction:[16]

subcc %L1,10,%G0  !compare %L1 to 10 and ignore the result, but set the flags

add and sub also have another modifier, X, which indicates whether the operation should set the carry bit:

addx %L1,100,%L1  !add 100 to the value in %L1 and track carry

SPARC V7 does not have multiplication or division instructions, but it does have MULSCC, which does one step of a multiplication testing one bit and conditionally adding the multiplicand to the product. This was because MULSCC can complete over one clock cycle in keeping with the RISC philosophy. SPARC V8 added UMUL (unsigned multiply), SMUL (signed multiply), UDIV (unsigned divide), and SDIV (signed divide) instructions, with both versions that do not update the condition codes and versions that do. MULSCC and the multiply instructions use the Y register to hold the upper 32 bits of the product; the divide instructions use it to hold the upper 32 bits of the dividend. The RDY instruction reads the value of the Y register into a general-purpose register; the WRY instruction writes the value of a general-purpose register to the Y register.[15]: 32  SPARC V9 added MULX, which multiplies two 64-bit values and produces a 64-bit result, SDIVX, which divides a 64-bit signed dividend by a 64-bit signed divisor and produces a 64-bit signed quotient, and UDIVX, which divides a 64-bit unsigned dividend by a 64-bit unsigned divisor and produces a 64-bit signed quotient; none of those instructions use the Y register.[10]: 199 

Branching

[edit]

Conditional branches test condition codes in a status register, as seen in many instruction sets such the IBM System/360 architecture and successors and the x86 architecture. This means that a test and branch is normally performed with two instructions; the first is an ALU instruction that sets the condition codes, followed by a branch instruction that examines one of those flags. The SPARC does not have specialized test instructions; tests are performed using normal ALU instructions with the destination set to %G0. For instance, to test if a register holds the value 10 and then branch to code that handles it, one would:

subcc %L1,10,%G0 !subtract 10 from %L1, setting the zero flag if %L1 is 10
be WASEQUAL      !if the zero flag is set, branch to the address marked WASEQUAL

In a conditional branch instruction, the icc or fcc field specifies the condition being tested. The 22-bit displacement field is the address, relative to the current PC, of the target, in words, so that conditional branches can go forward or backward up to 8 megabytes. The ANNUL (A) bit is used to get rid of some delay slots. If it is 0 in a conditional branch, the delay slot is executed as usual. If it is 1, the delay slot is only executed if the branch is taken. If it is not taken, the instruction following the conditional branch is skipped.

There are a wide variety of conditional branches: BA (branch always, essentially a jmp), BN (branch never), BE (equals), BNE (not equals), BL (less than), BLE (less or equal), BLEU (less or equal, unsigned), BG (greater), BGE (greater or equal), BGU (greater unsigned), BPOS (positive), BNEG (negative), BCC (carry clear), BCS (carry set), BVC (overflow clear), BVS (overflow set).[15]: 119–120 

The FPU and CP have sets of condition codes separate from the integer condition codes and from each other; two additional sets of branch instructions were defined to test those condition codes. Adding an F to the front of the branch instruction in the list above performs the test against the FPU's condition codes,[15]: 121–122  while, in SPARC V8, adding a C tests the flags in the otherwise undefined CP.[15]: 123–124 

The CALL (jump to subroutine) instruction uses a 30-bit program counter-relative word offset. As the target address is specifying the start of a word, not a byte, 30-bits is all that is needed to reach any address in the 4 gigabyte address space.[16] The CALL instruction deposits the return address in register R15, also known as output register O7.

The JMPL (jump and link) instruction is a three-operand instruction, with two operands representing values for the target address and one operand for a register in which to deposit the return address. The address is created by adding the two address operands to produce a 32-bit address. The second address operand may be a constant or a register.

