AMD vs. Intel: 64-bit Windows Dual-Core Shootout
|
When it looked like AMD was going to render Intel's multibillion dollar IA64 effort (aka Itanium) irrelevant with its brassy push into 64-bit computing using an extended x86 architecture, Intel responded with EM64T. When it was first revealed that Intel had an x86-64 project in the works, codenamed Yamhill, the company at first denied the project's existence. Eventually, though, EM64T surfaced, and it proved to be a near-clone of AMD's x86-64. There are some subtle differences in calling conventions between the two processors, but the differences are slight enough that compatibility isn't usually a worry.
Last week in Part I, we took a look at the AMD dual-core Athlon 64 X2 4800's performance on 64-bit Windows. As it turned out, Windows XP Pro X64 ran most 32-bit applications just fine on the X2 4800, and 64-bit code showed a few modest performance gains.
ADVERTISEMENT
Our article generated considerable interest in the community at large. We received several interesting emails on the topic. One of the most salient was from Joe Landman of Scalable Informatics. Scalable ran a variety of 32-bit-vs-64-bit tests on Opteron CPUs. Although somewhat synthetic, the tasks run were applicable to certain types of scientific computing. SI found that 64-bit code was almost always faster than 32-bit code, when properly written.
David Wren of Passmark Software, an Australian utility software company, echoed those findings, and allowed us to use Passmark's Performance Test, which is available in both 32-bit and 64-bit versions. PPT is synthetic in nature, but can yield interesting results that point to potential theoretical gains with 64-bit code.
Given Intel's renewed focus on both 64-bit x86, coupled with its rapid, if somewhat improvisational, push into dual-core processing, we wanted to see just how well Intel's dual-core processors fared with 64-bit Windows relative to AMD's best. So we loaded up a system with an Intel Pentium Extreme Edition 840 and took it for a spin. Intel is nothing if not a fierce competitor, so let's see how the dust settles.
source: http://prethi.blogspot.com/2007/08/amd-vs-intel-64-bit-windows-dual-core.html
AMD, Intel: Who's Zooming Who?
|
Race for bragging rights favors Intel again -- for now. But AMD is fighting back with a workaround it says will keep it in the running until it can bring out its own quad core. By Dan Goodin.By Dan Goodin.
Intel vs AMD
|
Intel vs AMD
Could it be the folks over at Intel are getting a little nervous?
Word on the street is the rival Advanced Micro Devices has been eating their lunch in the small, but very lucrative four-way server business (that’s four or more processors in one system). AMD also is doing brisk business in the entry-level server market.
Never a company to take any lost sales standing still, Intel announced today it “is accelerating the availability” of its dual-core Xeon and Xeon MP processors, introducing new chips at some unspecified time this year instead of next. They’ll be targets at “early adopters and evaluators of dual-core technology.”
In an oddly worded release, the chipmaker then adds that these new processors really are a prelude to two other chips that will be introduced in early 2006.
What gives? Behind the curtains, one might expect to find that Dell has been pressuring Intel to come up with something to help them compete with hot-sellig systems from HP, Sun and others, who have heartily been jumping on the AMD bandwagon.
And AMD execs likely are smirking at the idea that dual-core technology still needs a lot of testing, since their own products have been flying off the shelves. They say they have an inherent advantage over Intel because their architecture is vastly more energy -efficient, on top of being powerful in their own right.
While Intel is still making a bundle off its Xeon chips, it seems clear they’re feeling the heat for the first time in a long time from AMD. And if researchers from IDC or Gartner in coming weeks confirm that AMD continues to gain share in that space, it’ll be a black eye that will make Intel very unhappy indeed.
Intel prices up four wireless modules for 'Centrino 2'
|
Intel's upcoming WiMax module for laptops, 'Echo Peak', is going to add at least $34 to the price of a new notebook, if leaked pricing information is to be believed. But watch out for cheaper offerings.
Why? Because Intel will be offering two versions of Echo Peak. It's said as much itself, but thanks to Taiwanese industry moles cited by DigiTimes, we have the prices.
The top-of-the-line Intel WiMax/Wi-Fi Link 5350 module costs $54, the cheaper one $44. That $10 premium buys you a module capable of using a 3x3 multi-antennae array. The $44 unit - aka the Intel WiMax/Wi-Fi Link 5150 - can only connect to a 1x2 array.
That's three transmit and three pick-up antennae versus one transmit and two pick-ups, so the former should get you superior wireless reception. That's not to say the 5150 won't perform well, only that the 5350 will perform better.
So make sure you check the spec of the WiMax laptop you plan to buy, and don't assume all Intel WiMax/Wi-Fi Link modules are created equal.
The Echo Peak modules are due to debut in May with the arrival of 'Montevina', the next version of Centrino - Centrino 2, as it'll apparently be branded.
