Master bus

Figure 1 shows the hardware architecture of TOPFIT. There is one master processor which communicates with N slave processors through dual-ported RAM. The dual-ported RAM can be addressed by both the master bus and the slave bus. Each slave is an independent microcomputer with its own processor, memory, bus and operating system. Its computing power is close to 1 MIPS. 

The dual port design guarantees a very high communication bandwidth and, therefore, avoids a traffic bottleneck for the data exchange between master and slave. The number N of slaves to be controlled and served by the master processor is variable and can be installed by the operating system. It is limited only by the electrical characteristics of the master bus. 

The slave microcomputer consists of a CPU, random access memory, a memory management unit and a master bus interface. It has no ROM (read only memory). The whole slave operating system and the retrieval software are both downloaded from the master. Thus software updates and upgrades are easily implemented even on machines equipped with a large number of slaves. The slave operating system includes the following functions selfcheck, a simple command interpreter and master communication. 

The definitions of master, bus.active, and cdb handle the process of bus arbitration. We start by selecting an active result register, if one exists, and placing its index in master. We then define the boolean expression bus.active to be true if master indeed points to an active result register. Finally, we define cdb to be this active register. 

Click the Master FX button on the Master bus in the Mixer window. 

There are two places that FX can be added in ACID into a track or into the Mixer window. Within the Mixer window, there are three places where FX might go into an Assignable FX, into a pre-existing bus, or into the Master bus (see Figure 7.1). 

The last effects in a chain are best used at the project level as Master bus FX. A Reverb effect is often the last effect used in a chain. Reverb is also a delay-type effect and adds space to a mix, simulating various room types and environments. Use caution when applying reverb at the track level, so that various tracks don t sound like they were recorded in different rooms (although reverb is commonly used on drums and vocals to bring these parts out of a mix). EQ fdters can also be used at the project level to polish a mix in the final stages, although it is better to take the time and use EQ more precisely at the track level to highlight individual parts and instruments. 

It is not highly recommended that you use multiple EQ plug-ins in a chain. At a maximum, you may want to use EQ at the track level to fix a few problems and maybe then on the Master bus in the Mixer to polish off a project. You should think very carefully about what you are doing and whether there is a better way to more efficiently do the same thing. This is actually good advice for all plug-ins ... 

While you probably want to use reverb on the Master bus at the end of a project to keep all of the tracks in the same space, you may want to solo one of the primary tracks and adjust the reverb while listening to it alone. Reverb is more clearly beard in solo instrument and voice parts. Of course, before you settle on a final effect setting, you will need to fine tune the reverb for the whole project. 

MIDI tracks must be routed through the Master bus in order to be rendered with the project. All other devices are external to ACID. 

ACID MIDI tracks need to use the Master bus (DirectX SoftSynth) to render projects. Other MIDI devices listed on the Bus Selector (see Figure 10.2) act as external devices and are used only for playback. All MIDI tracks in a project automatically use the same device changing the device on one MIDI track changes all MIDI tracks. As one example, if you are using a Yamaha SXG software synthesizer, your project will sound as if the MIDI track was being mixed into the project. In reality, the audio tracks in the project are mixing in ACID, but the MIDI track is being processed by an external device (software, in this case) that is mixed with the output from ACID outside of ACID on the sound card. Therefore, when you render the project, the MIDI track is silent. There is a way around this problem, however. 

Also, most computer systems currently use ATA-2 technologies or above, which means that they can support four IDE drives. There will be a primary and a secondary IDE bus, and each will have its own master and slave drives. In these systems, you can have a primary master, primary slave, secondary master, and secondary slave drives. Which bus (primary or secondary) a device is shown as depends on the bus to which you connect the drive. From there, you can then designate each drive as master or slave as described in the next section. 

Most CMOS setup programs will display all the IDE buses and which devices are connected in which positions. You can check which bus and which designation (master or slave) a device is configured for by using "View drive setup" (or similar) in your computer s setup program. 

