Since Intel has recently launched its new i815E and i820E chipsets supporting ATA/100, the world turned to this interface. A lot of companies launch controller cards supporting ATA/100 protocol, such as HighPoint and Promise, for instance. Besides, almost all harddisk manufacturers are already offering their models with the new interface support as well. So, the interest in this new interface standard is constantly growing and that's why we decided to take a look at ATA/100 major peculiarities and features.

Before we actually start, we would like to offer you a brief tour into the history of data transfer interfaces and their development.

All of you have a thousand times come across the three letters: IDE. IDE is an abbreviation of either Intelligent Drive Electronics or Integrated Drive Electronics, depending on who you ask. An IDE interface is an interface for mass storage devices, in which the controller is integrated into the hard disk drive or CD-ROM drive. Although it really refers to a general technology, most people use the term to refer to the ATA (Advanced Technology Attachment) specification, which uses this technology. This is one of the oldest PC standards developed in 1989 by three companies: Imprimus (a division of Control Data Corporation), Western Digital and Compaq. Compaq badly needed a low-cost solution to connect the HDDs to their PCs, Imprimus was a large hard disk drive manufacturer and Western Digital had been dealing with HDD controller chips since 1984 (for IBM).

Although Imprimus didn't manufacture 3.5" hard drives Compaq was looking for, the latter had to order them from Conner Peripherals. The beginning of Conner HDDs retailing can be regarded as the first steps of ATA interface towards its popularity. In 1989 CDC sold Imprimus to Seagate and another 6 years later Seagate purchased Conner Peripherals. By that time Western Digital had also achieved significant success having become a large hard disk drive manufacturer. As for other smaller manufacturers, they little by little began joining the HDD manufacturers offering their new products as well.

The first standard, ATA, is also known as IDE. It was introduced in 1994 by NCITS (National Committee for Informational Technology Standards) and to be more precise, by its technical group, T13, organized especially for ATA. This protocol supported one or two hard drives and 16-bit interface. In 1999 it was removed from ANSI list upon T13's recommendation.

However, ATA interface still managed to cope with its major task: all IDE hard disks got a uniting feature. If you are old enough, you should remember that very often in the beginning of 1990s a combination of two IDE hard drives from different manufacturers could work properly only if one of them was defined as a master device and another as a slave. And every time you tried to swap the roles the whole system simply refused to function at all. Certainly, the single official standard helped to solve the problem of the HDD models incompatibility.

Two years later, in 1996, ATA interface was enhanced and acquired new faster working speeds. In other words, the world saw a new ATA-2 interface (AT Attachment with Extensions). Among these extensions we should mention logical block addressing (LBA) and block transfers (a block of requests generating only one IRQ). Besides, the system learned to identify the hard disk parameters much better than it used to. It supported faster PIO modes (3 and 4) and multiword DMA modes (1 and 2).

And then the most interesting things started happening. The PC software of those times was intended for HDDs with the capacity of 528MB at the most. Western Digital introduced their Enhanced BIOS specification, which allowed surpassing this obstacle and added another couple of new features, such as two ATA ports support. A combination of Enhanced BIOS and ATA-2 was called Enhanced IDE (or simply EIDE). So, strictly speaking, IDE hard drives should be only those from Western Digital. Though ATA-2 was ratified as a standard only in 1996, the spec had a much longer life. This mass storage device interface standard supported data rates of between 4 and 16.6MB/sec, about three to four times faster than the old IDE standard. In addition, it could support mass storage devices of up to 8.4GB, whereas the old standard was limited to 528MB. Because of its lower cost, enhanced EIDE replaced SCSI in many areas.

The reaction of Seagate and Quantum marketing departments followed right away. As a result, Fast ATA (or Fast IDE) and Fast ATA-2 appeared, which was essentially the same standard as ATA-2, developed and promoted by Seagate Technologies. Both of them implied ATA-2 HDDs support though Fast ATA had a bit limited data transfer mode: it supported only max PIO mode 3 and multiword DMA mode 1.

By the end of 90s most companies had already given up the idea of perplexing the users with all those numerous names of one and the same thing, Western Digital remained aloof and retained this name slightly modifying the interface standard itself every year. (At first there was no PIO mode 4 support, for instance). So, 8.4GB had been already surpassed, UltraATA/100 was knocking at your door, and Western Digital was still sticking to EIDE although it was hard to tell what they actually meant under this abbreviation then.

