AMD vs Nvidia: who makes the best graphics cards?

Mac vs PC. Microsoft vs Google. Intel vs AMD. There's no shortage of monumental rivalries in the tech industry. But the royal rumble between ATI and Nvidia for dominance of 3D graphics is one of the roughest of the lot.
The contest is really AMD vs Nvidia. After all, AMD snapped up Canadian graphics outfit ATI back in 2006.
More importantly, however you slice it the histories of ATI and Nvidia have been very closely intertwined. Both started out as specialists in PC graphics and have since branched out into non-PC platforms such as games consoles, mobile devices and set-top boxes.
More recently, ATI's acquisition by AMD seems to have set it on a very different path for the future from Nvidia. But let's start by reminiscing on a few of our favourite ATI vs Nvidia fisticuffs from yesteryear before taking a look at their current offerings.

AMD vs Nvidia: RIVA and Rage
The early days of graphics on the PC saw 3D cards like Nvidia's RIVA 128 and TNT2 take on ATI's Rage and Rage 128. But it was Nvidia who presaged the modern GPU or Graphics Processing Unit with the mighty GeForce 256 in 1999. It was the first graphics chip with hardware transform and lighting capabilities. And it was fast. Damned fast.
ATI responded in 2000 with the Radeon graphics card. Ever since, successive generations of GeForce and Radeon GPUs have been leapfrogging each other in the race for graphics supremacy. Nvidia had the early advantage with the GeForce, GeForce 2, GeForce 3 and GeForce 4 series arguably having the edge over ATI's Radeon, Radeon 7500 and Radeon 8500.

But in 2002 ATI turned the tables with the awesome Radeon 9700 Pro. The first GPU with fully programmable shaders, the 9700 Pro was massively more powerful than any graphics chip before. It took until early 2003 for Nvidia to respond with the ill-fated GeForce 5800 Ultra, a GPU that never lived up to expectations.
Nvidia was back on form a year later with the GeForce 6800 series. A tit for tat ensued with neither ATI nor Nvidia achieving a clear advantage. It was during this period that Nvidia introduced its revolutionary multi-GPU SLI technology and ATI responded with the copycat Crossfire platform. There really was nothing to separate them.

AMD vs Nvidia: Radeon rethink

At least, there wasn't until ATI released the underperforming Radeon HD 2900 XT. Like Nvidia's calamitous GeForce FX series, the 2900 arrived late, ran hot, underperformed and couldn't match its opposition, the GeForce 8800 Ultra.
But unlike the GeForce FX, it lead to a fundamental strategic rethink. AMD decided that in future ATI would no longer chase ultimate performance with its top GPU. Instead it would aim for maximum bang for buck and introduce dual-GPU boards to cater for enthusiasts demanding ultra-high performance.
The culmination of this rethink was the Radeon HD 4870. Launched in mid 2008, it was half the price of Nvidia's competing GeForce GTX 280 but delivered at least 80 per cent of the performance. It was a winning combination.

AMD vs Nvidia: The DX11 era
Of course, graphics technology waits for no man and much has changed since the Radeon HD 4000 series and GeForce GTX 200 hit the market in 2008. Late last year AMD unleashed the Radeon HD 5000 series, the world's first family of graphics chips with support for the latest DirectX 11 multimedia API from Microsoft, as seen in Windows 7 but also available as an update for Windows Vista.
It took a little longer for Nvidia to respond in kind with the GeForce GTX 400 family. It eventually turned up earlier this year and since then its been these two pixel pumping graphics architectures fighting it out for top DX11 honours.
Topping the current single-GPU tables, therefore, are the ATI Radeon HD 5870 and Nvidia Geforce GTX 480. Thanks to AMD's greater emphasis on value, the GTX 480 weighs in around £100 more expensive at £430 or so.
For the money Nvidia gives you an extra billion transistors for a faintly ridiculous total of three billion. You also get a little more memory as standard, 1.5GB to the 5870's 1GB. However, it's worth noting that 2GB variants of the 5870 are now available for less than the 1.5GB GTX 480.
Anyway, what you don't get from the 480 is a huge performance advantage. Yes it's a little quicker than the 5870. But not nearly as much as it needs to be given the extra cost and complexity.