Large constants

[edit]

As the instruction opcode takes up some bits of the 32-bit instruction word, there is no way to load a 32-bit constant using a single instruction. This is significant because addresses are manipulated through registers and they are 32-bits. To ease this, the special-purpose SETHI instruction copies its 22-bit immediate operand into the high-order 22 bits of any specified register, and sets each of the low-order 10 bits to 0. In general use, SETHI is followed by an or instruction with only the lower 10 bits of the value set. To ease this, the assembler includes the %hi(X) and %lo(X) macros. For example:[16]

sethi %hi(0x89ABCDEF),%L1       !sets the upper 22 bits of L1
or    %L1,%lo(0x89ABCDEF),%L1   !sets the lower 10 bits of L1 by ORing

The hi and lo macros are performed at assembly time, not runtime, so it has no performance hit yet makes it clearer that L1 is set to a single value, not two unrelated ones. To make this even easier, the assembler also includes a "synthetic instruction", set, that performs these two operations in a single line:

set   0x89ABCDEF,% L1

This outputs the two instructions above if the value is larger than 13 bits, otherwise it will emit a single ld with the value.[16]

Synthetic instructions

[edit]

As noted earlier, the SPARC assembler uses "synthetic instructions" to ease common coding tasks. Additional examples include (among others):[16]

SPARC synthetic instructions
mnemonic actual output purpose
nop sethi 0,%g0 do nothing
clr %reg or %g0,%g0,%reg set a register to zero
clr [address] st %g0,[address] set a memory address to zero
clrh [address] sth %g0,[address] set the half-word at memory address to zero
clrb [address] stb %g0,[address] set the byte at memory address to zero
cmp %reg1,%reg2 subcc %reg1,%reg2,%g0 compare two registers, set codes, discard results
cmp %reg,const subcc %reg,const,%g0 compare register with constant
mov %reg1,%reg2 or %g0,%reg1,%reg2 copy value from one register to another
mov const,%reg or %g0,const,%reg copy constant value into a register
inc %reg add %reg,1,%reg increment a register
inccc %reg addcc %reg,1,%reg increment a register, set conditions
dec %reg sub %reg,1,%reg decrement a register
deccc %reg subcc %reg,1,%reg decrement a register, set conditions
not %reg xnor %reg,%g0,%reg flip the bits in a register
neg %reg sub %g0,%reg,%reg two's complement a register
tst %reg orcc %reg,%g0,%g0 test whether the value in a register is > 0, 0, or < 0

SPARC architecture licensees

[edit]

The following organizations have licensed the SPARC architecture:

Implementations

[edit]
Name (codename) Model Frequency (MHz) Arch. version Year Total threads[note 1] Process (nm) Transistors (millions) Die size (mm2) IO pins Power (W) Voltage (V) L1 Dcache (KB) L1 Icache (KB) L2 cache (KB) L3 cache (KB)
SPARC MB86900 Fujitsu[1][3][2] 14.28–33 V7 1986 1×1=1 1300 0.11 256 0–128 (unified) none none
SPARC Various[note 2] 14.28–40 V7 1989–1992 1×1=1 800–1300 ~0.1–1.8 160–256 0–128 (unified) none none
MN10501 (KAP) Solbourne Computer,

Matsushita[18]