In 1997 ATA-3 came out. Though it seemed more correct to call it ATA-2.5, because it was a minor revision to ATA-2. To tell the truth, the main innovation introduced in ATA-3 was S.M.A.R.T. and a bit higher reliability. So, there was practically no equipment in the market supporting this standard and that's why the next standard, ATA/ATAPI-4 appeared very soon, in 1998.

Originally, the IDE/ATA interface was designed to work only with hard disks. CD-ROMs and tape drives used either proprietary interfaces or the floppy disk interface, which was slow and cumbersome. Then it became apparent that there would be enormous advantages to using the standard IDE/ATA interface to support devices other than hard disks, due to its high performance, relative simplicity, and universality. However, the structure of ATA commands made it impossible to simply put non-hard-disk devices on the IDE channel and expect it to work. That's why a group of CD-ROM drives manufacturers with the help from Western Digital and Oak Technology developed a special protocol called the ATA Packet Interface or ATAPI (the name "packet interface" comes from the fact that commands to ATAPI devices are sent in groups called packets). The ATAPI standard supported devices like CD-ROM drives and tape drives. It enabled them to plug into the standard IDE cable used by IDE/ATA hard disks, and to be configured as master or slave, etc. just like a hard disk would be. As a result, it didn't require any special controllers.

Internally, however, the ATAPI protocol was not identical to the standard ATA (ATA-2, etc.) command set used by hard disks. A special ATAPI driver was used to communicate with these devices. This driver had to be loaded into memory before the device could be accessed. Moreover, ATA was deprived of some useless and outdated commands and got some new but very important ones instead. And at last a new data transfer protocol appeared: multiword DMA mode 3, called UltraDMA, which allowed obtaining a much higher ATA throughput (up to 33MB/sec) and retaining the integrity of data (thanks to Cyclic Redundancy Check feature) transferred via a conventional 40-pin cable.

The situation again proved very similar to ATA-2 case. The companies' marketing departments again interfered and hence all hard disks supporting this standard entered the market as UltraATA/33 HDDs. Thank goodness the companies managed to agree upon a single marketing strategy!

Now T13 is working on ATA/ATAPI-5 standard, which will follow in the footsteps of ATA-3 and play an intermediate part between ATA/ATAPI-4 and ATA/ATAPI-6. The standard won't undergo any serious changes. In other words, there appeared two more data transfer rates: UltraDMA (ATA) with 44MB/sec and UltraDMA (ATA) with 66MB/sec throughput - a new version of ATA (ATA/66) proposed by Quantum Corporation, and supported by Intel.

The transfer rate increase turned out so great that the 40-pin cable couldn't provide the appropriate bandwidth, because it was initially designed for the rate of about 5MB/sec. This situation forced the developers to add another 40 conductors (ground lines) to this cable so that to increase the stability and protection against external interference.

You should remember the reaction of the manufacturers marketing departments that followed the launching of ATA/ATAPI-5 hard disk drives: they called them "UltraATA/66". This standard should allegedly be approved this year. Well, we'll see.

At the same time T13 has been working on ATA/ATAPI-6 as well since the end of 1999. This spec should include a lot of items missed out in the previous standard. Namely, LBA increased from 28bit to 64bit, new faster UltraDMA modes with the bandwidth of up to 100MB/sec, HDD noise reduction, new commands for audio/video streams transferring suggested by Quantum, Western Digital and Philips.

Actually, now all of us have a very good chance to watch hardware components developing rapidly. Their performance improves and grows incredibly fast, so that we hardly manage to make the parts of our system match each other in this respect. Besides, beautiful games, large audio and video files have turned more predominant recently and are now occupying most of the system storage space on nearly every PC. Indisputably, the connection between the controller and the storage device should keep pace with all these enhancements in order not to stymie the system. The hard disk drives internal data rates ideally shouldn't surpass the burst transfer rates of the interface supported by the system, otherwise the host-to-drive bus will appear the bottleneck, because it will be unable to cope with so much data at a time. As a result, the data will be queuing in the buffer memory of the HDD and slowing down the system performance. One of the possible ways-out in this case, which may seem the easiest, is to increase the size of the HDD buffer memory, so that the data could be held there before the transfer is successfully finished. However, it will hardly help to eliminate the problem completely, because the higher gets the HDD internal data rate, the more buffer memory is required to avoid delays. And this process may never end.

