New benchmarks Pentium 4 FSB 533 MHz. Intel

Written by Denis Luft. Posted in Processors. 11 views

Foreword

While the fanbase of AMD manufactured processor core with the codename Thoroughbred still waiting the new in 0, 13µm, Intel transition already 4 processors with their 0, 13µm Pentium to the next evolutionary stage. Is the new Pentium 4 models with a front side bus of 133 MHz or effective 533 MHz by the Quadspeed bus operated by and are now available with clock speeds from 2.26 GHz to 2.53 GHz in the trade. This comes not only a good deal closer Intel, the 3 GHz targeted for end of this year but can AMD extended to the competitors further also the lead. After the Pentium 4 2.4 GHz with a front side bus of 400 MHz could hit the Athlon XP 2000 + with a real frequency of 1.66 GHz in almost every area of our extensive benchmarking suite, an Athlon XP 2100 + against Intel’s was allowed to compete this time new high-end processors for the desktop market, the Pentium 4 with 2.4 GHz and 2.53 GHz with a front side bus of 533 MHz. Who won the battle of for the time being fastest processors, to emerge from the following pages.

Technology

As already mentioned, the front side bus has increased with the introduction of the new Pentium 4 models to 533 MHz, on the underlying Pentium 4 processor architecture has changed really not much compared to its predecessor. Also the new Pentium 4 is based on the in 0, 13µm manufactured Northwood-core, which has over a 512 kb large cache and which we ourselves already two times more looked in the past. Therefore want to we at this point not closer on Intel’s NetBurst architecture of the Pentium 4 go down and refer to our earlier article, which as the architecture for what still completely currently is. The following table shows the current processors from Intel and AMD at a glance. The old Pentium 4 with Willamette core, which was made in 0, 18µm, we have waived for reasons of topicality, as Intel of itself already the end for all it-based Pentium 4 processors has launched.

< div >

The processors overview
Features Pentium 4 (FSB400) Pentium 4 (FSB533) Athlon XP
Core Northwood Northwood Palomino
Production 0, 13µm 0, 13µm 0, 18µm
Socket Socket 478 Socket 478 SockelA
Clock rate 1600 MHz
1800 MHz
2000 MHz
2200 MHz
2400 MHz
2260 MHz
2400 MHz
2533 MHz
1333 MHz (1500 +)
1400 MHz (1600 +)
1466 MHz (1700 +)
1533 MHz (1800 +)
1600 MHz (1900 +)
At MHz (2000 +)
1733 MHz (2100 +)
Transistors 55 Million 55 Million 37.5 Million
DIE size 146 mm2 146 mm2 128 mm2
Front side bus 400 MHz QDR 533 MHz QDR 266 MHz DDR
L1-execution cache 12,000 µ OPS 12,000 µ OPS 64 KB
L1 data cache 8 KB 8 KB 64 KB
L1-stroke CPU clock CPU clock CPU clock
L2-cache 512KB 512 KB 256KB
L2 range 256-Bit 256-Bit 64-Bit
L2-cache clock CPU clock CPU clock CPU clock
Hardware data prefetching Yes Yes Yes
VCore 1.5 Volt 1.5 Volt 1.75 Volts
Instruction sets MMX
SSE / SSE2
MMX
SSE / SSE2
MMX / 3DNow!
3DNow! + / SSE
Temperature diode Yes Yes Yes
Multiprocessor capable No No No
CPU architecture 20-stage pipeline 20-stage pipeline 15-stage (FPU)
10-stage (ALU)
Pipeline

As we see the differences are not really serious. As mentioned above, Intel deploys the Northwood core at the new Pentium 4 based which was designed especially for high clock speeds on Intel NetBrust architecture and can therefore only have full performance at the high speed of the processor. In this context Intel had to criticism can be often fallen, that you wanted to drive only the clock frequency 4 introduction of the Pentium upward, without thereby breaking new speed records. At one point, the critics are here certainly right: A running at 1 GHz Pentium 4 would be much slower than a 1 GHz Pentium III or AMD Athlon. At a frequency of 2.4 GHz, which reached the Pentium 4 with Northwood core without greater thermal problems, is however a completely different result. With the introduction of 2.4 GHz Pentium 4 on 2 April increased casual around 200 MHz clock of the processor Intel and can set now the 2.53 GHz on it Pentium 4 again 133 MHz. In view of these major steps, it should be clear that Intel has left the largest competitor, the AMD Athlon XP, now significantly behind. Because AMD could only 66 MHz compared to the Athlon XP 2000 + with the introduction of Athlon XP 2100 + 1733 MHz a real processor clock on it, the two fast processors separate currently not less than 800 MHz from Intel and AMD. From Intel’s perspective, the idea of the Netbrust architecture was a resounding success – at the 20.11.2000 so or has now become one. This now several times mentioned architecture can this be outlined with six key words we have already extensively lit in a previous article:

  • Hyper pipelined technology
  • Rapid execution engine
  • Advanced dynamic execution
  • Execution trace cache
  • Quadspeed system bus
  • Streaming SIMD Extensions 2 (SSE2)

Who with the 533 MHz Pentium 4 revolutionary innovations expected should thus be something disappointed. So far nothing has changed in the Prozessorstepping. Currently, there are only the nB0 stepping, which is equipped with a total of 23 known bugs of the Northwood core.

But yet a visible change to the classic Northwood in the eye is by coincidence we like. So, the SMD resistors on the bottom of the Pentium 4 are now fixed with another material. While at the 400 MHz Pentium 4 2.4 GHz here nor a silver material is used, are all 533 MHz Pentium 4 models equipped with a golden material.

As Intel wrote us these are however a common change in the context of the usual development. However, no tangible improvements arising from this modification according to Intel.