33–36 V8 1990–1991 1x1=1 1.0[19] 8 8 0–256 none
microSPARC I (Tsunami) TI TMS390S10 40–50 V8 1992 1×1=1 800 0.8 225? 288 2.5 5 2 4 none none
SuperSPARC I (Viking) TI TMX390Z50 / Sun STP1020 33–60 V8 1992 1×1=1 800 3.1 293 14.3 5 16 20 0–2048 none
SPARClite Fujitsu MB8683x 66–108 V8E 1992 1×1=1 144, 176 2.5/3.3–5.0 V, 2.5–3.3 V 1, 2, 8, 16 1, 2, 8, 16 none none
hyperSPARC (Colorado 1) Ross RT620A 40–90 V8 1993 1×1=1 500 1.5 5? 0 8 128–256 none
microSPARC II (Swift) Fujitsu MB86904 / Sun STP1012 60–125 V8 1994 1×1=1 500 2.3 233 321 5 3.3 8 16 none none
hyperSPARC (Colorado 2) Ross RT620B 90–125 V8 1994 1×1=1 400 1.5 3.3 0 8 128–256 none
SuperSPARC II (Voyager) Sun STP1021 75–90 V8 1994 1×1=1 800 3.1 299 16 16 20 1024–2048 none
hyperSPARC (Colorado 3) Ross RT620C 125–166 V8 1995 1×1=1 350 1.5 3.3 0 8 512–1024 none
TurboSPARC Fujitsu MB86907 160–180 V8 1996 1×1=1 350 3.0 132 416 7 3.5 16 16 512 none
UltraSPARC (Spitfire) Sun STP1030 143–167 V9 1995 1×1=1 470 3.8 315 521 30[note 3] 3.3 16 16 512–1024 none
UltraSPARC (Hornet) Sun STP1030 200 V9 1995 1×1=1 420 5.2 265 521 3.3 16 16 512–1024 none
hyperSPARC (Colorado 4) Ross RT620D 180–200 V8 1996 1×1=1 350 1.7 3.3 16 16 512 none
SPARC64 Fujitsu (HAL) 101–118 V9 1995 1×1=1 400 Multichip 286 50 3.8 128 128
SPARC64 II Fujitsu (HAL) 141–161 V9 1996 1×1=1 350 Multichip 286 64 3.3 128 128
SPARC64 III Fujitsu (HAL) MBCS70301 250–330 V9 1998 1×1=1 240 17.6 240 2.5 64 64 8192
UltraSPARC IIs (Blackbird) Sun STP1031 250–400 V9 1997 1×1=1 350 5.4 149 521 25[note 4] 2.5 16 16 1024 or 4096 none
UltraSPARC IIs (Sapphire-Black) Sun STP1032 / STP1034 360–480 V9 1999 1×1=1 250 5.4 126 521 21[note 5] 1.9 16 16 1024–8192 none
UltraSPARC IIi (Sabre) Sun SME1040 270–360 V9 1997 1×1=1 350 5.4 156 587 21 1.9 16 16 256–2048 none
UltraSPARC IIi (Sapphire-Red) Sun SME1430 333–480 V9 1998 1×1=1 250 5.4 587 21[note 6] 1.9 16 16 2048 none
UltraSPARC IIe (Hummingbird) Sun SME1701 400–500 V9 1999 1×1=1 180 Al 370 13[note 7] 1.5–1.7 16 16 256 none
UltraSPARC IIi (IIe+) (Phantom) Sun SME1532 550–650 V9 2000 1×1=1 180 Cu 370 17.6 1.7 16 16 512 none
SPARC64 GP Fujitsu SFCB81147 400–563 V9 2000 1×1=1 180 30.2 217 1.8 128 128 8192
SPARC64 GP -- 600–810 V9 1×1=1 150 30.2 1.5 128 128 8192
SPARC64 IV Fujitsu MBCS80523 450–810 V9 2000 1×1=1 130 128 128 2048
UltraSPARC III (Cheetah) Sun SME1050 600 JPS1 2001 1×1=1 180 Al 29 330 1368 53 1.6 64 32 8192 none
UltraSPARC III (Cheetah) Sun SME1052 750–900 JPS1 2001 1×1=1 130 Al 29 1368 1.6 64 32 8192 none
UltraSPARC III Cu (Cheetah+) Sun SME1056 900–1200 JPS1 2001 1×1=1 130 Cu 29 232 1368 50[note 8] 1.6 64 32 8192 none
UltraSPARC IIIi (Jalapeño) Sun SME1603 1064–1593 JPS1 2003 1×1=1 130 87.5 206 959 52 1.3 64 32 1024 none
SPARC64 V (Zeus) Fujitsu 1100–1350 JPS1 2003 1×1=1 130 190 289 269 40 1.2 128 128 2048
SPARC64 V+ (Olympus-B) Fujitsu 1650–2160 JPS1 2004 1×1=1 90 400 297 279 65 1 128 128 4096
UltraSPARC IV (Jaguar) Sun SME1167 1050–1350 JPS2 2004 1×2=2 130 66 356 1368 108 1.35 64 32 16384 none
