Intel, Asus, Cooler Master, Corsair Memory, Dell, Compaq, Gigabyte, Mercury, zebronics, AMD, Nexus, Delta, IBM, HP, Apple, Acer, BenQ, Sony, Samsung, LG, Philips, Transcend, Nvidia, SiS, Logitech, Alps Electric Corporation, Creative Technology, ASRock, Asus

Blog Archive

Showing posts with label upgrade:. Show all posts
Showing posts with label upgrade:. Show all posts

Friday, 3 June 2011

Buying Guide: RAM upgrade: the complete guide

Not too long ago, a memory upgrade was one of the most effective, easiest ways to boost your PC's performance. High memory prices meant that new machines often shipped with the bare minimum, yet everything from multi-tasking to gaming lapped it up.

Bandwidth-critical image and video editing showed huge changes, especially if you upgraded to faster, lower latency memory. With all this in mind, you'd be daft not to consider a memory upgrade.

These days, you'll find that even the most modest of systems comes with 4GB of DDR3 RAM as standard, and if you're building a machine from scratch, there's no real incentive to start out with anything less. Does this mean that memory upgrades are a thing of the past, though?

Not at all - it simply means that if your machine has a decent amount of memory already, the post-upgrade improvements are going to be far more subtle than before.

If you're looking for specific numbers, 4GB should be seen as a minimum for most modern applications. In turn, this dictates that you'll need a 64-bit operating system to use all of the available capacity.

8GB has quickly become the norm for the more advanced enthusiast systems, while capacities of 12GB and higher are no longer the exclusive reserve of video jockeys, high-resolution image tweakers or programmers.

With the likes of the Intel X58 chipset boasting support for an addressable limit of 24GB, there's now plenty of scope for memory upgrades on the very latest machines.

As ever, you need to match any potential upgrade to your system's capabilities. With more recent systems, that means matching it with your processor. Modern CPUs have many of the features that were once reserved for the motherboard, and the most important one when it comes to the picking and upgrading of memory is the integrated memory controller in the processor itself.

Here, we'll cover everything you need to know to make the best decisions for upgrading your computer's memory. We'll start by considering the current state of the memory market and which technologies are on the horizon, then go on to highlight the best budget, performance and all-round kits that are out there right now. Finally, we'll show you how to get the top performance out of any memory kit you do buy.

The road ahead

The most recent change to the memory landscape has been the slow but inevitable transition from DDR2 to DDR3. The former is still around for older systems and laptops, but DDR3 is the future.

This makes the process of upgrading memory a lot simpler, because there's now only one standard to consider. The only question is, how many sticks do you need to buy?

X58

Bloomfield (and later Gulftown) Core i7 processors boast a triple-channel memory controller. You'll need to populate all three channels on your X58 motherboard to get the most out of this chip. This means using a minimum of three memory sticks and, depending on the capabilities of your motherboard, could make for some rather tough upgrade decisions.

Many of the more affordable X58 motherboards boast three or four memory slots. For those with three 2GB sticks, the jump to 12GB is going to be significantly more wasteful, not to mention pricier, than for those users who are trying to hit the same capacity using a motherboard with six memory slots.

The decision is easier for Intel's more recent Core i5/i7 processors and AMD's latest chips. In both cases, dual-channel DDR3 is on the respective roadmaps for the foreseeable future.

All that's left to consider is how much you get, and at what speed. Operating frequency is less important here than latency - faster responses to memory requests outweigh the benefits of just running the memory ever faster. This fact is keenly represented by the price of memory - low-latency memory costs more.

Prices continue to tumble, though, and even if you can afford the cash for your memory upgrade, you'll still find you're getting something of a bargain.

However, this price slashing has had a knock-on effect within the memory market, with many manufacturers blaming low margins as they pull out of production or downscale their operations. The current low prices we're seeing may not last forever, though.

Successors to current memory standards are some way off too. We don't mean from a technological point of view - Samsung has already shown DDR4 modules at the Consumer Electronics Show - but any new ideas are hampered by the fact that there's simply no need for replacement technology yet. DDR3 will be with us in some form for a good while yet.

Part of the problem can be pointed at Intel's door, since it included an integrated triple-channel memory controller in its first Nehalem processors. That third memory channel boosted the available bandwidth to the point that it simply wasn't a factor in normal operations anymore.

The subsequent decision to drop back to dual-channel controllers in the latest mainstream platforms shows that the world simply isn't ready for ever-faster memory standards.

Three top RAM choices

Corsair Dominator GT (4GB (2x 2GB) 2,133MHz)
Performance
Price: £108
Info: www.corsair.com

Corsair dominator gt

£140 is a lot of cash for a dual-channel 4GB memory kit, even one rated at a lofty 2,133MHz. In fact, look a little closer at the Corsair Dominator GT 2133MHz's specifications and the pricing looks even more preposterous. After all, timings of 9-10-9-24 are nothing special, even taking into account the flighty frequency.

However, that would be to disregard a number of benefits that come with this DDR3 kit. For starters, it's made by Corsair, and that means it's about as impeccably hewn and physically desirable as they come. There's nothing flashy or adolescent. It's just quality engineering and it bodes well for operating at high frequencies.

You also get Corsair's Xtreme Cooling System which involves a twin-fan cooling attachment that clips on top of the DIMMs. With all that in mind, the extra £4 Corsair is demanding compared to Geil Evo Two 2000MHz actually looks pretty reasonable.

Verdict: 4.5/5

Read our full Corsair Dominator GT review

Crucial Ballistix DDR3 (4GB (2x 2GB) 1,600MHz)
Budget
Price: £43
Info: www.crucial.com

Crucial ballistix ddr3

Crucial's Ballistix range of high performance DDR3 memory modules are well known amongst the overclocking fraternity offering faster clock speeds and lower latencies and in some instances fancy LED displays, than the standard Crucial product range.

The latest addition to the family line-up, is a 4GB memory kit of DDR3-1600 (PC3-12800) modules, carrying the part number Ballistix BL2Kit25664FN1608 if you want the full nomenclature.

The new modules also bring a change in module design to the product line and a new feature which some overclockers will find very useful.

The small discrete heatsink covers found on the rest of the Ballistix range have been replaced by the high profile finned heat spreaders favoured by most of Crucial's competitors on their high performance memory modules.

