Mashed Articles

Writen by Jason Kohrs

Hard Drives: ATA versus SATA

The performance of computer systems has been steadily increasing as faster processors, memory, and video cards are continuously being developed. The one key component that is often neglected when looking at improving the performance of a computer system is the hard drive. Hard drive manufacturers have been constantly evolving the basic hard drive used in modern computer systems for the last 25 years, and the last few years have seen some exciting developments from faster spindle speeds, larger caches, better reliability, and increased data transmission speeds.

The drive type used most in consumer grade computers is the hearty ATA type drive (commonly called an IDE drive). The ATA standard dates back to 1986 and is based on a 16-bit parallel interface has undergone many evolutions since its introduction to increase the speed and size of the drives that it can support. The latest standard is ATA-7 (first introduced in 2001 by the T13 Technical Committee (the group responsible for the ATA standard)) which supports data transfer rates up to 133MB/sec. This is expected to be the last update for the parallel ATA standard.

As long ago as 2000 it was seen that the parallel ATA standard was maxing out its limitations as to what it could handle. With data rates hitting the 133MB/sec mark on a parallel cable, you are inviting all sorts of problems because of signal timing, EMI (electromagnetic interference) and other data integrity issues; thus industry leaders got together and came up with a new standard known as Serial ATA (SATA). SATA has only been around a few years, but is destined to become “the standard” due to several benefits to be addressed in this Tech Tip.

The two technologies that we will be looking at are: ATA (Advanced Technology Attachment) – a 16-bit parallel interface used for controlling computer drives. Introduced in 1986, it has undergone many evolutions in the last 18+ years, with the latest version being called ATA-7. Wherever an item is referred to as being an ATA device, it is commonly a Parallel ATA device. ATA devices are also commonly called IDE, EIDE, Ultra-ATA, Ultra-DMA, ATAPI, PATA, etc. (each of these acronyms actually do refer to very specific items, but are commonly interchanged) SATA (Serial Advanced Technology Attachment) – a 1-bit serial evolution of the Parallel ATA physical storage interface.

Basic Features & Connections

SATA drives are easy to distinguish from their ATA cousins by the different data and power connections found on the back of the drives. A side-by-side comparison of the two interfaces can be seen in this PDF from Maxtor, and the following covers many of the differences…

Standard ATA drives, such as this 200GB Western Digital model, have somewhat bulky, two inch wide ribbon cable with 40-pin data connections and receive the 5V necessary to power them from the familiar 4-pin connection. The basic data cables for these drives have looked the same for years. A change was made with the introduction of the ATA-5 standard to better improve the signal quality by making an 80 wire cable used on the 40-pin connector (these are commonly called 40-pin/80-wire cables). To improve airflow within the computer system some manufacturers resorted to literally folding over the ribbon cable and taping it into that position. Another recent physical change also came with the advent of rounded cables. The performance of the rounded cables is equal to that of the flat ribbon, but many prefer the improved system air flow afforded, ease of wire management, and cooler appearance that come with them.

SATA drives, such as this 120GB Western Digital model, have a half inch wide, 7 “blade and beam” data connection, which results in a much thinner and easier to manage data cable. These cables take the convenience of the ATA rounded cables to the next level by being even narrower, more flexible and capable of being longer without fear of data loss. SATA cables have a maximum length of 1 meter (39.37 inches), which is much greater than the recommended 18 inch cable for ATA drives. The reduced footprint of SATA data connections frees up space on motherboards, potentially allowing for more convenient layouts and room for more onboard features!

A 15-pin power connection delivers the 250mV of necessary power to SATA drives. 15-pins for a SATA device sounds like it would require a much larger power cable than a 4-pin ATA device, but in reality the two power connectors are just about the same height. For the time being, many SATA drives are also coming with a legacy 4-pin power connector for convenience.

Many modern motherboards, such as this Chaintech motherboard, come with SATA drive connections onboard (many also including the ATA connectors as well for legacy drive compatibility), and new power supplies, such as this Ultra X-Connect, generally feature a few of the necessary 15-pin power connections, making it easy to use these drives on new systems. Older systems can easily be upgraded to support SATA drives by use of adapters, such as this PCI slot SATA controller and this 4-pin to 15-pin SATA power adapter.

Optical drives are also becoming more readily available with SATA connections. Drives such as the Plextor PX-712SA take advantage of the new interface, although the performance will not be any greater than a comparable optical drive with an ATA connection.

Performance

In addition to being more convenient to install and drawing less power, SATA drives have performance benefits that really set them apart from ATA drives.

The most interesting performance feature of SATA is the maximum bandwidth possible. As we have noted, the evolution of ATA drives has seen the data transfer rate reach its maximum at 133 MB/second, where the current SATA standard provides data transfers of up to 150 MB/second. The overall performance increase of SATA over ATA can currently be expected to be up to 5% (according to Seagate), but improvements in SATA technology will surely improve on that.

The future of SATA holds great things for those wanting even more speed, as drives with 300 MB/second transfer rates (SATA II) will be readily available in 2005, and by 2008 speeds of up to 600 MB/second can be expected. Those speeds are incredible, and are hard to imagine at this point.

Another performance benefit found on SATA drives is their built-in hot-swap capabilities. SATA drives can be brought on and offline without shutting down the computer system, providing a serious benefit to those who can’t afford downtime, or who want to move drives in and out of operation quickly. The higher number of wires in the power connection is partially explained by this, as six of the fifteen wires are dedicated to allowing the hot-swap feature.