UltraSPARC IV+ (Panther) Sun SME1167A 1500–2100 JPS2 2005 1×2=2 90 295 336 1368 90 1.1 64 64 2048 32768
UltraSPARC T1 (Niagara) Sun SME1905 1000–1400 UA2005 2005 4×8=32 90 300 340 1933 72 1.3 8 16 3072 none
SPARC64 VI (Olympus-C) Fujitsu 2150–2400 JPS2 2007 2×2=4 90 540 422 120–150 1.1 128×2 128×2 4096–6144 none
UltraSPARC T2 (Niagara 2) Sun SME1908A 1000–1600 UA2007 2007 8×8=64 65 503 342 1831 95 1.1–1.5 8 16 4096 none
UltraSPARC T2 Plus (Victoria Falls) Sun SME1910A 1200–1600 UA2007 2008 8×8=64 65 503 342 1831 8 16 4096 none
SPARC64 VII (Jupiter)[20] Fujitsu 2400–2880 JPS2 2008 2×4=8 65 600 445 150 64×4 64×4 6144 none
UltraSPARC "RK" (Rock)[21] Sun SME1832 2300 ???? canceled[22] 2×16=32 65 ? 396 2326 ? ? 32 32 2048 ?
SPARC64 VIIIfx (Venus)[23][24] Fujitsu 2000 JPS2 / HPC-ACE 2009 1×8=8 45 760 513 1271 58 ? 32×8 32×8 6144 none
LEON2FT Atmel AT697F 100 V8 2009 1×1=1 180 196 1 1.8/3.3 16 32 —|none
SPARC T3 (Rainbow Falls) Oracle/Sun 1650 UA2007 2010 8×16=128 40[25] ???? 371 ? 139 ? 8 16 6144 none
Galaxy FT-1500 NUDT (China) 1800 UA2007? 201? 8×16=128 40 ???? ??? ? 65 ? 16×16 16×16 512×16 4096
SPARC64 VII+ (Jupiter-E or M3)[26][27] Fujitsu 2667–3000 JPS2 2010 2×4=8 65 160 64×4 64×4 12288 none
LEON3FT Cobham Gaisler GR712RC 100 V8E 2011 1×2=2 180 1.5[note 9] 1.8/3.3 4x4Kb 4x4Kb none none
R1000 MCST (Russia) 1000 JPS2 2011 1×4=4 90 180 128 15 1, 1.8, 2.5 32 16 2048 none
SPARC T4 (Yosemite Falls)[28] Oracle 2850–3000 OSA2011 2011 8×8=64 40 855 403 ? 240 ? 16×8 16×8 128×8 4096
SPARC64 IXfx[29][30][31] Fujitsu 1850 JPS2 / HPC-ACE 2012 1x16=16 40 1870 484 1442 110 ? 32×16 32×16 12288 none
SPARC64 X (Athena)[32] Fujitsu 2800 OSA2011 / HPC-ACE 2012 2×16=32 28 2950 587.5 1500 270 ? 64×16 64×16 24576 none
SPARC T5 Oracle 3600 OSA2011 2013 8×16=128 28 1500 478 ? ? ? 16×16 16×16 128×16 8192
SPARC M5[33] Oracle 3600 OSA2011 2013 8×6=48 28 3900 511 ? ? ? 16×6 16×6 128×6 49152
SPARC M6[34] Oracle 3600 OSA2011 2013 8×12=96 28 4270 643 ? ? ? 16×12 16×12 128×12 49152
SPARC64 X+ (Athena+)[35] Fujitsu 3200–3700 OSA2011 / HPC-ACE 2014 2×16=32 28 2990 600 1500 392 ? 64×16 64×16 24M none
SPARC64 XIfx[36] Fujitsu 2200 JPS2 / HPC-ACE2 2014 1×(32+2)=34 20 3750 ? 1001 ? ? 64×34 64×34 12M×2 none
SPARC M7[37][38] Oracle 4133 OSA2015 2015 8×32=256 20 >10,000 ? ? ? ? 16×32 16×32 256×24 65536
SPARC S7[39][40] Oracle 4270 OSA2015 2016 8×8=64 20 ???? ? ? ? ? 16×8 16×8 256×2+256×4 16384
SPARC64 XII[41] Fujitsu 4250 OSA201? / HPC-ACE 2017 8×12=96 20 5500 795 1860 ? ? 64×12 64×12 512×12 32768
SPARC M8[42][43] Oracle 5000 OSA2017 2017 8×32=256 20 ? ? ? ? ? 32×32 16×32 128×32+256×8 65536
LEON4 Cobham Gaisler GR740 250[note 10] V8E 2017 1×4=4 32 1.2/2.5/3.3 4x4 4x4 2048 none
R2000 MCST (Russia) 2000 ? 2018 1×8=8 28 500 ? ? ? ? ? ? none
LEON5 Cobham Gaisler V8E 2019 ? ? ? ? 16–8192 none
Name (codename) Model Frequency (MHz) Arch. version Year Total threads[note 1] Process (nm) Transistors (millions) Die size (mm2) IO pins Power (W) Voltage (V) L1 Dcache (KB) L1 Icache (KB) L2 cache (KB) L3 cache (KB)