Verdict: 3.5/5

Read our full Crucial Ballistix DDR3 review

Kingston HyperX (8GB (2x 4GB) 1,600MHz)
All round
Price: £67
Info: www.kingston.com

Kingston hyperx

Kingston is a stalwart of the memory game, and it's ranges like HyperX that keep it so relevant when it comes to both performance and excellent value. Unlike the other two kits mentioned on this page, this isn't a 4GB package but a pair of 4GB modules, totalling 8GB. That's an incredible amount of storage for surprisingly little money.

With latency timings of 9-9-9-27, there are certainly speedier options around, but unlike much of the competition, there's some room for a little overclocking. With a base frequency of 1,600MHz, that latency certainly isn't as bad as it could be. If you're tempted to overclock these modules, you'll need to tweak the voltage in order to maintain stability.

The low-profile head spreaders mean that they'll fit into any chassis you can throw at them, and with 8GB of space on offer, it's a flexibility that's more than welcome. Only hardcore overclockers may feel a little constrained - everyone else will be delighted.

Verdict: 4.5/5

Memory is easy to install, but getting the best performance out of it takes a little more effort

laptop

We'd hazard a guess you know how to fit more memory to your PCs, replacing the current sticks if necessary. The mechanics of fitting a stick of memory into place haven't changed much in the last 10 years.

For laptops, it's simply a case of sliding the SODIMM into the slot and pushing the card into place until it clicks. There's nothing more to it from a physical perspective.

The process for desktops isn't much more taxing. Start by making sure the memory is the right way round, so that the notch lines up with the socket correctly. Then, slide the memory stick in firmly until the clips on either side flip into place, and that's it - you're done.

Installing memory is usually far more straightforward than working out what your upgrade options are, or configuring it for optimal performance.

It's taken some time, but motherboard manufacturers are finally cottoning on to the fact that graphics card slots are often in the way of memory clips, and some of them are now using clipping mechanisms that only have a clip at one side. This means that you have to slide the memory stick in and rotate it into place, but at least you don't graze your knuckles on the back of the graphics cards.

The most crucial changes have happened in the Serial Presence Detect EEPROM that resides on each memory module. The SPD contains the timing information for the module, but due to the way companies traditionally erred on the side of compatibility above all else, getting the best performance out of a module was left to overclockers.

This changed when Nvidia and Corsair decided to use the spare capacity of the SPD to define performance at different speeds to the standard JEDEC settings. Most manufacturers now use 256-byte EEPROMs for the SPD, while the specification only defines the first 64 bytes for timing information.

Known as Enhanced Performance Profiles, or EPP for short, the extension to the SPD for DDR2 modules meant this unused space could be used to store faster memory settings than those defined by JEDEC. With a compatible motherboard, this meant the memory could be configured to run painlessly at faster speeds.

Intel followed this lead with its own profile extensions for DDR3, Extreme Memory Profiles, using the unused EEPROM space to the same effect. This is how memory brands can sell kit rated above JEDEC specifications, yet users can still exploit the faster speeds easily. To check you've got your memory configured optimally, see the walkthrough below.

Optimise your RAM

1. Check your settings

step 1

When you install memory, the BIOS should get the timing information from the SPD on the modules. The only potential problem is if you've got two types rated at diff erent speeds.

CPU-Z lets you check memory settings within Windows. You'll need to run RAM at the slowest latency settings supported by all your machine's modules or you'll have memory misses, which will slow operations down.

2. Going extreme

step 2

The JEDEC memory profiles err on the safe side, so even if your memory is sold as running at 1,600MHz, it won't operate at that speed by default.

Enter your BIOS configuration screen, go to the memory settings and elect to use the XMP (or EPP) settings instead. Set the speed at which the memory should operate, and the actual timings are all done for you. Check that the changes have taken effect with CPU-Z.

3. Overclock it yourself

step 3

Most recent memory modules will overclock further than even the XMP defines, though you'll need to experiment a little to get the best performance.

Enter your BIOS and increase the memory speed up to the next stock frequency, checking it still boots. Rinse and repeat until the machines stops working, reset the BIOS to the frequency you know works. Benchmark the performance increase to see if it's worth it.



Buying Guide: Hard drive upgrade: what to buy and how to fit it

Hard drives - or non-volatile storage devices, to give them their proper name - have long been essential for computing, but you could be forgiven for thinking that the HDD's days are numbered.

If you believe the cloud computing fundamentalists, we'll be living and working in a browser soon, accessing applications on a North American server farm and maybe storing data elsewhere. This style of computing has its advantages, but we're great fans of owning our own applications and controlling our own data.

Given this philosophical position, we're going to need somewhere to store all our data and - despite technology's constant evolution - there's little better than a humble hard disk.

The current market is split roughly in two. On one hand are the traditional mechanical hard disks, with spinning platters and flicking heads. On the other are solid state hard disks. These have forgone magnetism and mechanics in favour of flash memory.

Once the stuff of fantasy, SSDs are now a financial reality if your pockets are a bit deeper than average. The question is, which do you choose?

The answer, as we'll see, isn't quite as straightforward as it might seem. If you're after a silent PC that goes like the clappers then an SSD might seem like the natural choice, but there's more to it than that. Sure, SSDs have come down in price lately, but even so, traditional hard disks still offer a cost per gigabyte that's enough to make an SSD blush.

SSD

There are also certain situations in which mechanical hard disks will outperform their more technically advanced brethren. We'll explore these later.

Don't get disheartened, though - there's a smart middle ground that lets you enjoy the inherent benefits of both technologies: why not fit both types of drive into your PC?

It's perfectly feasible to install a small SSD as your main boot drive. This should ensure that your OS and a few key applications boot quickly. Alongside this, you can then install a cheap yet capacious mechanical drive for your games, MP3s and videos. This might seem like an overly complicated procedure, but if you want to create a well balanced system then it may be the perfect compromise.

However you choose to upgrade your machine, the following pages should prove invaluable when you're considering how to store your data.

What's the best form of storage right now, and what will it be in a year?

hard drive

Storage is currently in a state of flux. Given technology's relentless march, it might seem as though disk manufacturers are always changing this and upgrading that, but some fundamental shifts really are taking place at the moment.