AMD vs Nvidia: Cut-down cards

It's a similar story further at the next rung down the graphics ladder. Both AMD and Nvidia offer slightly cut down versions of their top GPUs. The Radeon HD 5850 is yours for £225 and retains 1,440 of the 5870's 1,600 stream shaders. Meanwhile, Nvidia's GeForce GTX 470 weighs in around £295 and packs 448 of Nvidia's mighty CUDA cores. The GTX 480, for the record, has 480 cores.
Once again, the 470 is a little quicker than its AMD equivalent, but it's also much more expensive. From there, things get a little more complicated. AMD does a silly-money dual-GPU Radeon HD 5000 board, the 5970. In most tests of pixel pumping prowess, it's the quickest thing out there (NVIDIA has yet to wheel out a dual-GPU take on the GTX 400 series). But just occasionally its dual-GPU architecture and split-memory set up gets the better of it.
Move into mid-range territory and direct comparisons between ATI and Nvidia are currently a bit tricky. That's because Nvidia has yet to release more affordable chips based on Fermi, the new DX11 architecture that underpins the GTX 480 and 470 GPUs.
Consequently, the Radeon HD 5770 (£125), Radeon HD 5670 (£85) and Radeon HD 5570 (£72) are lording it without any DX11 competition.
Instead, Nvidia makes do with older chipsets based on DX10 tech, such as the GeForce GTS 250 (£125) and GeForce GT 240 (£72).

Advantage Nvidia?
That said, Nvidia has recently released an even more cut-down version of the Fermi chip in the new GeForce GTX 465, on sale now from around £230. But what it really needs is some pukka mid-range DX11 chips to take the fight to AMD. And it needs them soon. AMD has already released its family of second-generation DirectX 11 GPUs.
Read ourAMD Radeon HD 6870 review
In the meantime, it's not all bad news for Nvidia. Arguably, it has an edge in at least one important DX11 feature, the hardware tessellator. Designed to spew out huge numbers of polygons and therefore give games more geometric detail and realism than ever before, the tessellator could prove to be the killer feature in DX11. Early tests suggest Nvidia's chips have more tessellation power than ATI's.
Nvidia is also way ahead of ATI when it comes to stereoscopic 3D. Nvidia's 3D Vision technology is the best way to get 3D on your PC today. It works with a large number of games and is also compatible with certain formats of 3D movies including Blu-ray 3D. But like most other 3D display technologies, wearing a pair of geeky goggles is the price of participation.

source of article 

What's the difference between an Intel Core i3, i5 and i7?

What's the difference between Intel Cores?

 

 

What's the difference between an Intel Core i3, i5 and i7?

'What's the difference between an Intel Core i3, i5 and i7?' is a pretty confusing and intimidating question for the majority of people. On the surface, processors are a little tougher to get your head around than, say, a hard drive, as they are not measured by a metric as simple as GB size. However, once you've spent a couple of minutes reading this article, you can be confident of knowing 'What's the difference between an Intel Core i3, i5 and i7?' See also Group test: what's the best processor?

Intel Core i3, Core i5, and Core i7 CPUs have been around for over a year now, but some buyers still get stumped whenever they attempt to build their own systems and are forced to choose among the three. With the more recent Sandy Bridge architecture now on store shelves, we expect the latest wave of buyers to ask the same kind of questions.

What the difference between Intel Cores: Core i3, Core i5, Core i7 — the difference in a nutshell

If you want a plain and simple answer, then generally speaking, Core i7s are better than Core i5s, which are in turn better than Core i3s. Nope, Core i7 does not have seven cores nor does Core i3 have three cores. The numbers are simply indicative of their relative processing powers. See all PC Component reviews.

Their relative levels of processing power are also signified by their Intel Processor Star Ratings, which are based on a collection of criteria involving their number of cores, clockspeed (in GHz), size of cache, as well as some new Intel technologies like Turbo Boost and Hyper-Threading.

Core i3s are rated with three stars, i5s have four stars, and i7s have five. If you’re wondering why the ratings start with three, well they actually don’t. The entry-level Intel CPUs — Celeron and Pentium — get one and two stars respectively.

Note: Core processors can be grouped in terms of their target devices, i.e., those for laptops and those for desktops. Each has its own specific characteristics/specs. To avoid confusion, we’ll focus on the desktop variants. Note also that we’ll be focusing on the 2nd Generation (Sandy Bridge) Core CPUs.