Notes:

  1. ^ a b Threads per core × number of cores
  2. ^ Various SPARC V7 implementations were produced by Fujitsu, LSI Logic, Weitek, Texas Instruments, Cypress and Temic. A SPARC V7 processor generally consisted of several discrete chips, usually comprising an integer unit (IU), a floating-point unit (FPU), a memory management unit (MMU) and cache memory. Conversely, the Atmel (now Microchip Technology) TSC695 is a single-chip SPARC V7 implementation.
  3. ^ @167 MHz
  4. ^ @250 MHz
  5. ^ @400 MHz
  6. ^ @440 MHz
  7. ^ max. @500 MHz
  8. ^ @1200 MHz
  9. ^ excluding I/O buses
  10. ^ nominal; specification from 100 to 424 MHz depending on attached RAM capabilities

Operating system support

[edit]

SPARC machines have generally used Sun's SunOS, Solaris, JavaOS, or OpenSolaris including derivatives illumos and OpenIndiana, but other operating systems have also been used, such as NeXTSTEP, RTEMS, FreeBSD, OpenBSD, NetBSD, and Linux.

In 1993, Intergraph announced a port of Windows NT to the SPARC architecture,[44] but it was later cancelled.

In October 2015, Oracle announced a "Linux for SPARC reference platform".[45]

Open source implementations

[edit]

Several fully open source implementations of the SPARC architecture exist:

  • LEON, a 32-bit radiation-tolerant, SPARC V8 implementation, designed especially for space use. Source code is written in VHDL, and licensed under the GPL.
  • OpenSPARC T1, released in 2006, a 64-bit, 32-thread implementation conforming to the UltraSPARC Architecture 2005 and to SPARC Version 9 (Level 1). Source code is written in Verilog, and licensed under many licenses. Most OpenSPARC T1 source code is licensed under the GPL. Source based on existent open source projects will continue to be licensed under their current licenses. Binary programs are licensed under a binary software license agreement.
  • S1, a 64-bit Wishbone compliant CPU core based on the OpenSPARC T1 design. It is a single UltraSPARC V9 core capable of 4-way SMT. Like the T1, the source code is licensed under the GPL.
  • OpenSPARC T2, released in 2008, a 64-bit, 64-thread implementation conforming to the UltraSPARC Architecture 2007 and to SPARC Version 9 (Level 1). Source code is written in Verilog, and licensed under many licenses. Most OpenSPARC T2 source code is licensed under the GPL. Source based on existing open source projects will continue to be licensed under their current licenses. Binary programs are licensed under a binary Software License Agreement.

A fully open source simulator for the SPARC architecture also exists:

  • RAMP Gold, a 32-bit, 64-thread SPARC Version 8 implementation, designed for FPGA-based architecture simulation. RAMP Gold is written in ~36,000 lines of SystemVerilog, and licensed under the BSD licenses.