Mechanical hard drives, and to a lesser extent SSDs, are now undergoing significant alterations that are changing the capacity and speed the technologies can offer. The most immediate of these shifts is happening in the hard drive space, and it's another instance of hard drives smashing through a capacity limitation that only serves to highlight the legacy protocols modern computers are built on.

The 2.19TB limit of the 32-bit Master Boot Record (MBR) is compounded by the standard PC BIOS limitation of not being able to boot from a drive that's larger than 2.19TB. Solutions to both limitations already exist, but you'll need a modern machine if you want to use new, larger drives without having to rely on an expansion card.

Windows 7, and indeed Windows Vista, offer support for the replacement for the MBR, namely the GUID Partition Table (GPT). The replacement for the BIOS has been a long time coming, but the Extensible Firmware Interface (EFI) has finally made an appearance with the release of Sandy Bridge. It's supported by motherboards that boast P67, H67 and H61 chipsets.

Before Sandy Bridge landed, Western Digital launched its first drives to have broken through this limit. The 2.5TB and 3TB Caviar Green models might not have set the world alight with their relatively slow 5,400rpm spin speeds, but the capacity on offer is a hint at what's coming.

Importantly, when these drives were released, Western Digital included an AHCI daughter card that provided full access to their capacities on machines hobbled by the ageing BIOS.

Now that Sandy Bridge is out there in the wild, we can expect other hard drive manufacturers to follow WD's lead and release models larger than 2TB. With these limits smashed, the focus will return once again to increasing the density of the data packed into a magnetic platter.

The areal density, as it's called, has turned into something of a point of obsession among disk makers. Two technologies are currently in development that will further increase the storage capacity of hard drives - heat assisted magnetic recording (HAMR) and patterned media.

The differences for the SSD market may seem less profound, but we could see both costs and speeds improving within a few months. On the price side of things, the relatively high margins to be had from SSDs are now starting to attract more manufacturers, and we can expect to come across plenty more companies selling drives this year.

Sandforce

SSD transfer speeds are expected to jump yet again as well, spearheaded by the new SF-2000 controller chip from SandForce, with the other brands no doubt following suit. SandForce is a key player in the SSD market, having shipped a million controller chips in 2010. Intel is still a company to watch when it comes to SSDs too.

OCZ 240GB IBIS
Performance
Price: £540
Info: www.ocztechnology.com

OCZ ibis 240gb

It's fashionable among the technologistas to grumble about the quality and performance of solid-state drive controller chipsets. But controllers aren't the only problem with SSD performance; increasingly, storage interfaces are creating a bottleneck.

Enter the OCZ IBIS HSDL 240GB, an ultra-high performance SSD designed to sidestep performance issues related to the SATA I/O interface.

The OCZ IBIS HSDL 240GB does this courtesy of its own unique storage connection. Out goes SATA, in comes OCZ's proprietary High-Speed Data Link or HSDL for short. In simple terms, HSDL is a four-lane PCI Express link. PCI Express 2.0, of course, supports 500MB/s per lane.

Potentially, therefore, HSDL can manage a massive 2GB/s of raw storage bandwidth. This first incantation of HSDL is PCIE 1.1 and thus half that speed, but on paper it's still a fair bit quicker than SATA.

Verdict: 4/5

Read the full OCZ 240GB IBIS SSD review

Seagate 1.5TB Barracuda 7200.11
Budget
Price: £54
Info: www.seagate.com

Seagate barracuda 1.5tb 7200.11

Seagate's 11th generation 1.5TB Barracuda 7200.11 hard disk was the first 1.5TB disk to market, and although it's been around for a while, it's still a very capable and popular drive.

It's easy to see why, with its combination of performance, capacity, price and fairly low power consumption. It's a reflection of just how far storage technology has come in a relatively short space of time that a drive that was state of the art just over a year ago - 7,200RPM spindle speed, Perpendicular Magnetic Recording (PMR) technology, a compact disk density of 375GB, Native Command Queuing (NDQ) - is old hat among the 2.5TB and 3TB monsters available now.

The ST31500341AS uses four platters to get to its 1.5TB capacity, backed by a 7,200rpm spindle speed and a 32MB cache. Despite having been around for a while, the ST31500341AS offers an even better sweet spot for price versus capacity now than when it was launched, especially considering hard drive prices are now as cheap as they've ever been.

Verdict: 4/5

Read the full Seagate 1.5TB Barracuda 7200.11 review

Crucial 64GB RealSSD C300
All round
Price: £89
Info: www.crucial.com

SSD

Solid state drives are definitely coming of age. Not only are they increasingly the norm for mid-range laptops and pre-built gaming systems, but we've seen huge leaps in performance lately. From Kingston's SSD Now V+ series to the SandForce-powered Agility 2 and Vertex 2 from OCZ, things are looking good for SSDs.

Crucial's latest realSSD C300 drive is available in three sizes, 64GB, 128GB and 256GB. Inside the drive, the memory chips naturally come from Crucial's parent company Micron, and the controller is by way of Marvell.

There are two things that are really interesting about the realSSD C300, though. First up, it's reasonably priced - a little bit more expensive than Kingston's perhaps, but a lot cheaper than OCZ's offerings.

Secondly, it's one of the first internal drives of any flavour we've seen that make use of the new 6Gbps SATA standard.

SSD read speeds have been getting very close to the limits of the older SATA 3Gbps bandwidth cap. Does changing the interface let them off the leash?

Verdict: 5/5

Read the full Crucial 64GB RealSSD C300 review

How to install a new drive

Modern hard drives are easy to install. Connect the power cable and the SATA data cable and boot the machine to make sure the drive works. All being fine, screw the drive into place.

If you've gone down the SSD route then you'll need to use a mounting bracket to secure the 2.5-inch drive in a 3.5-inch drive bay. Beyond fighting with the power cabling from your PSU, and the strange way that manufacturers can place the drive connectors in the middle of the cables, that's about as complex as it gets when it comes to physically installing a drive.

Things get a little trickier when it comes to laptops, netbooks and all-in-ones, because there's not generally enough room to have two drives in the machine at the same time. This means that you'll need to replace what's currently there and usually reinstall Windows anew.

It's worth checking the connection standard used by your current hard drive in your laptop before buying an upgrade though, as not only do PATA still pop up from time to time, but manufacturer-specific interfaces are also known to rear their ugly heads.