So too will 'Shirley Peak' - formerly 'Dana Point' - a WLAN module without WiMax. Shirley Peak will likewise appear in two forms: the 3x3 Wi-Fi Link 5300 and the 1x2 Wi-Fi Link 5100. It's said they'll be priced at $30 and $20, respectively.
© The Register.
AMD readies third Radeon HD 3800-series GPU line
|
AMD will next month launch the ATI Radeon HD 3830, an 'RV670'-based graphics card said to have been designed to pull the rug from under Nvidia's low-end-of-the-top-range offerings.
So claim industry moles cited by Chinese-language site HKEPC. They say the 3830 will contain the same 320 unified shader processors as the 3850 and 3870, but connect to memory over a 128-bit bus rather than the 256-bit bus the higher-spec GPUs use.he 3830 will also be limited to 256MB of graphics memory.
The moles suggest the 3830 will be released in the middle of April, with boards costing gamers between $119 and $129 or equivalent. AMD will apparently pitch the GPU as a product more capable than Nvidia's GeForce 8600 GTS but cheaper than the recently released 9600 GT.
© The Register.
Intel 'Nehalem' CPU 'borrows' AMD Phenom cache plan
|
Intel's 45nm 'Nehalem' processor architecture, due for release later this year, will see the chip maker adopt AMD's approach to cache structure: small per-core Level 1 and Level 2 caches connected to a big, shared Level 3 cache. 
Nehalem, which will form the basis for two-, four- and eight-core processors, will contain 64KB of L1 cache per core, split 50:50 between memory reserved for program instructions and for data. That's current how Core 2 CPUs work, but while today's desktop and mobile CPUs complement that with a big, multi-megabyte L2 caches shared between pairs of cores, each Nehalem core will get 256KB of L2 cache of its own.
All two, four or eight cores will then be able to access a shared pool of up to 8MB of L3 cache memory, allowing them to take as much or as little as they need for the threads they're running up to the overall limit.
Intel's Nehalem: native quad-core
It's an approach AMD introduced with its Phenom chips. Earlier AMD processors gave each CPU both its own L1 cache and L2 memory. Intel previously poo-poo'd this design, claiming better performance could be achieved using a shared L2. Whatever the reason, the Phenom CPU line introduced a third tier of cache, this time shared.
The Phenom 9600, for example, has 2MB of L2 divided into four 512KB blocks, each assigned to a single core. All four cores share a further 2MB of L3. Each core has 128KB of L1 cache.
It's a logical move for Intel as it was for AMD. The exclusive L2 caches give each core a pool of fast-access memory, while the shared cache acts as a buffer to trap data and instructions other cores may have requested and which another core can now grab more quickly that going out to main memory or peeking onto other cores' personal storage.
More to the point, since Nehalem is essentially Intel's first design - as AMD's have been for some time - that doesn't build four-core CPUs out of groups of two two-core dies. With no shared L3, the core-pairs in today's Core 2 Quad and Core 2 Extreme processors have to look in other core-pairs' caches, which can hinder performance.
Each Nehalem core uses Intel HyperThreading technology to handle up to two processing threads in execution simultaneously, allowing a four-core chip to appear to the host OS as an eight-core part.
Nehalem will initially be a 'true' quad-core part, but Intel promised future, eight-core parts that are built natively rather than from a part of quad-core CPUs bolted together.
The CPU design incorporates an out-of-order window running to 128 instructions, up from Core 2's 96 instructions. That allows the new chip to look ahead to a greater number of instructions to see which can be pulled out of the program sequence and processed without affecting the results of operations further down the line. It's also able to keep 33 per cent more micro-ops in flight at once than its predecessor could
© The Register.Intel E6750 Conclusions
|
On all our tests Core 2 Duo E6750 performance was identical to Core 2 Duo E6700, even though it uses a faster front side bus – and we simulated the use of several different types of application.
So what is the advantage of using a Core 2 Duo E6750 instead of a Core 2 Duo E6700? Based on the results from our tests, none.
Then why Intel is launching this new 1,333 MHz external bus? We have two answers for that.
First, on quad-core CPUs the higher external clock rate may increase the system performance. This happens because currently on Intel quad-core CPUs the cores are arranged into two pairs. The cores inside each pair can talk directly to each other, but if they need to talk to a core that belongs to the other pair they need to go to the front side bus and make this connection going thru outside the processor. Increasing the CPU external clock rate increases, at least in theory, the speed each core pair can talk to each other.