The first type is the simplest. There are two sets of pins, one labeled master/ single, the other labeled slave (Figure 4.13). If you have one drive, you jumper the master side and leave the slave side jumper off. If you have two drives, you jumper the master side only on the first drive (at the end of the cable) and jumper the slave side only on the other drive(s). A variant of this type uses no jumpers on either to indicate just one drive on the bus. 

If you have two drives on a bus and both are set to master, or both are set to slave, neither drive will work. In the first case (two masters), they will be fighting each other for control of the disks. In the latter case (two slaves), the disks won t know where to get their instructions from. However, if you have an EIDE interface system, you can have two masters as long as the drives are on two separate IDE buses. 

With DMA channels, a device can write directly to memory. But what if a device needs to read or write directly to another device (like the hard disk) For this purpose, bus designers came up with bus mastering. This is a feature that allows a... 

The ISA bus is easily identifiable by the presence of the small bus connector behind the 8-bit connector, as shown in Figure 5.3. This additional connector adds several signal lines to make the bus a full 16-bit bus. The other connector is a regular 8-bit bus connector. The ISA bus has eight more interrupts than the 8-bit bus and four additional DMA channels. Also, this bus can operate at nearly twice the speed of the older, 8-bit bus (ISA can run at 8MHz, and Turbo models can run as fast as lOMHz reliably). Finally, this bus can use one bus-mastering device, if necessary. 

MCA was a major step forward in bus design. First, it was available in either 16-bit or 32-bit versions. Second, it could have multiple bus-mastering devices installed. Third, the bus clock speed was slightly faster (lOMHz instead of 8MHz). And finally, it offered the ability to change configurations with software rather than with jumpers and DIP switches. 

There were several new, desirable features introduced with EISA. Its creators took the best of MCA s features and added to them. As we have already mentioned, EISA has a 32-bit data path. Additionally, it has more FO addresses, it allows expansion cards to be set up using software, there is no need for interrupts or DMA channels, and it allows for multiple bus-mastering devices. However, despite all these advances, it still uses the 8MHz clock speed of ISA (to ensure backward compatibility with ISA cards). 

As this book is being written, PCMCIA s bus width is only 16 bits, but a 32-bit version is on the way. Also, PC cards support only one IRQ (a problem if you need to install in a PC card bus two devices that both need interrupts). PC cards also do not support bus mastering or DMA. However, because of its flexibility, PCMCIA has quickly become a very popular bus for all types of computers (not just laptops). 

Bus Type Bus Width (Bits) Maximum Speed (MHz) Uses Bus Mastering Configuration... 

The most important goal of the network adapter card is to optimize network performance and minimize the amount of time needed to transfer data packets across the network. There are several ways of doing this, including assigning a DMA channel, use of a shared memory adapter, and deciding to allow bus mastering. 

PCI Abbreviation for Peripheral Component Interconnect. A specification introduced by Intel that defines a local bus that allows up to 10 PCI-compliant expansion cards to be plugged into the computer. One of these 10 cards must be the PCI controller card, but the others can include a video card, network interface card, SCSI interface, or any other basic input/output function. The PCI controller exchanges information with the computer s processor as 32- or 64-bits and allows intelligent PCI adapters to perform certain tasks concurrently with the main processor by using bus mastering techniques. 

VL bus Also known as VL local bus. Abbreviation for the VESA local bus, a bus architecture introduced by the Video Electronics Standards Association (VESA), in which up to three adapter slots are built into the motherboard. The VL bus allows for bus mastering. 

A. Direct Memory Access (DMA) channels allow a device to write directly to memory. Bus mastering allows devices to write directly to each other. See Chapter 5 for more information. 

As a consequence, we introduce the term of an application frame. It consists of the ordered sequence of real world events that happen in between two rendering steps. A master node thus collects an application frame, chooses a proper encoding, and sends the frame as a complete collection to all the slave nodes, where the sequence will be reconstructed and injected to the local event bus. As shown above in Fig. 3.58, this will result in the same state of the domain objects right before rendering. 

---