The Serial ATA standard stipulates that there's only one device per connector, which means the problems of clashing master/slave settings with the old PATA standard are a thing of the past. Indeed, there's little reason to touch a PATA drive these days.

Even if you're looking at upgrading an old machine, we'd recommend getting an affordable SATA expansion card that will bring you up to date. Either that or buy a brand new motherboard.

Realistically, the only thing that can really go wrong with a modern drive installation, other than you having a faulty drive, is a dodgy cable. Even now, SATA cables are still prone to popping out of the connectors, despite there being various clips and springs designed to keep them in place.

Also, the plastic shields on the connectors can perish as you're trying to push them into the socket. The tiny size of the connectors can make slotting them home pretty frustrating as well.

Indeed, the toughest part of installing a new hard drive is working out how to integrate it into your current configuration. Should you install Windows from fresh, or copy your current installation across using the likes of Norton Ghost and Acronis True Image?

We'd generally recommend installing Windows from new, if only for the opportunity to start again and get rid of all of that software you never use anyway.

Troubleshooting drive installation problems

1. Check the BIOS

step 1

Your BIOS (or EFI if you have a newer motherboard) should pick up the hard drive without problems. Even so, it's still the first place to check if Windows is refusing to play ball with the new hardware.

It should be listed in the hard drive section of the BIOS. If it isn't, try plugging the drive into a different SATA port, or use a different cable. Put simply, if you can't see it in the BIOS, you won't be able to see it in Windows.

2. Administrative privileges

step 2

The next stage in troubleshooting your missing drive is to use the Administrative Tools, which can be found in the Control Panel.

Select 'Computer management | Storage | Disk management' to see a list of all the drives attached to your system. The drive should be here. If it's refusing to cooperate, you may have to remove any existing partitions, create a new one, format it for use and assign a drive letter.

3. Upper management

step 3

Windows may still not present the drive in Explorer, though this is generally down to a driver issue or confusion in the device manager.

The easiest fix is to open the Device Manager (hold [Windows]+[Pause/Break] and then click 'Device Manager') and remove the drive entry from the Disk Drives list. Scan for changes and the issue should be sorted. Failing that, check the drive works in another machine before continuing.



Buying Guide: Graphics card upgrade: your options explained

The graphics card is the racehorse of your PC components stable. It's a high-value add-in board that's traditionally done one thing and one thing only: let you play games with all the latest graphical wizardry.

Increasingly though, the graphics card is becoming far more than just a gamer's luxury. With architecture improving year on year, 3D graphics aren't the only thing your discrete GPU can do.

It can now be used to enhance your web browsing experience and enjoyment of high-definition media, let you explore your creative side with enhancements for productivity software and even help cure terminal diseases through projects like Folding@home.

So as well as producing some stunning visuals, your graphics card can also help save lives.

In the last decade, GPUs have been following in the footsteps of the CPU market, with increased core and thread counts. Speed in MHz or GHz is no longer the only measure of a chip's power, whether it's a GPU or a CPU.

What counts now is the number of cores, and how much data the chip can process at any one time. In CPU terms, the maximum on the desktop is six cores and 12 threads, and a full-fat 12 cores in the server space.

The top-end Nvidia GPU – the GeForce GTX 580 – has 512 CUDA (Compute Unified Device Architecture) cores. Meanwhile, the AMD Radeon HD 6970 has 1,536 shader processors, all of which are simple processors capable of taking on tasks such as video encoding, where some simple parallel processing is needed to enhance speed.

Nvidia was first to take this on with its CUDA cores, which let programmers write code in industry-standard languages such as C++. This code is run using all the shader processors (or CUDA cores) in Nvidia's GTX 8-series onwards.

Microsoft's latest update to its graphics API – DirectX 11, has done a similar thing with its Compute feature, which enables general purpose applications to run through a DirectX-capable GPU rather than taxing the processor.

If the GPU is becoming ever more powerful, why is there such doom and gloom around the discrete graphics card market? According to Intel and AMD, the future is fusion.

Are integrated graphics the next big step in the great graphics war?

Fusion

There are many reasons to be upbeat about the future of discrete graphics cards. There isn't going to be a new games console release for another couple of years now, and the mid-range cards of today are far superior to anything the Xbox 360 or PS3 contain, so the PC is the platform to go for if you want to see the top releases looking their best.

Integrated graphics (the graphics processing power that traditionally comes with your processor chipset combination of CPU and motherboard) are catching up, though. They're changing as well – moving from the motherboard onto the CPU. All the big boys are getting involved.

First there was Intel and its Arrandale processors, which packaged a GPU and CPU on the same chip. Then came the company's Sandy Bridge, with its fully integrated processor graphics.

AMD has recently released its first Fusion board to the world, housing a tiny CPU and GPU setup – the first new CPU architecture we've seen from the company in years.

At this year's Consumer Electronics Show, held in Las Vegas in January, Nvidia announced Project Denver, its own collaboration with ARM to create a powerful desktop CPU with Nvidia's GPU architecture built right in. This may not shake up the high end of the discrete graphics market – after all, the latest 3D games are still going to need a power-hungry graphics card sitting in that PCI Express slot – but the value end of the market is going to change.

Processor graphics will be more than capable of coping with high definition video, encoding and casual gaming, so why would you choose to spend £50 on a separate card that will do the same job?

That said, times move quickly in the graphics card market, and tomorrow's £50 GPU will make processor graphics weep. AMD and Nvidia will be launching a slew of low-end cards to prop up their latest HD 6xxx and 5xx series respectively.

The high end will probably see the biggest battle. Nvidia's GTX 580 is currently top dog, but AMD is due to release its dual-GPU Antilles behemoth in the next few months, possibly at the CeBIT show in Germany. Details are scarce, but if AMD follows the example set by its previous dual-GPU releases, you can expect two Cayman Pro GPUs wired into one slice of AMD-red PCB.

HD 6950

Those are the chips powering the superlative Radeon HD 6950, and will make for one hell of a card.

Don't expect Nvidia to be keeping quiet, though. When we spoke with Tom Petersen, the company's Director of Technical Marketing, at the secretive preview of the GTX 580 last year, we asked if he expected to see a dual-GPU Fermi card any time soon.