Price

Comparing ATA drives to SATA drives can be tricky given all of the variables, but in general it is the case that SATA drives will still cost just a bit more than a comparable ATA drive. The gap is closing rapidly though, and as SATA drives gain in popularity and availability a distinct shift in prices can be expected. Considering the benefits of SATA over ATA, the potential difference of a few dollars can easily be justified when considering an upgrade. Computer Geeks currently has a limited selection of SATA drives, but several technical sites, such as The Tech Zone and The Tech Lounge, offer real time price guides to see how comparable drives stack up.

Final Words

The current SATA standard provides significant benefits over ATA in terms of convenience, power consumption and, most importantly, performance. The main thing ATA has going for it right now is history, as it has been the standard for so long that it will not likely disappear any time soon. The future of SATA will be even more interesting as speed increases will help hard drive development keep pace with other key system components.

Jason Kohrs Computer Geeks tech tips and computer advice

Tags: acronym, API, caches, Computer, computers, Development, device, geek, integrity, interface, iss, memory, performance, power consumption, processor, rms, sap, space, stack, storage, system, Technology, type, XP

Writen by Syed Feroz Zainvi

The article summarizes the Hyperthreading technology as best as possible in just ten steps.

1. Hyperthreading (HT) is term coined by Intel for Simultaneous Multi-Threading (SMT).

2. Hyperthreading is meant for dual-core or multi-core processors like Pentium 4 with HT or Xeon MP.

3. The technology aims at better utilization of execution units in a processor which otherwise would be idle.

4. This technique is better than its predecessor Multithreading and Superthreading in the sense that in each instruction cycle, a processor can execute instructions from mutliple threads.

5. Hardware implemetation of Hyperthreading is not very complex as reported by Intel “Implementing HT took only 5% more die space in Xeon”.

6. Implemeting HT needs hardware resources to be divided into Replicated(e.g. Instruction Pointers), Partitioned (e.g.Queues) and Shared (e.g. Caches)

7. For actual performance gain, your applications have to hyperthreaded i.e. you have to parallelize your code.

8. Hyperthreaded applications use Shared Memory Programming Model. Most famous Shared Memory Programming standard is OpenMP (Open specifications for Multi-Processing), an effort initiated by Silicon Graphics Inc. and now supported by major industry players. OpenMP library is available for C,C++ and Fortran on both Windows and Unix platforms. OpenMP programming constructs consists of few compiler directives for parallelizing the code.

9. There are compilers and other automating tools that support shared memory parallel programming libraries like OpenMP. Some examples are Intel C++ Compiler, Intel VTune Performance Analyzer and High-Level tools from KAI Inc.

10. Hyperthreading does not necessarily increase performance. With wrong mix of code, it can even decrease performance.

Syed Feroz Zainvi has obtained M.Tech. (Comp Sc & Engg) degree from Indian Institute of Technology, Delhi (INDIA). His areas of interests are distributed computing, computer graphics and Internet Technologies. Currently, he is involved in Software Project Planning, Development and Management. His other interests include writing for magazines and contributing utility softwares on Magazine’s CDs. He also have flair for teaching computer science with new teaching methodologies.

His web-page URL is http://www.zainvi.tophnors.com

Tags: application, caches, Computer, Computing, Development, Hardware, ims, Internet, memory, methodologies, partition, performance, processor, rms, Science, silicon, smt, Software, space, Technology, tools, unix, web, Windows

At the risk of offending all those who love to talk for hours about cores, caches and clock speeds, I have to say that I personally find discussions about the innards of silicon chips and how they are wired together intensely boring. In fact, I’ve probably already used all the wrong words and phrases, even in that first sentence, which is no doubt going to annoy some people further.

So, when Tony, Martin and I were invited to a dinner to meet with some of AMD’s European executives, I was understandably in two minds about attending, especially as I am also not really into all this wining and dining stuff as some other analyst are.

I went along, though, and I’m glad I did. Sure, I found myself sucked into the odd eye glazing conversation that I only partially understood, but something that came across clearly was that AMD is investing quite a bit in ‘reaching through’ relationships with its direct customers (largely the OEMs) to the ultimate customers – Enterprises, SMBs and consumers.

Of course there is nothing new or unique in this, in fact I ran a team at Nortel Networks back in the early 00’s which did exactly the same thing (in that case, reaching through the mobile operators to understand how 3G related to their subscribers). The basic idea is that you can gain insights and tune your R&D based on direct end user/buyer input that would not be possible if you worked second hand through your customer as an intermediary. To do this well, however, you really need people who understand that end user environment and the trends that are taking place within it, and that’s not necessarily the same people that deal with your core product design from an internal perspective.

Anyway, this end-user oriented view of the world shifted discussions to more familiar territory for me during the dinner, and I enjoyed hearing people like Giuseppe Amato, who goes under the title “Director, Value Proposition Team”, explaining how the whole process works in relation to data centre evolution, high performance computing and mobile working. It changed my perception of AMD quite a bit from simply “the alternative to Intel” to that of an independent player that is committed to driving industry development in its own way.

While I am not qualified to comment on the relative merits of AMD technology versus the competition, nor its ability to execute in the cut throat world of OEM deals and supply chains, I now have a much better appreciation of why what AMD does actually matters. It is not just about price/performance or performance per watt of energy consumed, it is about shifting thresholds to make things economically or practically possible in the mainstream market that previously were not. That’s why the “what if you could….?” conversations with end customers as suppliers like AMD reach through to them are so important. And also why, for the first time in my life, I actually had some genuinely interesting conversations about silicon that were directly relevant to the world in which I live.

Tags: amd, caches, Computing, consumers, conversations, cores, Development, energy, Environment, internal perspective, mobile operators, network, nortel networks, oems, performance, Personal, relationship, relationships, silicon, silicon chips, smbs, Technology, XP