All in all is the only real also notable change in the new processors thus increasing the front-side bus 533 MHz. But why should the performance of the entire system increase just this? Already in the past has shown that the Pentium 4 benefits greatly from high bandwidth of the system bus. There is a theoretical bandwidth of 3.2 GB per second at a front side bus of 400 MHz. The Pentium 4 can thus up 3.2 GHz with the memory controller, communicate Controller Hub (MCH) memory, which among other things, the AGP bus and the memory bus are addressed. In a front-side bus of 533 MHz as the new Pentium 4 with it brings him, increases the theoretical bandwidth to over 30 percent and thus is whole 4.2 GB per second.

Bandwidth

Because we have dealt in the past rather less with the bandwidth of the bus systems, I should like briefly at this point to some basics and why they are so important. As a processor can process more data with increasing clock speed, it is important that it can be supplied with sufficient data per clock or per second. What good is the fastest CPU, if he is not sufficiently quickly supplied with data? As well there of an immensely high memory bandwidth nothing, if the processor is not able that is just so bad incoming data to process or the connection of the processor, that the data can not get to the processor.

The Pentium 4 is a 64-bit wide bus with the Northbridge in Intel the memory controller hub (MCH), connected. Intel is able to transfer four data information to this bus per cycle as a whole. Can the bandwidth of the bus be here for the Pentium as follows 4 processors:

  • 400 MHz: 64-bit bus width * 100 MHz bus clock * 4 packets = 25600 MBit per second
  • 533: MHz 64 bit bus width * 133 MHz bus clock * 4 packets 34048 MBit per second =

The result by 8 divided to get bytes per second is one like previously mentioned the system bus 3.2 and 4.2 GB per second of bandwidth. The new Pentium 4 can send so 4.2 GB per second or received. Thus Controller Hub (MCH) depends it largely from memory, system performance is ultimately how high in the Pentium 4. Here, the Northbridge run together at the MCH and the data from the memory of the AGP bus and all other system components. This certainly, the most important factor is the memory because “filed” here open applications. Basically, it can be said that the memory has significant influence on overall system performance. The graphics card as well as the remaining components play only a minor role in particular in modern systems.

The introduction of accelerated graphics port (AGP) was originally intended to improve the connectivity of the graphics card memory, so that these textures can outsource “faster”. For this reason, it has doubled the clock frequency of AGP over the PCI 33 MHz on 66 MHz. Because it is still a 32-bit wide bus you increased 1 so the bandwidth of 133 MB / s at the PCI to 266 MB / s for AGP. Now can you by small tricks for the AGP four packets per stroke send or receive so the bandwidth with AGP 4 x rose to little more than 1 GB per second. But already at the stage of AGP 2 x on AGP 4 x was observed only small performance gains despite doubling theoretical bandwidth. But why?

One reason for this is certainly the fact that modern video cards more must take no textures due to their large graphics card store at all in memory. Because the processor also no longer must be used for transformation or lighting calculations (keyword T & L), less data between the processor and video card must also be replaced. Thus only a fraction remains of the theoretical bus load of 1 GB / s.

Thus, it is clear that the connection of memory is particularly important. Continue to the read data for the time being land drives of disks or CD-ROM in memory, the memory is the instance that has a major impact on system performance. To take its full performance Pentium 4 with a bus connection 3.2 and 4.2 GB / s, is needed so fast memory. For this reason, Intel has decided with the introduction of the Pentium 4 for Rambus as storage technology. He alone was to market able to deliver the bandwidth for the Pentium 4 with 400 MHz FSB. Rambus with a frequency of 400 MHz (PC800) can reach a bandwidth of 3.2 GB / s. To also the new 533 MHz Pentium 4 with enough data to provide, there called PC1066 Rambus which can theoretically reach a 4.2 GB / s bandwidth. However, new Rambus chipset supports Intel’s officially only the slower Rambus, the faster version runs but nevertheless totally stable on such boards, so that manufacturers such as ASUS court their based products even with such support Rambus.

However, Rambus is still not the popularity, that is wanted Intel, even though the memory of the price is her no longer more expensive than good DDR SDRAM. Since Rambus for Intel still is the first choice in the Highperfomance area, also the new Pentium 4 with a front side bus of 533 MHz starts for the time being with a new Rambus chipset (i850-E). Intel will present but a DDR SDRAM chip set (i845-E) in the near future, which is also suitable for the new Pentium 4 with 533 MHz. However, this is to develop the full power of the Pentium 4. The i845-E only memory will support after PC200 and PC266 standard. The bandwidth of PC266 is only 2.1 GB per second, without wanting to open large invoices at this point. If it matters, simply too little for the old Pentium 4 and especially for the new Pentium 4, because the bandwidth increase in this case can not be exploited. Also PC333 memory that is supported by some Pentium 4 chipsets such as the SiS 645DX, or VIA P4X333, is here with its 2.7 GB per second still not fast enough. The performance difference is ultimately really how much in practice, we will still show a later review.

The AMD Athlon XP, we have also a 64-bit wide bus. This is clocked at 133 MHz and works with the double data rate procedure. In this way, the Athlon XP with the Northbridge with a range of 2.1 GB per second can communicate. Thus can DDR SDRAM PC266 standard with its 2.1 GB / s the Athlon so with enough data supply. The use of faster memory such as, for example, PC333 has only a small impact on the overall performance of an Athlon system. A small increase in performance can be seen but still, because normally, not only the processor accesses the memory. This information only to the edge.