He explained that, now the thermal issues seen in the first high-end Fermi card (the GTX 480) had been solved in the GTX 580 and GTX 570, there really wasn't a barrier any more.

So pretend to be surprised when Nvidia announces a GTX 595 just as AMD starts to get excited about its Antilles card.

Three top graphics card choices

Zotac GTX 580 AMP
Performance
Price: £480
Info: www.zotac.com

Zotac gtx 580

On the basis that money is no object in your search for graphics perfection, you'll be hard-pressed to find a more impressive pixel-pusher than Zotac's recently launched, overclocked GTX 580 AMP.

This souped-up version of Nvidia's GPU is the fastest thing on two PCIe power cables. Based almost entirely on the first Fermi card, the GTX 480, it's undoubtedly what the brand wanted to release originally.

The GTX 480 had a cutdown version of its low-yielding GF 100, with one streaming microprocessor turned off. That meant a lowly 480 CUDA cores instead of the full 512 we were expecting.

The GTX 580 came out of nowhere last year with the full complement, plus nifty power and cooling advances. So it's quicker, cooler, quieter and far more power-efficient. In short, it's just better.

The AMP version is ever so slightly overclocked, but will also give a little more headroom should you wish to push it further. At these speeds though, you won't need to for a few years at least.

Verdict: 4.5/5

Asus GTX 460 Top 768MB
Budget
Price: £130
Info: www.asus.com

Asus gtx 460

We've already seen the stock GTX 460 768MB, and now it's the turn of the overclocked cards in the shape of Asus' GTX 460 768MB TOP edition.

The GTX 460 looks set to be the most successful iteration of the Fermi architecture that Nvidia has released to date. That's mainly thanks to a redesigned chip, still based on the same technology that made the GTX 480 such a blisteringly fast, and hot, card.

This new GF104 GPU is a far more streamlined chip compared to the fairly bestial GF100.

It still has the same basic premise running through it, but more cores have been squeezed into fewer streaming microprocessors (SMs) and more texture and special funtion units have been jammed in there too.

Verdict: 3.5/5

Read the full Asus GTX 460 Top 768MB review

Sapphire Radeon HD 6950
All round
Price: £228
Info: www.sapphiretech.com

HD 6950

AMD's Radeon HD 6950 is the must-have card of the moment, its price tag hitting the sweet spot in terms of cost/performance ratios.

The card is based on AMD's latest Cayman GPU, and with its redesigned approach to tessellation, offers some serious competition for the far more expensive GTX 570. It's also the only card under £250 that can take on the tessellation-heavy Metro 2033 at an eye-bleeding 2,560 x 1,600 resolution and still come out smiling.

The Cayman GPU's twin tessellation engines make the HD 6950 an excellent DirectX 11 card. On the DX 10 benchmarks it loses ground to the new GTX 560 Ti from Nvidia, but the AMD card has the better scores in the newest titles and comes with an impressive trick up its sleeve.

With a simple BIOS flash you can upgrade your HD 6950 and turn it into an HD 6970 – a £270 card – for free. That's not an overclock; it's unlocking dormant parts of the GPU and setting them free. That makes it the card of choice right now.

Verdict: 5/5

Nvidia

With any sort of gaming, the biggest performance boost you can get is by making a change to your graphics card. However, that doesn't mean you need to kick your existing card into touch and dip your hand into your wallet to fund a quicker one.

Overclocking your graphics card is the quickest, cheapest and (most likely) easiest way to wring a little extra performance out of your current GPU. It was originally seen as a way to make older components competitive for longer, whereas it's now an expected consideration when you're choosing any new graphics card.

It's also not as dangerous as it once was; it's fairly difficult to brick a GPU with the simple overclocking you can do in Windows.

Generally speaking, you'll be lucky to get a 10 per cent increase in performance, but those precious few extra frames per second can mean the difference between almost unplayable choppiness and smooth, slick action.

To get started, all you need is a software tool like MSI's Afterburner, a graphics stress test such as Heaven 2.1 and a bit of patience. With a few cards, including the Radeon HD 6950 we've already talked about, it's possible to make some BIOS tweaks to garner a boost in performance.

The HD 6950, for example, runs the same GPU as the HD 6970. All that's different is that the chip was chosen to go in the lower-price card and has some features turned off. Normally, these changes are hard-wired out of the GPU, but AMD decided to implement the castration with software pincers only, making for easy reversal.

This is an extremely rare state of affairs though, and we haven't seen similar things happen in the GPU game for years.

Another way to improve things is by taking the more heavy-handed approach of adding a second GPU to the equation. Multi-GPU graphics have come on leaps and bounds in the last couple of years. Now we're seeing performance with a second card hit the 2x boost we'd always hoped for – and even more in some cases.

The difficulty is with the motherboard. You need one that supports multi-GPU and, unfortunately, that's a tougher, more expensive journey if you want to go down the Nvidia SLI route.

However, most boards will support AMD's CrossFire technology as standard on extra PCIe lanes, so that's becoming far more popular.

Easy overclocking

1. The setup

steo 1

MSI's Afterburner has an impressive hardware monitoring display. Detach this from the main console and it will give you a better view.

Start up the Heaven 2.1 benchmark in Windowed mode, at a lower resolution than your desktop, but with high settings. This will stress your GPU as you apply the overclock to help you judge stability. If there are no graphical glitches on-screen, the OC is stable.

2. The push

step 2

You'll need to approach the memory and GPU overclocking of your card separately to judge the maximum overclock on each. Start with the memory and push the slider up in 5MHz increments.

After each step, watch the Heaven benchmark closely for artefacts. Keep doing this until you see problems, then step back the OC until they go. Reset the memory slider. Repeat for the processor clock.

3. The stress

step 3

Once you've discovered the top overclock for the GPU and memory clocks, set both sliders to their maximums and watch the benchmark. You'll have to knock back the relevant sliders to create a stable OC.

Pixel-sized artefacts mean there's a memory problem, and bright colour blocks represent GPU issues. When there are no artefacts, your OC is stable. Now stress test it in fullscreen mode to make sure.



Buying Guide: Processor upgrade: how to choose the right CPU

Upgrading your PC's processor isn't the easiest or cheapest way to boost its performance, but it can deliver dramatic results. The right chip can transform a sluggish machine into a powerhouse.