Supercomputers

[edit]

For HPC loads Fujitsu builds specialized SPARC64 fx processors with a new instruction extensions set, called HPC-ACE (High Performance Computing – Arithmetic Computational Extensions).

Fujitsu's K computer ranked No. 1 in the TOP500 June 2011 and November 2011 lists. It combines 88,128 SPARC64 VIIIfx CPUs, each with eight cores, for a total of 705,024 cores—almost twice as many as any other system in the TOP500 at that time. The K Computer was more powerful than the next five systems on the list combined, and had the highest performance-to-power ratio of any supercomputer system.[46] It also ranked No. 6 in the Green500 June 2011 list, with a score of 824.56 MFLOPS/W.[47] In the November 2012 release of TOP500, the K computer ranked No. 3, using by far the most power of the top three.[48] It ranked No. 85 on the corresponding Green500 release.[49] Newer HPC processors, IXfx and XIfx, were included in recent PRIMEHPC FX10 and FX100 supercomputers.

Tianhe-2 (TOP500 No. 1 as of November 2014[50]) has a number of nodes with Galaxy FT-1500 OpenSPARC-based processors developed in China. However, those processors did not contribute to the LINPACK score.[51][52]

See also

[edit]
  • ERC32 — based on SPARC V7 specification
  • Ross Technology, Inc. — a SPARC microprocessor developer during the 1980s and 1990s
  • Sparcle — a modified SPARC with multiprocessing support used by the MIT Alewife project
  • LEON — a space rated SPARC V8 processor.
  • R1000 — a Russian quad-core microprocessor based on SPARC V9 specification
  • Galaxy FT-1500 — a Chinese 16-core OpenSPARC-based processor