Power consumption

Before we chase our test candidates by the benchmark course, we can first compete them in usage. Should be clear from the outset that the new Northwood here should have an advantage due to its smaller manufacturing technology and lower core voltage. But with every increase in clock increases significantly also the consumption of each processor. Accordingly also the cooler must be more powerful, to also pay the resulting heat output. The specified by the manufacturers of processor thermal design power is here the typical thermal power, a cooler in normal operation shall be at least can drain. The following table shows the performance data specified for their processors from Intel and AMD.

< div >

Performance data in comparison
Processor Core voltage Consumption
Athlon XP processors
Athlon XP 2100 + (1.73 GHz) 1.75 Volts 64.3 Watt
Athlon XP 2000 + (1.66 GHz) 1.75 Volts 62.5 Watts
Athlon XP 1900 + (1.60 GHz) 1.75 Volts 60.7 Watt
Athlon XP 1800 + (1.53 GHz) 1.75 Volts 59.2 Watts
Athlon XP 1700 + (1.46 GHz) 1.75 Volts 57.4 Watt
Pentium 4 processors
Pentium 4 2.53 GHz (Northwood) 1.5 Volts 59.3 Watt
Pentium 4 2.40 GHz (Northwood) 1.5 Volts 57.8 Watt
Pentium 4 2.20 GHz (Northwood) 1.5 Volts 55.1 Watt
Pentium 4 2.0A (Nothwood) 1.5 Volts 52.4 Watt
Pentium 4 2.0 (Willamette) 1.75 Volts 75.3 Watt

The table shows clearly that the consumption of the Pentium 4 with Northwood core below which an Athlons. In addition, you can see the savings in consumption by moving to the 0 13µm manufacturing technology also very beautiful. At same speed, the Pentium 4 with 2.0 GHz and Northwood-core needs whole 23 Watts less than its predecessor with Willamette core and 0 18µm technology. The new Pentium 4 with its 2.53 GHz needs still significantly less than the models of competition and is approximately with the Athlon XP 1800 + at a level.

In keeping with the power consumption is naturally also the radiated heat. Under full load the temperature of the Pentium 4 2.53 GHz at open housing was 40 ° C, about which he warmer than the 2.4 GHz Pentium well 2 ° C was a 133 MHz front side bus. The temperature of Athlon XP 2100 + was read out with really powerful and loud fan with 47 ° C, which makes the AMD processor still significantly warmer than its Intel counterpart.

As we can see from the table, Intel has left the core voltage unchanged in the new processors. But the tension is only then, if the processor is not charged, or better put, if it flows a stream of 0 amps. Also, the processor voltage is changed when increasing load of the processor. All Pentium 4 models have this behavior, however we have not made so far on this feature. The 1.5 volt form only the tension in the current referred to. When a current of 10 amperes flowing through the processor, the processor power to at least 19 mV and maximum 75mV should be reduced. The current and the voltage are this is linearly dependent. The Pentium 4 2.4 GHz is crossed according to Intel at full load of 49.8 amps Max. Under this load condition, the voltage shall be not less than 1,330 volts and not exceeding 1,405 volts. In this case it is core voltage at anliegende the processor voltage noticeably below the usually specified from 1.5 volts. Since the Pentium 4 2.53 GHz can pull up to 51.5 amps, the processor voltage in this case from the motherboard must automatically be reduced to a value between 1,325 and Safari 1,400 V. In this way, the heat radiation of the Pentium 4 is throttled.

Conditions

While the on presented 2 April 2002 2.4 GHz Pentium 4 with its 400 MHz front-side bus without problems in all the time of available motherboards could be used, a new chipset needed is actually for the new evolution with 533 MHz. Together with the launch of the new Pentium, Intel as mentioned 4-processor sends the new i850-E chipset with Rambus support in the race. This chipset, we have devoted an own article in which we also specifically address the new features of the chipset. Also, Intel will soon present a slightly revised i845 chipset which will bear the name i845-E and is also designed for the new Pentium 4. In addition to Intel, even SiS with the SiS645DX has a FSB533 capable chip set in the program. Also VIA an appropriate chipset has introduced with the P4X333. Since however still the disputes between VIA and Intel regarding the licensing of the Pentium 4 bus not resolved are, for the time being only VIA corresponding with his own motherboard brand VPSD products in the program will have. Here a small overview of the currently available and appropriate for the new Pentium 4 chipsets.

< div >

The chipsets compared
Features Intel 850 E VIA P4X333 SiS 645DX
Supported CPU sockets Socket 478 Socket 478 Socket 478
Northbridge features
Northbridge KC82850E VT8754 645DX
Multiprocessor support No No No
Front side bus
400 MHz Yes Yes Yes
533 MHz Yes Yes Yes
Memory clock
100 MHz Yes (400) Yes Yes
133 MHz Yes (533) Yes Yes
166 MHz No Yes Yes
Asynchronous memory clock Yes Yes Yes
Memory slots (max) 4 3 3
Storage support
SDRAM No No Yes
DDR-SDRAM No Yes Yes
(Dual)-Rambus Yes No No
AGP support
1 x No No No
2 x Yes Yes Yes
4 x Yes Yes Yes
Southbridge features
Southbridge Intel 82801 BA VT8235 SiS 961B
Disk controller
ATA 33 / 66 Yes Yes Yes
ATA 100 Yes Yes Yes
ATA 133 No Yes Yes
Number of PCI-slots (max) 6 6 6
USB ports (max) 4 4 6
USB 2.0 No Yes No
Other
I/O-link Hub interface
(266MB / s)
V link
(533MB / s)
MuTIOL
(533MB / s)

However, owners of older Pentium should not despair 4 motherboard if they find their chipset in this collection. So are many previous chips in a position to operate the new Pentium 4 with a front side bus of 533 MHz outside its own specifications. For example, good mainboards with i845, i850, or SiS645 are such as those of us in the past tested ASUS mainboards P4B266-E or P4S333 able. On the safe side, it is however only if one can call his own a Mainboard with a chipset in the table. For example, the newly introduced by Asus and tested by us P4T533-C among the representatives of appropriate under warranty.