However, it is also possible to approach this problem from a different side: to enhance the bus between the harddisk and controller and to increase its bandwidth, so that the data could be transferred at a higher rate. In other words, the data won't be kept in the hard disk drive buffer memory for long and will be transferred via the faster bus almost as soon as it gets there. That's why there will be no need in large buffer memory and it won't require any additional spendings from the HDD developers.

Having weighed all pro and contra, Quantum developed a new ATA/100 interface with additional speed, which allows transferring the data at 100MB/sec along the host-to-drive bus and hence to unload the HDD buffer memory. This new interface incorporates all the innovations made to the cables and connectors supporting ATA/66. Just to remind you let us say once again that ATA/100 interface requires the same good old 40-pin IDE cable used in ATA/66. However, since the burst transfer rates have become considerably higher, better protection against crosstalk and electromagnetic noise interference turned out necessary. For this particular purpose 40 additional ground lines were tied together and acted as shields between the lines that actually carried live signals back and forth. So, altogether the new ATA/100 cable is composed of 80 conductors. However, the developers wanted to retain the conventional 40-pin connector, so that to ensure plug compatibility with existing drives and systems. In other words, we can state that ATA/100 has another worthy feature: it is backward compatible with ATA/33 and ATA66 devices, including hard disk drives, removable media disk drives (for example, ZIP, JAZ), CD-ROM drives, CD-R/RW drives, ATA tape drives, and DVD-ROM drives.

The new interface, ATA/100, also includes a Cyclic Redundancy Check (CRC) feature, which serves to ensure the integrity of transferred data and to detect data transmission errors. The CRC is a very powerful but easily implemented technique to obtain data reliability. It is used to protect blocks of data called frames. Using this technique, the transmitter appends an extra n-bit sequence to every frame called Frame Check Sequence (FCS). This verification code produced for each burst data transfer by both the host and the drive is stored in their respective CRC registers. The FCS holds redundant information about the frame that helps the transmitter detect errors in the frame, i.e. the host and drive CRC register contents are compared with each other to make sure that the data burst sent matches the data burst received. If they don't match, then the process is repeated automatically, until it succeeds. This technique gained its popularity because it combines the following three advantages:

• Extreme error detection capabilities;
• Little overhead;
• Ease of implementation.

Well, if you got really attracted by all these promising details, then you should definitely try it in practice. However, your system should meet certain requirements, otherwise you will fail to enjoy the thing. Here you are:

• You should use an external controller card supporting ATA/100 interface or an integrated controller on your mainboard;
• Your hard disk drive should support ATA/100 interface;
• You should have an 80-conductor cable like the one used for ATA/66 interface connection.

The controller on your mainboard may be integrated into the set of your mainboard core logic or simply soldered to the board. As for the core logic, the today's choice in chipsets supporting ATA/100 is not so rich, however, for the time being it should be OK. One of the largest chipset manufacturers, Intel, announced that its i820E and i815E chipsets would support ATA/100 interface. Note that only if the chipset has this "E" in its name it means that it supports ATA/100 interface. That's why you should be very careful when buying one because you may get an ordinary i815, for instance, with ATA/66 support only instead of i815E. The difference lies in the South Bridge chipset. The ordinary i815 uses 82801AA chip, while the "enhanced" one - 82801BA chip, supporting ATA/100. The same thing is valid for i820 and i820E as well.

Besides, on June 27 VIA Technologies brought out a press release and announced support for the high-speed ATA/100 hard disk drive interface in its upcoming South Bridge chipsets. The first one is VT82C686B, which is pin compatible with the old 686A, i.e. it can be used for all the today's chipsets so that to make them support ATA/100. And the other two South Bridges are totally new: VT8231 (value) and VT8233.

ATA/100 standard is currently referred to as Parallel ATA since it transfers data in parallel. In other words, multiple bits of data are transferred simultaneously. But this method has a lot of drawbacks. First of all, manufacturing an 80-pin cable is a rather expensive task. Besides, the cable like this occupies a lot of space in the PC case, which undoubtedly tells on the cooling preventing the airflow from circulating properly. Frankly speaking, ATA/100 has every chance to become the last standard of this type, because it may turn out pretty hard to push the standard much further than 100-133MB/sec.