We will have to wait until we review a quad-core CPU based on this new FSB to see if this is really the case. The second reason we can see is regarding DDR3-1333 memories, which will be supported by the Intel X38 chipset to be launched in July. Using DDR3-1333 memories with a Core 2 running externally at 1,333 MHz you will be able to match the FSB clock with the memory clock, allowing you to achieve the maximum performance with current technology. In retrospect we must keep in mind that this is the first time in six years that Intel is launching a new front side bus speed with memories matching it. When the Pentium 4 was launched, for example, it used the then-new 400 MHz FSB and there were no 400 MHz memories at the time – this was PC-100 times and DDR-400 wasn’t available yet. The same thing happened when they launched the 533 MHz, the 800 MHz and even the 1,066 MHz external clock speeds. Even to this date Intel chipsets do not officially support DDR2-1066 (even though Intel P965 and P35 can access memories at 1,066 MHz just fine and P35 officially support DDR3-1066). DDR3-1333 availability may be an issue, but that is a totally different story.
Intel E6750 Testing: SYSmark2004
|
SYSmark2004 is a program that simulates the use of real-world applications. Thus, we consider this the best software to measure, in practical terms, the system performance.
The benchmarks are divided into two groups:
- Internet Content Creation: Simulates the authoring of a website containing text, images, videos and animations. The following programs are used: Adobe After Effects 5.5, Adobe Photoshop 7.01, Adobe Premiere 6.5, Discreet 3ds Max 5.1, Macromedia Dreamweaver MX, Macromedia Flash MX, Microsoft Windows Media Encoder 9, McAfee VirusScan 7.0 and Winzip 8.1.
- Office Productivity: Simulates the use of an office suite, i.e. simulates sending e-mails, word processing, spreadsheets, presentations, etc. The following programs are used: Adobe Acrobat 5.05, Microsoft Office XP SP2, Internet Explorer 6.0 SP1, NaturallySpeaking 6, McAfee VirusScan 7.0 and Winzip 8.1.
This software delivers several results, all of them using a specific SYSmark2004 unit. First we have a SYSmark2004 overall score. Then we have a group result for each batch listed above. And for each batch, we have specific results: 3D Creation, 2D Creation and Web Publication for Internet Content Creation; and Communication, Document Creation and Data Analysis for Office Productivity.
For a better visualization, we separated the results into three graphs: overall score, Internet Content Creation score and Office Productivity score.
| SYSmark2004 – Overall | Score | Difference |
| Core 2 Extreme QX6700 (2.66 GHz) | 368 | 6.36% |
| Core 2 Extreme X6800 (2.93 GHz) | 367 | 6.07% |
| Core 2 Duo E6700 (2.66 GHz) | 346 | 0.29% |
| Core 2 Duo E6750 (2.66 GHz) | 345 |
|
| Pentium 4 550 (3.4 GHz) | 203 | 70.44% |
On the overall score Core 2 Duo E6750 achieved the same performance of Core 2 Duo E6700.
| Difference | Internet Content Creation | 3D Creation | 2D Creation | Web Publication |
| Core 2 Extreme QX6700 (2.66 GHz) | 14.16% | 16.54% | 16.22% | 9.82% |
| Core 2 Extreme X6800 (2.93 GHz) | 7.42% | 9.27% | 6.13% | 7.30% |
| Core 2 Duo E6700 (2.66 GHz) | -0.45% | -1.25% | -1.80% | 2.02% |
| Pentium 4 550 (3.4 GHz) | -48.09% | -44.61% | -48.29% | -51.13% |
Core 2 Duo E6750 achieved a performance similar to Core 2 Duo E6700 on Internet Content Creation score and in all of its individual tests. Here you can also see that Core 2 Extreme QX6700 had a good advantage because of its four cores, even though these three CPUs run at 2.66 GHz.
| Difference | Office Productivity | Communication | Document Creation | Data Analysis |
| Core 2 Extreme X6800 (2.93 GHz) | 5.22% | 7.50% | 6.43% | 2.17% |
| Core 2 Duo E6700 (2.66 GHz) | 0.75% | 10.63% | -4.83% | -2.80% |
| Core 2 Extreme QX6700 (2.66 GHz) | -0.75% | 3.13% | -2.14% | -3.11% |
| Pentium 4 550 (3.4 GHz) | -33.58% | -15.00% | -45.31% | -37.27% |
Core 2 Duo E6750 achieved the same performance level of E6700 on the Office Productivity score. However, on Communication Core 2 Duo E6700 was 10.63% faster, but this CPU was 4.83% slower than the reviewed CPU on Document Creation. On Data Analysis both CPUs achieved the same performance level. Here you can see that the extra cores brought by Core 2 Extreme QX6700 made no difference, as office applications do not take benefit of having more than two cores (or two CPUs) on the system. Thus the CPU with the highest clock – Core 2 Extreme QX6800 – was the fastest.
Subscribe to:
Posts (Atom)