Overclocking

On our last processor reviews, we have treated the overclocking while often in principle, have implemented but not our thoughts. Because this time for our processor test that was used for all benchmarks, in addition to the Intel motherboard, even the previously mentioned P4T533-C motherboard from ASUS with its numerous overclocking features available to us, we could operate themselves. Even we were surprised. We could totally stable run Pentium 4 2.53 GHz here using air cooling (Intel boxed fan) and without VCore increase to 2.85 GHz (19 x 150 MHz). In this case under full load, the processor temperature was around 45 ° C, a perfectly acceptable value. The memory was running in this case with 450 MHz. No problem for our PC1066 module that by default for 533 MHz are designed. Our 32ns Samsungmodule not verkrafteten the attempt to operate our Rambus with synchronous front-side bus 600 MHz, unfortunately.

Even if the motherboard has allowed us not setting a higher bar in the BIOS, this test shows but impressively, which performance potential in the Pentium 4. Considering this fact it should be easy for Intel, to continuously increase the processor clock on a regular basis at 3 GHz.

Test system

So now we come to the most interesting part of this article. In order to assess the performance of processors came a slightly modified test system to support the new Pentium 4 processors used. Motherboard Intel i850-E reference named D850-EMV2 served as us, that as a special extra through a USB 2.0 controller from NEC has. However taktete the Board somewhat conservative, the processors so that the 2.53 GHz Pentium was ultimately only operated with around 2.52 GHz. The comparison test of Intel D850 EMV2 and the ASUS P4T533-C published also at the start of the new Pentium 4 shows what performance brings the choice of a different motherboard. As a network card, an Intel EtherExpress served 100 Pro. Our proven GeForce3 Ti500 from Inno3D was responsible for the screen output. Tested both with the officially support Rambus, PC800, as well as with PC1066 memory that should have the full power of the processor.

For the fastest Athlon XP performance evaluation, we initially wanted to use the MSI KT3Ultra-ARU. Since its performance with our DDR333 however not convinced (Cl2, 5) memory from crucial, our KT266A reference the EPoX 8KHA + for this processor test being used had to without further ADO. The good piece heard in spite of already available KT333 yet long not on the scrap. DDR266 modules (CL2) from Infineon were used as storage for the Board of EPoX. Hereby, we could take the sharpest settings.

The usual us Windows XP Professional was used as the operating system. Functions such as system restore and automated system updates have been disabled. In order to create a normal working environment, we have worked as a quality settings for the system interface with smoothed fonts, and the default Windows XP user interface. All other Visual options were also enabled. As a graphics card driver, they served last official nVidia detonator version 28.32. To prevent issues now follows a complete comparison of the hardware used and the installed driver.

  • Processor
    • Intel Pentium 4 2.53 GHz (FSB533)
    • Intel Pentium 4 2.40 GHz (FSB533)
    • Intel Pentium 4 2.40 GHz (FSB400)
    • Intel Pentium 4 2.20 GHz (FSB400) (on the old platform)
    • Intel Pentium 4 2, 0a GHz (FSB400) (on the old platform)
    • AMD Athlon XP 2100 +
    • AMD Athlon XP 2000 + (on the old platform)
    • AMD Athlon XP 1900 + (on the old platform)
  • Motherboard
    • AMD platform (new): EPox 8KHA + (KT266A)
    • AMD platform (old *): ASUS A7V266-E (KT266A)
    • Intel platform (new): Intel D850EMV2 (i850-E)
    • Intel platform (old *): Intel D850MD (i850)
  • Memory
    • 256 MB PC266 CL2 DDR SDRAM
    • 2x128MB PC800 Samsung Rambus
    • 2x128MB PC1066 Samsung Rambus
  • Graphics card
    • Inno3D tornado Geforce3 Ti500
  • Peripherals
    • ASUS CRW 4012A
    • IBM IC35LC040
    • Intel EtherExpress 100 Pro
  • Driver versions
    • nVidia detonator 28.32
    • AMD platform: via 4in1 4.38
    • Intel platform: Intel inf driver 3.20.1008 + Intel Application Accelerator 2.2
  • Software
    • Windows XP Professional

Benchmarks

For assessing the processing power a slightly stripped-down test scenario was this time used. So the now rather old 3DMark exiled 2000 test runs from our, at least for the time being. Also SuperPI was allowed to pack up his things, since he us too unrealistic appears. Who calculates make at home from fun PI to more than one million. In addition, he confirmed only the test results of Sandra 2002. Also the MP3 Maker Platinum by MAGIX has unfortunately no longer managed in this test. In later tests he will be with but again rested and quiet things of the game. Still we have replaced 2001 2002 the SYSmark by his successor, the SYSmark, to assess the performance in the Office life with current applications. In addition to the disposal, we can welcome but also a new addition. It is the PCMark 2002, which is designed to support Sandra 2002. Thus our used benchmarks should cover all areas relevant to the user.

  • Game performance
    • 3DMark 2001 SE
    • GLMark
    • Quake 3 arena
  • Office performance
    • SYSmark 2002
    • Winace 2.11
    • FlaskMPEG (DiVX 5.01)
    • Lame
  • 3D of rendering performance
    • Cinema 4 d
    • Lightwave 7.0 (b)
    • Spec ViewPerf 6.1.2
  • Other
    • Sandra 2002
    • PCMark 2002
    • SETI@home version

Who would like to understand the benchmarks in their home even once, will find much of the above listed test programs with us in the download section.