References

[edit]
  1. ^ a b c "Fujitsu to take ARM into the realm of Super". The CPU Shack Museum. June 21, 2016. Archived from the original on June 30, 2019. Retrieved June 30, 2019.
  2. ^ a b c d "Timeline". SPARC International. Archived from the original on April 24, 2019. Retrieved June 30, 2019.
  3. ^ a b "Fujitsu SPARC". cpu-collection.de. Archived from the original on August 6, 2016. Retrieved June 30, 2019.
  4. ^ Vaughan-Nichols, Steven J. (September 5, 2017). "Sun set: Oracle closes down last Sun product lines". ZDNet. Archived from the original on September 10, 2017. Retrieved September 11, 2017.
  5. ^ Nichols, Shaun (August 31, 2017). "Oracle finally decides to stop prolonging the inevitable, begins hardware layoffs". The Register. Archived from the original on September 12, 2017. Retrieved September 11, 2017.
  6. ^ "Roadmap: Fujitsu Global". www.fujitsu.com. Retrieved February 15, 2022.
  7. ^ a b "Oracle SPARC Architecture 2015: One Architecture ... Multiple Innovative Implementations" (PDF). Draft D1.0.0. January 12, 2016. Archived (PDF) from the original on April 24, 2016. Retrieved June 13, 2016. IMPL. DEP. #2-V8: An Oracle SPARC Architecture implementation may contain from 72 to 640 general-purpose 64-bit R registers. This corresponds to a grouping of the registers into MAXPGL + 1 sets of global R registers plus a circular stack of N_REG_WINDOWS sets of 16 registers each, known as register windows. The number of register windows present (N_REG_WINDOWS) is implementation dependent, within the range of 3 to 32 (inclusive).
  8. ^ "SPARC Options", Using the GNU Compiler Collection (GCC), GNU, archived from the original on January 9, 2013, retrieved January 8, 2013
  9. ^ SPARC Optimizations With GCC, OSNews, February 23, 2004, archived from the original on May 23, 2013, retrieved January 8, 2013
  10. ^ a b c d e f Weaver, D. L.; Germond, T., eds. (1994). The SPARC Architecture Manual, Version 9. SPARC International, Inc.: Prentice Hall. ISBN 0-13-825001-4. Retrieved May 27, 2023.
  11. ^ "SPARC Behavior and Implementation". Numerical Computation Guide – Sun Studio 10. Sun Microsystems, Inc. 2004. Archived from the original on January 25, 2022. Retrieved September 24, 2011. There are four situations, however, when the hardware will not successfully complete a floating-point instruction: ... The instruction is not implemented by the hardware (such as ... quad-precision instructions on any SPARC FPU).
  12. ^ "Oracle SPARC Architecture 2011" (PDF), Oracle Corporation, May 21, 2014, archived (PDF) from the original on September 24, 2015, retrieved November 25, 2015
  13. ^ Soat, John. "SPARC M7 Innovation". Oracle web site. Oracle Corporation. Archived from the original on September 5, 2015. Retrieved October 13, 2015.
  14. ^ "Software in Silicon Cloud - Oracle". www.oracle.com. Archived from the original on January 21, 2019. Retrieved January 21, 2019.
  15. ^ a b c d e f g h i The SPARC Architecture Manual, Version 8. SPARC International, Inc. 1992. Retrieved May 27, 2023.
  16. ^ a b c d e f g h i "SPARC Fundamental Instructions".
  17. ^ "SPARC64 IXfx Extensions Fujitsu Limited Ver 12, 2 Dec. 2013" (PDF). p. 103-104. Retrieved December 17, 2023.
  18. ^ "Floodgap Retrobits presents the Solbourne Solace: a shrine to the forgotten SPARC". www.floodgap.com. Archived from the original on December 1, 2020. Retrieved January 14, 2020.
  19. ^ Sager, D.; Hinton, G.; Upton, M.; Chappell, T.; Fletcher, T.D.; Samaan, S.; Murray, R. (2001). "A 0.18 μm CMOS IA32 microprocessor with a 4 GHZ integer execution unit". 2001 IEEE International Solid-State Circuits Conference. Digest of Technical Papers. ISSCC (Cat. No.01CH37177). San Francisco, CA, USA: IEEE. pp. 324–325. doi:10.1109/ISSCC.2001.912658. ISBN 978-0-7803-6608-4.
  20. ^ FX1 Key Features & Specifications (PDF), Fujitsu, February 19, 2008, archived (PDF) from the original on January 18, 2012, retrieved December 6, 2011
  21. ^ Tremblay, Marc; Chaudhry, Shailender (February 19, 2008), "A Third-Generation 65nm 16-Core 32-Thread Plus 32-Scout-Thread CMT SPARC(R) Processor" (PDF), OpenSPARC, Sun Microsystems, archived from the original on January 16, 2013, retrieved December 6, 2011
  22. ^ Vance, Ashlee (June 15, 2009), "Sun Is Said to Cancel Big Chip Project", The New York Times, archived from the original on November 4, 2011, retrieved May 23, 2010
  23. ^ "Fujitsu shows off SPARC64 VII", heise online, August 28, 2008, archived from the original on May 23, 2013, retrieved December 6, 2011