The practicalities of fitting a processor have remained largely unchanged since the early days of PCs, but selecting a new chip isn't so simple. Once it was just a matter of picking the one with the highest GHz rating. Now you've got to keep that in mind, along with an increasingly complex set of other variables.

First, you need to look at the CPU socket on the motherboard to see whether you need an Intel or AMD processor. Next, you need to consider the processor itself. Each manufacturer will have a slew of subtly different technologies, all packaged inside chips with names that tell you little about them.

When you've chosen a chip type, you'll need to weigh up how many processor cores you need. Should you pick fewer faster cores or go for more cores that run at a comparatively slow speed? Then there are nuances like automatic turbo modes and hypertheading to consider.

Finally, you need to think about the interplay between your CPU and graphics processing unit.

Lets consider one factor in detail: the number of cores. As perverse as it sounds, more cores don't necessarily mean more speed. Speed is largely determined by the applications you're running, and how they use those cores.

Thankfully, many video editing suites and games are now written to use multiple cores simultaneously. The same applications (including Photoshop, Adobe Premiere and HD video players) can take advantage of GPU acceleration. You therefore need to think about your system holistically, considering your requirements.

Video-editing professionals, for example, might opt for a dual-core MacBook Pro, with a GeForce chip for encoding HD streams with Final Cut Pro. This would do a better job than a PC with a quad-core CPU and a non-programmable graphics card.

There are other factors to consider, too. For example, Intel's new Sandy Bridge CPUs have an astonishingly good Quick Sync video encoding engine, which is often better than GPU acceleration.

We'll guide you through the business of picking the best chip. It might not be an easy decision, but with the array of new silicon, you're sure to find one that's a good fit for you and your system.

The great processor war heats up

The battle between AMD and Intel has been raging for decades, but it feels like Intel may have gained the upper hand. Even the billion-dollar recall of its new 6 Series motherboards failed to upset the launch of its Sandy Bridge processors - the week the recall was announced, Intel's share price actually rose. Such is the market's confidence in Intel's silicon.

AMD has been trailing in raw performance since the launch of Intel's Core 2, which blew the Athlon 64 out of the water. Throwing cores at the problem hasn't helped either - six-core Phenom II chips can't keep pace with their Intel quad-core rivals.

Don't write AMD off just yet, though. This summer we're expecting to see AMD processors based on a completely new design, currently called Bulldozer, which could pose a real threat to Intel's dominance.

Bulldozer's design looks very advanced and efficient. AMD has yet to confirm which chips it's planning to launch, but the internal structure is very different from what's gone before.

The basic unit within Bulldozer is a dual-core module, which can share resources across its two execution engines or divide them up for increased parallelism. The first Bulldozer chips, codenamed Zambezi, will be come in four, six or eight-core variants, which suggests they'll build on AMD's current strategy of offering more cores for less than Intel.

The biggest problem AMD faces will be branding. It's already begun its Fusion project, which - like Intel's Sandy Bridge chips - integrates a graphics core onto the same die as a CPU. Bulldozer, however, won't be part of the Fusion line-up for over a year.

AMD's first Fusion processors have been built around the low-end Bobcat core, which is aimed at netbooks and thin and light laptops - somewhere above an Atom, but below a Core i3.

The first Fusion chips for desktop use will be based on Phenom II, and will launch around the same time as Bulldozer. That gives AMD the problem of trying to communicate the relative strengths and weaknesses of many different CPU types at the same time.

By comparison, Intel has already begun retiring many of its Nehalem-based Core processors, giving it two main platforms - Atom and Core - which are relatively easy to understand. Pentium and Celeron will continue to exist as value brands, but even these will use Sandy Bridge designs and the new socket 1155 motherboards.

Only at Intel's extreme high end, where six-core CPUs based on the older Gulftown design are still being released, is there any inconsistency - and these are already starting to look irrelevant in the wake of Sandy Bridge's performance (see the 980x review).

By the time the new Bulldozer Fusion chips arrive, Intel should have completed its plans to migrate Sandy Bridge onto a 22nm process. With that sort of performance increase and cost reduction, AMD could be left trailing behind.

12 CPU choices

Intel processors: The high end

Intel high end

Intel Core i7 980X
Performance
Price: £750

Want the finest CPU money can buy? Look no further than the Intel Core i7 980X. It's the undisputed world heavyweight champ among chips. Just remember you'll need a lot of it. The money, that is.

But what a processor you get in return. The Core I7 980X is, of course, a six-core beast. But this is six-cores Intel style, so that's two threads per core and a dozen of those little green graphs when you fire up task manager. That's unparalleled, er, parallelism in a PC processor.

In that context, even AMD's Phenom II X6 1090T Black Edition looks rather ordinary. It may have six cores. But each is single-threaded only. Moreover, AMD's underlying CPU architecture is pretty ancient while the Intel Core i7 980X is literally the latest thing, right down to is impossibly tiny 32nm underpinnings. What it isn't however, is unique. The recently released Core i7 970 is largely the same six-core, 12-thread processor at a slightly less offensive price point.

Read the full Intel Core i7 980X review

Verdict: 4.5/5

Intel Core i5 2500K
Budget
Price: £190

Ready or not, here they come. Intel is rolling out a thoroughly overhauled range of PC processors based on its new Sandy Bridge microarchitecture. Our first taste of the new chips comes in the form of the Intel Core i5-2500K and Intel Core i7-2600K desktop CPUs.

Thanks to the baffling array of chips, sockets and brands, we've barely got to grips with Intel's existing CPU range. Certainly Intel's main rival, AMD, has no answer in outright performance terms to the chips Intel already offers, but the relentless march of technology must go on.

So, ignore the familiar Core i5 and Core i7 branding. These are all new processors and they're ready to roll.

As it happens, Intel could actually do with more powerful and, crucially, more power efficient processors for laptop PCs. Deep down, that's what Sandy Bridge is really about. However, as we'll learn, Sandy Bridge has a lot to offer for the desktop, too, including exciting new features such as a hardware video transcoding engine and much-improved integrated graphics.

Read the full Intel Core i5 2500K review

Verdict: 4.5/5

Intel Core i7 2600S
All Rounder
Price: £235

Another day another new Sandy Bridge CPU, and another new suffix to get your head around. This time it's the Intel Core i7 2600S.