Boottime

The program developed by Microsoft Bootvis was used to measure the speed of the boat. This program has actually been developed to optimize the boot process and analyzes startup for this purpose in every detail. Among other things in / output examines the boat activity, CPU utilization, disk the driver delay and much more. As our experience showed, the boot times are different between the individual processors of a platform only slightly. For this reason we satisfied this time with two Deputy measurements. So Bootvis identified in the 2.4 GHz Pentium 4 (533 MHz) one boot time of approximately 10.4 seconds. Our AMD Athlon XP 2100 + test system needed for the startup procedure 12.2 seconds and was thus easily delayed prepared to work. Differ from each other significantly more than the two test platforms and in particular the mainboard at boot is important, you should transfer not system the result on any Pentium 4 or Athlon XP.

Sandra 2002

Before we compete against the team of processors in real-world applications, we want to look once more the theoretical values of the car at this point. To this end we have used Sandra, since both the 3DNow!, extension of processors correctly recognize SSE1 as well as the SSE2 and accordingly also exploited.

Sandra processor test
Sandra 2002 provides equal to two benchmarks, which are to determine only the performance of the processor. This is the Dhrystone benchmark on the one used, which was originally developed by Siemens, to measure the performance of the main processor. On the other hand, the performance of the co processor is determined via the whetstone benchmark. Both tests be carried out without considering the extended such instruction sets. In addition to still the SSE2 command extension was the whetstone test used.

Sandra 2002 – Dhrystone

P4 2.53 / 533 – ENERGIESPAR
4.895
XP 2100 + – PC266
4,800
XP 2000 + – PC266 *.
4.642
P4 2.40 / 533 – ENERGIESPAR
4.549
P4 2.40 / 400 – ENERGIESPAR
4.513
XP 1900 + – PC266 *.
4.456
P4 2.20 / 400 – ENERGIESPAR *.
4.211
P4 2,0A / 400 – ENERGIESPAR *.
3.723
Data in MIPS

Since the memory bandwidth in this test has no outstanding importance, we have renounced the publication of our test results with the fast PC1066 memory. In the context of the measurement uncertainty, it makes no difference whether PC800 or PC1066 Rambus is used. As we see the Athlon XP makes a really good figure 2100 + as well as his predecessor in this purely theoretical test. Not even 100 points separate the processor in this test the 800 MHz higher clocked opponent.

Sandra 2002 – whetstone

P4 2.53 / 533 – ENERGIESPAR SSE2
3,076
P4 2.40 / 533 – ENERGIESPAR SSE2
2,916
P4 2.40 / 400 – ENERGIESPAR SSE2
2.859
P4 2.20 / 400 – ENERGIESPAR SSE2 *.
2.706
P4 2,0A / 400 – ENERGIESPAR SSE2 *.
2.457
XP 2100 + – PC266
2.403
XP 2000 + – PC266 *.
2.320
XP 1900 + – PC266 *.
2.228
P4 2.53 / 533 – ENERGIESPAR
1.316
P4 2.40 / 400 – ENERGIESPAR
1,250
P4 2.40 / 533 – ENERGIESPAR
1,247
P4 2.20 / 400 – ENERGIESPAR *.
1,143
P4 2,0A / 400 – ENERGIESPAR *.
1.039
Information in MFLOPS

In the Sandra processor test whetstone is once again an ambivalent result. When the Pentium 4 processors with its SSE2 extension, they are for the time being no AMD processor stop. Once however forced the benchmark for the use of SSE1, the performance of the Pentium can not really convince 4 processors. This benchmark is only then for the Pentium 4 architecture, if the SSE2 instruction set is used.

Sandra multimedia test
The Sandra 2002 multimedia test used an algorithm, which is used also when generating realistic natural objects such as mountains and clouds. There is talk here of chaos theory, which was drawn up by Mandelbrot. This benchmark consider the extended command sets of the Pentium 4, or that the Athlon XP. Since the implementation of SSE1 in this partial testing better than that of 3DNow! is, we have the Athlon XP with its SSE1 unit (i.e. 3DNow! professional) work for you. The Pentium 4 was, however, both SSE1, including SSE2 to use.

Sandra 2002 – integer

SSE1 and SSE2 optimization:
P4 2.53 / 533 – ENERGIESPAR SSE2
10.022
XP 2100 + – PC266
9.530
P4 2.40 / 533 – ENERGIESPAR SSE2
9.491
XP 2000 + – PC266 *.
9.207
P4 2.40 / 400 – ENERGIESPAR SSE2
9.059
XP 1900 + – PC266 *.
8.839
P4 2.20 / 400 – ENERGIESPAR SSE2 *.
8.692
P4 2.53 / 533 – ENERGIESPAR
8.518
P4 2.40 / 400 – ENERGIESPAR
8.087
P4 2.40 / 533 – ENERGIESPAR
8.067
P4 2,0A / 400 – ENERGIESPAR SSE2 *.
7.913
P4 2.20 / 400 – ENERGIESPAR *.
7.396
P4 2,0A / 400 – ENERGIESPAR *.
6.721
Information in instructions per second (it/s)

At the multimedia benchmark Sandra 2002 the Pentium 4 2.53 GHz can call while his own the first place, the distance to the Athon XP 2100 + is not particularly large. The full performance can be achieved only with enabled SSE2 command extension. Particularly interesting about this test should be the fact, that the new 2.4 GHz Pentium 4 by system bus bandwidth increased to 4.2 GB / s alone reached 400 points more as equal switched mode counterparts with a front side bus of 400 MHz. The performance difference is only 4.8 percent. Opposite to the 2.4 GHz Pentium 4 to 5.4% higher clocked 2.53 GHz Pentium 4 can implement his “more” to stroke in 5.6 percent more performance by the way.