  24. ^ Barak, Sylvie (May 14, 2009), "Fujitsu unveils world's fastest CPU", The Inquirer, archived from the original on May 17, 2009, retrieved December 6, 2011{{citation}}: CS1 maint: unfit URL (link)
  25. ^ "Sparc T3 processor" (PDF), Oracle Corporation, archived (PDF) from the original on April 24, 2016, retrieved December 6, 2011
  26. ^ Morgan, Timothy Prickett (December 3, 2010), "Ellison: Sparc T4 due next year", The Register, archived from the original on March 7, 2012, retrieved December 6, 2011
  27. ^ "SPARC Enterprise M-series Servers Architecture" (PDF), Fujitsu, April 2011, archived (PDF) from the original on March 4, 2016, retrieved November 5, 2011
  28. ^ Morgan, Timothy Prickett (August 22, 2011), "Oracle's Sparc T4 chip", The Register, archived from the original on November 30, 2011, retrieved December 6, 2011
  29. ^ Morgan, Timothy Prickett (November 21, 2011), "Fujitsu parades 16-core Sparc64 super stunner", The Register, archived from the original on November 24, 2011, retrieved December 8, 2011
  30. ^ "Fujitsu Launches PRIMEHPC FX10 Supercomputer", Fujitsu, November 7, 2011, archived from the original on January 18, 2012, retrieved February 3, 2012
  31. ^ "Ixfx Download" (PDF). fujitsu.com. Archived (PDF) from the original on May 18, 2015. Retrieved May 17, 2015.
  32. ^ "Images of SPARC64" (PDF). fujitsu.com. Archived (PDF) from the original on April 22, 2016. Retrieved August 29, 2017.
  33. ^ "Oracle Products" (PDF). oracle.com. Archived (PDF) from the original on March 8, 2017. Retrieved August 29, 2017.
  34. ^ "Oracle SPARC products" (PDF). oracle.com. Archived (PDF) from the original on September 26, 2018. Retrieved August 29, 2017.
  35. ^ "Fujitsu Presentation pdf" (PDF). fujitsu.com. Archived (PDF) from the original on April 22, 2016. Retrieved August 29, 2017.
  36. ^ "Fujitsu Global Images" (PDF). fujitsu.com. Archived from the original (PDF) on May 18, 2015. Retrieved August 29, 2017.
  37. ^ "M7: Next Generation SPARC. Hotchips 26" (PDF). swisdev.oracle.com. Archived (PDF) from the original on October 31, 2014. Retrieved August 12, 2014.
  38. ^ "Oracle's SPARC T7 and SPARC M7 Server Architecture" (PDF). oracle.com. Archived (PDF) from the original on November 6, 2015. Retrieved October 10, 2015.
  39. ^ Vinaik, Basant; Puri, Rahoul (August 24, 2015). "Hot Chips – August 23–25, 2015 – Conf. Day1 – Oracle's Sonoma Processor: Advanced low-cost SPARC processor for enterprise workloads" (PDF). hotchips.org. Archived (PDF) from the original on October 9, 2022. Retrieved January 25, 2022.
  40. ^ "Blueprints revealed: Oracle crams Sparc M7 and InfiniBand into cheaper 'Sonoma' chips". theregister.co.uk. Archived from the original on August 29, 2017. Retrieved August 29, 2017.
  41. ^ "Documents at Fujitsu" (PDF). fujitsu.com. Archived (PDF) from the original on August 29, 2017. Retrieved August 29, 2017.
  42. ^ "Oracle's New SPARC Systems Deliver 2-7x Better Performance, Security Capabilities, and Efficiency than Intel-based Systems". oracle.com. Archived from the original on September 18, 2017. Retrieved September 18, 2017.
  43. ^ "SPARC M8 Processor" (PDF). oracle.com. Archived (PDF) from the original on February 28, 2019. Retrieved September 18, 2017.
  44. ^ McLaughlin, John (July 7, 1993), "Intergraph to Port Windows NT to SPARC", The Florida SunFlash, 55 (11), archived from the original on July 23, 2014, retrieved December 6, 2011
  45. ^ Project: Linux for SPARC - oss.oracle.com, October 12, 2015, archived from the original on December 8, 2015, retrieved December 4, 2015
  46. ^ "TOP500 List (1-100)", TOP500, June 2011, archived from the original on June 23, 2011, retrieved December 6, 2011
  47. ^ "The Green500 List", Green500, June 2011, archived from the original on July 3, 2011
  48. ^ "Top500 List – November 2012 | TOP500 Supercomputer Sites", TOP500, November 2012, archived from the original on November 13, 2012, retrieved January 8, 2013
  49. ^ "The Green500 List – November 2012 | The Green500", Green500, November 2012, archived from the original on June 6, 2016, retrieved January 8, 2013
  50. ^ "Tianhe-2 (MilkyWay-2)", TOP500, May 2015, archived from the original on May 26, 2015, retrieved May 27, 2015
  51. ^ Keane, Andy, "Tesla Supercomputing" (mp4), Nvidia, archived from the original on February 25, 2021, retrieved December 6, 2011
  52. ^ Thibodeau, Patrick (November 4, 2010). "U.S. says China building 'entirely indigenous' supercomputer". Computerworld. Archived from the original on October 11, 2012. Retrieved August 28, 2017.
[edit]