What the difference between Intel Cores: Number of cores

 

 

The more cores there are, the more tasks (known as threads) can be served at the same time. The lowest number of cores can be found in Core i3 CPUs, i.e., which have only two cores. Currently, all Core i3s are dual-core processors.

Currently all Core i5 processors, except for the i5-661, are quad cores in Australia. The Core i5-661 is only a dual-core processor with a clockspeed of 3.33 GHz. Remember that all Core i3s are also dual cores. Furthermore, the i3-560 is also 3.33GHz, yet a lot cheaper. Sounds like it might be a better buy than the i5. What gives?

At this point, I’d like to grab the opportunity to illustrate how a number of factors affect the overall processing power of a CPU and determine whether it should be considered an i3, an i5, or an i7.

Even if the i5-661 normally runs at the same clockspeed as Core i3-560, and even if they all have the same number of cores, the i5-661 benefits from a technology known as Turbo Boost.

What the difference between Intel Cores: Intel Turbo Boost

The Intel Turbo Boost Technology allows a processor to dynamically increase its clockspeed whenever the need arises. The maximum amount that Turbo Boost can raise clockspeed at any given time is dependent on the number of active cores, the estimated current consumption, the estimated power consumption, and the processor temperature.

For the Core i5-661, its maximum allowable processor frequency is 3.6 GHz. Because none of the Core i3 CPUs have Turbo Boost, the i5-661 can outrun them when it needs to. Because all Core i5 processors are equipped with the latest version of this technology — Turbo Boost 2.0 — all of them can outrun any Core i3.

What the difference between Intel Cores: Cache size

Whenever the CPU finds that it keeps on using the same data over and over, it stores that data in its cache. Cache is just like RAM, only faster — because it’s built into the CPU itself. Both RAM and cache serve as holding areas for frequently used data. Without them, the CPU would have to keep on reading from the hard disk drive, which would take a lot more time.

Basically, RAM minimises interaction with the hard disk, while cache minimises interaction with the RAM. Obviously, with a larger cache, more data can be accessed quickly. All Core i3 processors have 3MB of cache. All Core i5s, except again for the 661 (only 4MB), have 6MB of cache. Finally, all Core i7 CPUs have 8MB of cache. This is clearly one reason why an i7 outperforms an i5 — and why an i5 outperforms an i3.

What the difference between Intel Cores: Hyper-Threading

Strictly speaking, only one thread can be served by one core at a time. So if a CPU is a dual core, then supposedly only two threads can be served simultaneously. However, Intel has introduced a technology called Hyper-Threading. This enables a single core to serve multiple threads.

For instance, a Core i3, which is only a dual core, can actually serve two threads per core. In other words, a total of four threads can run simultaneously. Thus, even if Core i5 processors are quad cores, since they don’t support Hyper-Threading (again, except the i5-661) the number of threads they can serve at the same time is just about equal to those of their Core i3 counterparts.

This is one of the many reasons why Core i7 processors are the creme de la creme. Not only are they quad cores, they also support Hyper-Threading. Thus, a total of eight threads can run on them at the same time. Combine that with 8MB of cache and Intel Turbo Boost Technology, which all of them have, and you’ll see what sets the Core i7 apart from its siblings.

The upshot is that if you do a lot of things at the same time on your PC, then it might be worth forking out a bit more for an i5 or i7. However, if you use your PC to check emails, do some banking, read the news, and download a bit of music, you might be equally served by the cheaper i3.

At DCA Computers, we regularly hear across the sales counter, “I don’t mind paying for a computer that will last, which CPU should I buy?” The sales tech invariably responds “Well that depends on what you use your computer for.” If it’s the scenario described above, we pretty much tell our customers to save their money and buy an i3 or AMD dual core.

Another factor in this deliberation is that more and more programs are being released with multithread capability. That is they can use more than one CPU thread to execute a single command. So things happen more quickly. Some photo editors and video editing programs are multi-threaded, for example. However, the Internet browser you use to access Netbank or your email client is not, and is unlikely to be in the foreseeable future.

Hopefully this gives you some insight for your next CPU selection. Happy computing!

Source of article: http://www.pcadvisor.co.uk/