Sandra 2002 – floating point

SSE1 and SSE2 optimization:
P4 2.53 / 533 – ENERGIESPAR
12.422
P4 2.53 / 533 – ENERGIESPAR SSE2
12.209
P4 2.40 / 400 – ENERGIESPAR
11.789
P4 2.40 / 533 – ENERGIESPAR
11.759
P4 2.40 / 400 – ENERGIESPAR SSE2
11.594
P4 2.40 / 533 – ENERGIESPAR SSE2
11.564
XP 2100 + – PC266
11.085
P4 2.20 / 400 – ENERGIESPAR *.
10.784
XP 2000 + – PC266 *.
10.666
P4 2.20 / 400 – ENERGIESPAR SSE2 *.
10.597
XP 1900 + – PC266 *.
10.238
P4 2,0A / 400 – ENERGIESPAR *.
9.807
P4 2,0A / 400 – ENERGIESPAR SSE2 *.
9.625
Information in instructions per second (it/s)

The multimedia test with floating point numbers, the Pentium 4 looks actually quite well. However, one must admit envy that the Athlon XP 2100 + can exceptionally well pass this test. He follows an even slightly ahead of the 2.2 GHz Pentium.

Sandra memory test
The memory test by Sandra occupies at least 50 percent of available memory. This benchmark to determine the performance of the storage subsystem, as well as the cache. For this, both arithmetic and floating point operations are carried out. As this test very much depends of the platforms, we want to take only a short look at.

Sandra 2002 – RAM

Integer data:
P4 2.53 / 533 – PC1066
3.312
P4 2.40 / 533 – PC1066
3.304
P4 2.53 / 533 – ENERGIESPAR
2,802
P4 2.40 / 533 – ENERGIESPAR
2.764
P4 2.40 / 400 – ENERGIESPAR
2.491
XP 2100 + – PC266
1.980
Floating point:
P4 2.53 / 533 – PC1066
3.312
P4 2.40 / 533 – PC1066
3.303
P4 2.53 / 533 – ENERGIESPAR
2.796
P4 2.40 / 533 – ENERGIESPAR
2.764
P4 2.40 / 400 – ENERGIESPAR
2.489
XP 2100 + – PC266
1,900
Information in megabytes per second (MB/s)

The Sandra 2002 memory test to the Pentium 4 models operated with PC1066 Rambus could decide clearly for themselves. The memory could send at least 3.3 GB/s through the lines of the theoretically possible 4.2 GB/s. The Athlon XP with his PC266 memory is really behind this test. In particular this result is responsible in the one or other place for the poor performance of the Athlon XP compared to the Pentium 4. However you should be wary, to push the Grouch in the shoes of the memory performance for all the bad results.

PCMark 2002

The PCMark 2002 is a comparatively new benchmark, as the name says, designed to measure the overall system performance. For this, he performs a series of tests, which primarily call for the processor and memory. Also the hard disk is checked for their performance there. However, we have given to the publication of the subtotal. Thus, PCMark provides us with 2002 just a score for the performance of the processor and the memory.

PCMark 2002

CPU mark:
P4 2.53 / 533 – PC1066
6.146
P4 2.53 / 533 – ENERGIESPAR
6.118
P4 2.40 / 533 – PC1066
5.892
P4 2.40 / 533 – ENERGIESPAR
5.845
P4 2.40 / 400 – ENERGIESPAR
5.786
XP 2100 + – PC266
5.164
RAM mark:
P4 2.53 / 533 – PC1066
6.599
P4 2.40 / 533 – PC1066
6.427
P4 2.53 / 533 – ENERGIESPAR
5.806
P4 2.40 / 533 – ENERGIESPAR
5.579
P4 2.40 / 400 – ENERGIESPAR
5.315
XP 2100 + – PC266
3.278
Information on points

On to PCMark should serve 2002 to verify the results obtained in Sandra 2002. At least the memory performance assessment confirmed the values from Sandra fully assessing processor Athlon XP truncates much better 2100 + considerably worse, and the Pentium 4. The performance gain by the higher front side bus is here by the way with 60 points (110 points in PC1066 memory) or just 1 percent (2%) in the weight. The 2.53 GHz Pentium 4 can unleash 4.7 percent more performance, however, from his 5.4% higher clock.

3DMark 2001SE

The importance of 3DMark 2001 has grown continuously in recent months. Since the predecessor served has long been only as, we have decided in the context of this test for this, to use only a synthetic DirectX benchmark. At the 3DMark 2001 the combination of graphics card and processor is especially important so that clear statements about the performance of the processor can be active. By the way, the benchmark ran into the default run with the driver settings for best graphics in the mipmap level of detail.

3DMark 2001SE

P4 2.53 / 533 – PC1066
9.221
P4 2.40 / 533 – PC1066
9.125
P4 2.53 / 533 – ENERGIESPAR
9.034
P4 2.40 / 533 – ENERGIESPAR
8.893
P4 2.40 / 400 – ENERGIESPAR
8.703
P4 2.20 / 400 – ENERGIESPAR *.
8.407
XP 2100 + – PC266
8.393
XP 2000 + – PC266 *.
8.320
XP 1900 + – PC266 *.
8.245
P4 2,0A / 400 – ENERGIESPAR *.
8.160
Information on points

Already in our last review of the processor the speed Crown at the 3DMark went in 2001 on the 2.4 GHz Pentium 4 over and hand remains as expected in Intel’s. Particularly interesting in this context seems, that the 2, 4 / 533 GHz Pentium 4 by using faster Rambus can draw on the 2.53 GHz Pentium 4 past. The higher Speichbandbreite can be implemented here in about 2 percent more power. Not just the world, but all the same. The Athlon XP 2001 + be found “only” in the midfield.