Well, ATA/100 seems to be pretty cool, however, the developers of Intel, IBM, Maxtor, Seagate, Western Digital, Quantum, Dell, and other companies are currently working on a totally new interface, which should open a new era of super high transfer rates and new design concepts. This new interface is known as Serial ATA. So, how does the Serial ATA work? In fact, the name speaks for itself. Serial ATA transfers the data serially, i.e. one bit at a time. The primary advantage of this standard is the possibility to make the connector much smaller and the cable much narrower without losing its efficiency. You will be surprised, but there are only 4 signal pins and just a few more standing for power and ground lines.

The promised features of the serial ATA interface will look as follows:

• Backward compatibility with Parallel ATA interfaces;
• Separate channel for each HDD (no need to share bandwidth);
• Lower power consumption;
• Hot plug support (optional).

Summing up all the facts and figures about the data transfer interfaces we get the following picture:

Actually, looks very promising. But this is still the future, though a rather near one. And now that we have already discussed all the theoretical stuff, let's give our mind a short rest and take a look at a bit of practice.

We decided to find out if the new ATA/100 interface is really worth the stir around it and tested a hard disk drive with the three main protocols: ATA/33, ATA/66 and ATA/100. The tests were carried out for two different platforms. The first one is based on i815E mainboard (the test were run for all three protocols) and the second one is on VIA Apollo Pro133 (the tests were run only for ATA/33 and ATA/66 protocols).

The HDD we chose for our experiment was a 46.1GB IBM DTLA-307045 drive. This HDD features 7,200rpm rotation speed, 8.5msec average seek time and can boast a 2MB data buffer.

We have to point out that during our tests we switched from one protocol to another with the help of IBMATASW (version 1.40) utility, which you can download here.

So, the tests. The testing system was configured as follows:

• Intel Pentium III 933 CPU;
• Chaintech 6ATA4 (VIA Apollo Pro133A), ASUS CUSL2 (i815E) mainboards;
• 128MB PC133 SDRAM by Micron;
• Creative 3D Blaster Annihilator Pro graphics card;
• Creative SoundBlaster Live! sound card.

Here are the results obtained in WinBench 99.

As you can see from the chart, in ATA/66 the disk performed better in the system on i815E. The same thing refers to ATA/33, where i815E based system proved a slightly better platform for our HDD. In fact, it is quite possible that we should blame the controller for this, but as for us, we think the major reason is the drivers. However, as it comes to ATA/100, the performance gain appears not very noticeable though we have to admit that it exists.

And now let's take a look at the hard disk drive performance in all sorts of office applications:

These results once again prove our point expressed above. Testing hard disk drives is a nice and illustrative thing, of course. However, we also wanted to take a look at the influence of ATA/100 interface on the overall system performance and hence tested our systems with Content Creation Winstone2000. the results turned out practically the same for all interfaces. The largest difference mad only 0.1 point that's why we made up our mind not to overload our article with a row of columns of the same size. :)

In order to give you a better idea of how linear read rate can influence the disk performance, we suggest looking at the data transfer rate taken in the beginning and in the end of the HDD with different protocols. The chart clearly shows that since the liner read rate is higher than the bus bandwidth, when working with ATA/33, the bandwidth appears insufficient. Besides, you have probably noticed that there is no difference between the linear read rate in case of ATA/66 and in case of ATA/100. It simply means that ATA/66 is quite enough to provide the appropriate linear read rate. However, as you can see from other tests, the difference between them still exists. Namely in real life applications when the HDD is requested more randomly, the data can be read from the disk buffer memory as well as from the disk itself. In case the data is read from the fast HDD buffer, the bus bandwidth turns the limiting factor and the advantages of ATA/100 prove indisputable.

Conclusions

Well, as you see, there is no cause for concern even if you have ATA/33 interface and if you don't work with huge audio and video files. As for ATA/66, it is enough for your today's needs as well. And as far as ATA/100 is concerned, of course it provides a certain performance gain compared to ATA/66. However, the today's hard drives are too slow to make this gain a significant argument in favor of ATA/100 interface. But certainly every innovation is always aimed at the future rather than at the present. Everything keeps developing and very soon hard disk drives may appear much faster rotating at 10,000rpm, for instance, or even more. This is exactly when the new interface will manage to show off.

Besides, we have to admit that the IDE controller in i815E proved more efficient than that by VIA. Maybe it is the drivers, which are the reason of this success, and maybe it's simply the hardware quality. Anyway, let's wait for the promised ATA/100 support in VIA South Bridge. And then we'll see, who is the best…