Many people have heard of the Core i7 2600K by now, the unlocked overclocking demon that was part of the original Sandy Bridge launch, but the Core i7 2600S you may not have heard of.

It's also pretty well known that if a second-gen Core CPU doesn't have a K at the end of the model number, its pretty much game over for any sort of serious overclocking. So what does the S stand for, and does it mean even more features turned off?

Well no, the S models represent more low power units, but just not as low powered as the T models. So instead of the 95W of a standard 2600 part, they have a TDP of 65W, as compared to the 45W of the T series.

Apart from the low power rating, the Intel Core i7 2600S still retains all the familiar features of the 2600 family: four cores, eight threads and 8MB of Smart Cache. But as with all the S class chips, it's clocked slower than the rest of the their family.

In the case of the Core i7 2600S, this means it runs at 2.8GHz compared to the 3.4GHz (3.8GHz max Turbo) of the i7 2600 and Intel Core i7 2600K.

Read the full Intel Core i7 2600S review

Verdict: 4/5

Intel processors: The mid-range

Intel mid range

Intel Core i5 2300
Performance
Price: £150

This is the lowest specced of Intel's new Core i5 series, but it only costs a little less than the faster 2500. That said, there's not a lot of difference in performance either unless you're planning to overclock, so you may as well opt for this slightly cheaper model and save a little cash.

This quad-core Sandy Bridge CPU has a base clockspeed of 2.8GHz, and can accelerate to 3.1GHz under Turbo mode. Like all i5s, it doesn't have Hyperthreading enabled, but it's still more than a match for all but the very fastest of AMD's six-core Phenom IIs in everything except demanding video encoding applications.

Even though it's Intel's least expensive quad-core, the i5 2300 still isn't cheap. Factor in the price of an H67 motherboard, and it highlights a significant problem for a lot of the current Core line-up: there are a lot of three, four and even six-core AMD processors available for less, and if you're on a really tight budget there's every reason to be swayed by them.

Verdict: 4/5

Pentium G6950
Budget
Price: £75

This the cheapest CPU currently available from Intel, costing just £75, but it still offers an impressive balance of performance and value.

Its low price can be attributed to the fact that it's not a new chip. Launched last year and based on the Clarkdale architecture, the Pentium Dual Core G6950 is almost identical to the outgoing Core i3s in that it requires a Socket 1156 motherboard, has an on-board graphics accelerator (albeit a low power one) and is designed around the superb Nehalem core.

With a base clock of 2.8GHz and no Turbo Boost or Hyperthreading abilities, the G6950 can't keep up with its Core i3 brethren, but it's proved itself to be a top overclocker in the past.

Ultimately though, if you're looking at spending less than £100 on a new CPU, you'll probably be more attracted to the even cheaper Athlon IIs. They aren't as fast as the G6950, but they'll get the job done perfectly well for less.

Verdict: 3/5

Intel Core i3 2100
All Rounder
Price: £108

After all the Sandy Bridge goodness without the quad-core price-tag? Then the dual-core Intel Core i3 2100 might well be up your street.

While the high-end unlocked Sandy Bridge CPUs, the Intel Core i7 2600K and Intel Core i5 2500K were rightfully taking all the plaudits for being overclocking monsters, the 2600K especially, not many people were looking at the other end of the food chain.

That is to say in the value end of the market where the lowly Intel Core i3 2100 is to be found. As with all the current Sandy Bridge processors, it's built on the 32nm process and manages to pack 504 million transistors into its die.

The Core i3 2100 is clocked at 3.1GHz with 3MB of L3 cache, which sounds like it should be a fairly blazing chip. However it has no Turbo Boost and is totally locked down, so there's no overclocking fun available on the processing side.

This is a pity, because some of the best overclockable Intel chips in the past have come from this segment of the market.

You may not be able to overclock the CPU core but you can though do a smidgen of tinkering to the HD2000 graphics core integrated into this second generation Core CPU.

Read the full Intel Core i3 2100 review

Verdict: 3/5

AMD processors: The high-end

AMD high-end

AMD Phenom II X6 1100T BE
Performance
Price: £208

This new hex-core chip is symptomatic of AMD's current predicament: that its most expensive PC processors sell for barely one-third the price of Intel's.

And that's not a situation the new AMD Phenom II X6 1100T Black Edition is going to change – despite the fact that it's officially AMD's fastest and most expensive chip.

With a retail sticker around £215, the AMD Phenom II X6 1100T Black Edition is priced on a par with the very cheapest of Intel's Core i7 processors such as the Intel Core i7 870. Nevertheless, it serves up six execution cores to the 870's four. Six-core Intel CPUs are far more expensive, starting around £700.

Then again, Intel's cores do much more work per cycle. Until it releases the long awaited Bulldozer CPU architecture, AMD needs to sell more cores for less cash. In the meantime, this revised six-core Phenom II X6 raises AMD's game incrementally with an increase in clockspeed from 3.2GHz to 3.3GHz.

For the most part, the AMD Phenom II X6 1100T Black Edition is not a new processor. It's based on AMD's increasingly familiar six-core Thuban die, a chip that can trace its roots directly back to the AMD Hammer CPU architecture first seen in 2003. Thus, it's a 45nm CPU with 512k cache per core and a further 6MB of shared cache memory.

Read the full AMD Phenom II X6 1100T BE review

Verdict: 4.5/5

Phenom II X4 970BE
Budget
Price: £140

By now, we were hoping to able to get our hands on AMD's new Bulldozer archtecture processors, but there's still no sign of them, so instead we back with the familiar Phenom II theme. Give it up for the new AMD Phenom II X4 970 Black Edition.

Based on the 45nm Deneb core that's being doing duty in quad-core Phenom II chips for nearly two years, the AMD Phenom II X4 970 Black Edition sports a heady stock clockspeed of 3.5GHz. Yup, that's precisely 100MHz or three per cent faster than the Phenom II X4 965 Black Edition, previously the fastest X4 model.

This kind of incremental upgrade is symptomatic of an ageing processor architecture on it last legs. But while Bulldozer can't come soon enough, there's no reason why the 970 shouldn't be an attractive CPU. It all comes down to pricing and positioning. At £140, the 970 squares up directly against Intel's Core i5 760. Game on.