GLMark

The GLMark has been developed by the German software company Vulpine and is a purely synthetic benchmark for OpenGL as well as 3DMark2000 and 3DMark2001 for DirectX. The benchmark is highly complex and can take advantage of some of the new GeForce3 features. Also here, you can achieve a performance by a new processor. The benchmark this time expired with disabled VSync, whereas we activated this option in past reviews, because one usually works as a normal user with VSync. However, since the performance differences between the respective processors in this case disappears low, now each processor was able to show what’s really in it.

GLMark

640 x 480 x 32:
P4 2.53 / 533 – PC1066
129,3
P4 2.40 / 533 – PC1066
127,8
P4 2.53 / 533 – ENERGIESPAR
125.8
P4 2.40 / 533 – ENERGIESPAR
124,6
P4 2.40 / 400 – ENERGIESPAR
121,8
XP 2100 + – PC266
115,9
1024 x 768 x 32:
P4 2.53 / 533 – PC1066
85,9
P4 2.40 / 533 – PC1066
85,4
P4 2.53 / 533 – ENERGIESPAR
84,7
P4 2.40 / 533 – ENERGIESPAR
84.2
P4 2.40 / 400 – ENERGIESPAR
83.7
XP 2100 + – PC266
82.8
Information in frames per second (FPS)

Also in the Vulpine GLMark, Intel, can expand the lead with the idea of new processors. The increase in the average framerate to 3 frames per second only by the higher front side bus is particularly striking in this context. However, there is the result particularly in 1024 x 768 quite closely to each other, because here is the video card to the limit.

Quake 3 arena

Quake3 uses only OpenGL arena. First person shooter is characterized by a large number of polygons and complex scenarios. Because many games on the Quake III engine based, the tests are very practical with Quake III. We came in addition to the in Quake 1.17 already contained demo 001 still the NV15 demo by nVidia used that affects the processor to the utmost. It was tested with the normal config.

Quake 3 arena – demo 001

640 x 480:
P4 2.53 / 533 – PC1066
351,3
P4 2.40 / 533 – PC1066
341,8
P4 2.53 / 533 – ENERGIESPAR
326,5
P4 2.40 / 533 – ENERGIESPAR
319,2
P4 2.40 / 400 – ENERGIESPAR
302,4
XP 2100 + – PC266
244,8
1024 x 768:
P4 2.53 / 533 – PC1066
279,0
P4 2.40 / 533 – PC1066
277,0
P4 2.53 / 533 – ENERGIESPAR
272,4
P4 2.40 / 533 – ENERGIESPAR
269,8
P4 2.40 / 400 – ENERGIESPAR
262,6
XP 2100 + – PC266
227,6
Information in frames per second (FPS)

Quake 3 arena – demo NV15

640 x 480:
P4 2.53 / 533 – PC1066
75,0
P4 2.40 / 533 – PC1066
72.6
P4 2.53 / 533 – ENERGIESPAR
70,2
P4 2.40 / 533 – ENERGIESPAR
68,5
P4 2.40 / 400 – ENERGIESPAR
64,8
XP 2100 + – PC266
57.2
1024 x 768:
P4 2.53 / 533 – PC1066
74,7
P4 2.40 / 533 – PC1066
72,5
P4 2.53 / 533 – ENERGIESPAR
70,0
P4 2.40 / 533 – ENERGIESPAR
68,2
P4 2.40 / 400 – ENERGIESPAR
64.6
XP 2100 + – PC266
56.9
Information in frames per second (FPS)

Quake 3 arena as a representative of a large team at computer games once again impressively shows what services through the combination of fast RAM and processor are possible. From his clock increase can the 2.53 GHz Pentium 4 only 2.3 percent more performance out hit. When comparing the 2.4 GHz Pentium 4 processors but reveals which that Quake one of the games is that benefit most from the higher front side bus. Good 15 frames or 5.6 per cent are in it solely by this small modification. Performance increases also PC1066 Rambus is used again to 7.6 percent. However given that the measure of PC1066 compared to PC800 Rambus is around 31 percent higher, no overwhelming result.

SYSmark 2002

The SYSmark 2001 has often caused excitement on the part of the AMD fan base in the past. For this reason we decided in 2002 in this processor test for the successor of SYSmark. But also the new version of the benchmark is divided up into two areas. The first area simulates the work environment of a “Webmaster” or “Web Designer”, bearing the name Internet content creation. The following real-world applications with a scripting language are controlled in this test section: Macromedia Dreamweaver 5, Adobe Photoshop 6.0.1, Adobe Premiere 6.0, Microsoft Windows Media Encoder 7.1 and Macromedia Flash 5.0. The version of Windows Media Encoder 7.0 was used, which is however not correctly recognized the SSE1 command extension of Athlon XP because he checked processors on Intel only whether the processor signals a support of these commands in SYSmark 2001. However, because the new version of Microsoft with the error is equipped, the SYSmark 2002 now forms a comparatively good valuation basis for the system performance, which is significantly influenced by the processor.

In the second area of SYSmark 2002 the Office life (Office content creation) is measured with a whole latte applications, raised in parallel to each other via multitasking. These applications include Microsoft Office 2002, Dragon Naturally speaking, Netscape Communicator 6.0, WinZip 8.0 and McAfee VirusScan 5.13. The respective partial results arising from the Internet content creation and Office content creations, enter each to 50 percent in the final result. But let us look at the individual first once in detail:

SYSmark 2002 – Internet content

P4 2.53 / 533 – PC1066
353
P4 2.53 / 533 – ENERGIESPAR
335
P4 2.40 / 533 – PC1066
331
P4 2.40 / 533 – ENERGIESPAR
325
P4 2.40 / 400 – ENERGIESPAR
320
XP 2100 + – PC266
214
Information on points

Already the first part of SYSmark 2002 shows that AMD has lost the tuning competition. Even with the new system benchmark, the Athlon XP 2100 + is just behind. The poor performance in the partial test also affects the overall result. Clear winner in this area is Intel’s new 2.53 GHz Pentium using PC1066 memory.

SYSmark 2002 – Office content

P4 2.53 / 533 – PC1066
188
P4 2.40 / 533 – PC1066
185
P4 2.53 / 533 – ENERGIESPAR
181
P4 2.40 / 533 – ENERGIESPAR
179
P4 2.40 / 400 – ENERGIESPAR
174
XP 2100 + – PC266
145
Information on points

The results in the second part of test in favour of the Pentium 4 processors. The Athlon XP is clearly behind, however differences in clock should think in this case also of sizes between the two processors. An improved performance by 5 points could earn Intel by the way only by the higher front side bus.

SYSmark 2002 – overall

P4 2.53 / 533 – PC1066
254
P4 2.40 / 533 – PC1066
247
P4 2.53 / 533 – ENERGIESPAR
245
P4 2.40 / 533 – ENERGIESPAR
241
P4 2.40 / 400 – ENERGIESPAR
236
XP 2100 + – PC266
176
Information on points

One is one together one and one it becomes clear that the Pentium 4 is for the time being better suitable for the Office life. The difference in speed to the Athlon XP here really strongly catches the eye.

SETI@home version

We are especially proud in our comparison test course on the inclusion of Textclienten of SETI@home version (3.03). Due to the long term, a clear picture of the performance of individual processors should arise here. To keep the results comparable, always the same work unit came with one angle range of 0,417 used.

SETI@home version

P4 2.53 / 533 – PC1066
10.072
P4 2.40 / 533 – PC1066
10.517
P4 2.53 / 533 – ENERGIESPAR
10.594
P4 2.40 / 533 – ENERGIESPAR
11.088
P4 2.40 / 400 – ENERGIESPAR
11.208
XP 2100 + – PC266
11.667
P4 2.20 / 400 – ENERGIESPAR *.
11.952
XP 2000 + – PC266 *.
12.096
XP 1900 + – PC266 *.
12.348
P4 2,0A / 400 – ENERGIESPAR *.
13.176
Information in seconds

In SETI, all processors are actually quite closely to each other. The new Pentium 4 is of course winner with a time of converted 2 hours and 48 minutes. The performance benefit by the higher front side bus is by the way, vanishingly small.

WinACE

Further to clarify the sense and nonsense of a faster processor for frequently used applications, we moved at this point to use WinACE in version 2.11. The program was to grab a 200 MB large .wav sound file with maximum compression level (4096k) as ACE file.

WinACE

< td rowspa
P4 2.53 / 533 – PC1066
193
P4 2.40 / 533 – PC1066
200
P4 2.53 / 533 – ENERGIESPAR
212
P4 2.40 / 533 – ENERGIESPAR
216
P4 2.40 / 400 – ENERGIESPAR
227
P4 2.20 / 400 – ENERGIESPAR *.
244
P4 2,0A / 400 – ENERGIESPAR *.
256

Recent Computer Articles

Wood and plasma tv stands: the epitome of style and sustainability in the scenario

Written by Benedikt Kappel. Posted in Home – Garden & Home Improvement

There are various facets of modern technology which has fascinated many people. It is a well known fact that beautiful wooden tv stands because it offers an old and classic style to your television. Usually, it is better in the workplace elderly because they appreciate the sustainability to the modern style. You must ensure that that they are treated appropriately after the purchase. Many of them are very expensive and should be handled with...

Collaboration between Konica Minolta and Komori Corporation for commercial printing market

Written by Benedikt Kappel. Posted in News

Recently, the commercial print market uses more and more digital printing systems, the increasing of customers very changing missions, low volume and in the short term, and after the sharp rise of the impressions marketing ‘one-to-one’ and variable application request. Konica Minolta BT develops and produces systems digital printing for the market of the printing of production since 1999, on the basis of its extensive assortment of...

Income tax is obligatory and pre-registration

Written by Maike Lehrer. Posted in Finance – Stock Market And Stock Markets

In 2008, filing begins with 2/18. Such people forget VOL.3, US tax! And is rather high. You dispose of stock or mutual fund during the 2007 (transfer income, such as stocks that people) who is eligible, their choice is a general account of a special account or a special account at “withholding is” is, “without withholding information stated please check first. People have to tax The maximum cost is tax definitely...

Full live interview strategy linking the

Written by Swanhilda Hertzog. Posted in Commercial - Business & Economics

The interview is not a gamble. -Evaluating interviewer and elaboration. Interview is a logical process to evaluate you. Compatibility, or lucky, such vague isn’t saying. “See Tate says, also fell into! 」 “I’m always falls in the last interview! 」 “I was what I strained the good! 」 “For example is it if you answered the question? 」 In the toughest event produced and evaluated...

Insurance that if divorce becomes a single mother?

Written by Bernd Abt. Posted in Finance – Stock Market And Stock Markets

To his father’s security! So when single mothers, worst mother even children enough living expenses and school expenses, must be properly to ensure security. However, thinking different reason became a single mother. And because the bigger “bereavement” and divides into a divorce or unmarried birth “, but here, divorced or unwed birth (part from somebody below.)” of consider the case. Responsibility becomes...

Copyright © 2012 Computer Articles. All rights reserved.