Read the full AMD Phenom II X4 970 Black Edition review

Verdict: 4/5

Phenom II X6 1055T
All rounder
Price: £140

Remember when AMD launched its first quad-core processor in 2007? We can, because back then it seemed like AMD was months from keeling over stone dead. Today, the company is in much finer fettle. For proof, look no further than the new AMD Phenom II X6 1055T.

Somehow, AMD has managed to produce a six-core PC processor and sell it for just over £10 more than its best quad-core chip, the Phenom II X4 965 Black Edition. It's actually cheaper than several Intel quads. Inspect the detail specifications and the Phenom II X6 1055T only gets more impressive.

The transistor count has grown from 758 million transistors to 904 million. And yet the smaller quad-core 965 is rated at 140 Watts while this new six-core 1055T is a 125 Watt chip. Of course, at 2.8GHz, the 1055T is clocked quite a bit lower than the 3.4GHz 965 BE. But AMD has clearly done something right.

Read the full AMD Phenom II X6 1055T review

Verdict: 5/5

AMD processors: The mid-range

AMD mid range

Phenom II X4 955BE
Performance
Price: £107

When AMD rolled out its chips for the new AM3 CPU socket, we were baffled. AM3 dragged AMD into the DDR3 era. But with that in mind, why were the first AM3-compatible CPUs groveling models with cut-down caches, lower frequencies or fewer cores?

Surely the new socket deserved a brand new processor to showcase its beefed up bandwidth and multi-core majesty? Finally that chip has arrived in the form of the new Phenom II X4 955 Black Edition.

For those of you without a double first in cryptography, allow us to decode the ludicrous product name. For starters, it's a Phenom II chip and that means cool-running, high-clocking 45nm silicon and hence not AMD's utterly rubbish 65nm transistors. X4, of course, indicates four cores, which is as good as it currently gets.

Read the full Phenom II X4 955 Black Edition review

Verdict: 4/5

Athlon II X2 260
Budget
Price: £50

Want to build a cheap media server? You may have been tempted by one those bare-bones net-tops with a dual-core Atom processor and an Nvidia video card, but why would you go for one of those when it's easy to build a system around this chip for less than £250?

Unlike the Atom, the dual core Athlon II is a 'proper' processor that supports out-of-order execution. It'll handle most games and multitasking, and doesn't suffer from the same grinding slowdown as an Atom.

The fact that it isn't as power-efficient as Intel's mobile marvel isn't such a concern if you're after a low-end desktop. In fact, the difference in overall performance between this and the Phenom II X2 565BE won't be tangible to most people.

It sits between five and ten frames per second behind its stablemate in games, and seconds (rather than minutes) behind in video encoding. It's one of the best ways to build a silent, low-cost second system for simple tasks. The question is, are you sure you want to?

Verdict: 4/5

Phenom II X2 565BE
All Rounder
Price: £90

Whatever happened to the triple core X3 processors that AMD was hoping would win over the sub- £100 crowd? They seem to have sold out or vanished from most of our favourite online retail stores, leaving us with this twin-cored chip as AMD's best performer at that price.

The X2 565BE certainly looks like good value for money. A high 3.4GHz clockspeed and unlocked multiplier make it a solid performer for older games. The problem is that, overall, it doesn't convince us that it's worth saving the £17 compared to the quad core X4 955BE, which is almost twice as fast for video encoding and has the framerate advantage in the increasing number of well threaded games.

Intel's lowest priced retail CPU, the Pentium G6950, will turn in better performance for less too. If you're after some real bargain silicon - for a silently-running second machine, for example - then you should be looking at the cheaper Athlon IIs instead. The 565BE is likely to be a disappointment.

Verdict: 2/5

How to keep your CPU cool

You could be forgiven for thinking that a processor upgrade is the most complex system improvement you can make. Sure, PC open-heart surgery can be nerve-wracking, but when everything is in place modern motherboards work in concert with processors to make the business of configuration very easy.

Indeed, everything is pretty much automatic. If you do run into problems, your chip may be overheating. This is increasingly likely if you're re-using your existing cooler. Read on as we detail the correct procedure for keeping new and existing chips cool.

Optimise your cooling

1. Check current temperatures

step 1

Before you start taking anything out of your machine, you need to check how hot your processor is running. CPU Hardware Monitor provides detailed temperature information for your CPU cores and motherboard, and is free.

Take a reading with nothing else running (idle) and then when another running Super Pi (under load). Then check these figures against your CPU's datasheet.

2. Remove your existing fan

step 2

Removing processor fans is easy. For Intel components, unlock the four holding pins by turning them clockwise; you should then able to pull the whole thing out.

The fans that ship with standard AMD processors are also simple to remove - just lift and release the holding arm and then gently ease the cooler out. Don't forget to unplug the fan power cable before removing the cooler completely.

3. Wipe off the thermal grease

step 3

Before you put the cooler down on your workbench, clean off the existing thermal grease using a tissue and possibly a little solvent.

You will need some to add more thermal grease later though, so if you don't have any new stuff to hand, it's best to place the cooler on a piece of paper without cleaning it so that no impurities are mixed into the existing grease. Remove the CPU and clean the grease off that too.

4. Clean the blades

step 4

Use a piece of folded tissue or a soft brush to clean along the line of the fan blade. Don't push too hard, though, as these can break off under pressure. Dust tends to accumulate on the top of the fan blades, so it shouldn't be too hard to clean away.

If your fan is in a housing, take it apart if needed so that you can get better access to the heatsink and the fan.

5. Add new thermal grease

step 5

Put the CPU back into its socket and apply a new layer of thermal grease to the top of the processor. Remember, your aim is to have a very fine layer to fill in any gaps between the CPU and heatsink, not to make a grease sandwich. A small blob should be sufficient.

If you didn't remove the old thermal grease, check how much is on the processor and remove any if possible.

6. Reseat and retest

step 6

Put the cooler back in its home and wiggle it in place a little to make sure that the thermal grease is spread evenly between the cooler and CPU.

Next, clip the retaining arms or pins back into place and reconnect the fan to your system. Boot your machine and run the same tests that you did originally. If your CPU is still running too hot then a new cooler is needed.



TOP PRODUCTS

Related Posts Plugin for WordPress, Blogger...
Design by ROCKY| computer hardware by ROCKY