|
Why Concurrent PCI from
Intel delivers superior
multimedia performance.
Better Video.
Intel’s Concurrent PCI Architecture, available with the Intel 430TX
PCIset, enables more efficient operation of CPU, PCI and ISA transactions for faster
and smoother multimedia performance. Concurrent PCI’s Multi Transaction Timer
(MTT) optimizes performance for “short-burst” applications such as RGB video capture
and tuner cards. In video applications, you’ll see smooth video frames and higher overall
system performance.
Better Audio.
Concurrent PCI’s Passive Release reduces CPU utilization for soft
MPEG clips. That’s because Concurrent PCI allows CPU and PCI masters interleaved
access to the bus during ISA, DMA and bus master cycles. CPU and PCI masters have
concurrent access to memory, even while a relatively slow ISA transaction “owns” the
bus. Best of all, these performance advantages are available without reprogramming.
Better Scalability.
Intel’s Concurrent PCI Architecture provides deeper write buffers
to enable superior multimedia performance for both CPU and PCI add-in cards. The Intel
430TX PCIset enables the use of PCI 2.1 and 2.0 compatible add-in cards for long system
life, built-in scalability and the flexibility to adapt your system to future applications.
| FEATURE | TECHNICAL SUMMARY | BENEFIT TO APPLICATION |
| Multi Transaction
Timer (MTT) |
The MTT allocates a minimum time slice to a
PCI Master that holds its request line asserted.
A PCI Master of this type performs multiple
“short bursts” on PCI without having to
rearbitrate for each burst.
|
For “Short Bursting” cards with time sensitive
data (e.g. Frame Grabber boards), MTT reduces overhead
to enable fast multiple data xfers in one PCI request.
Video data streams from Frame Grabber cards appear
consistent and smooth,
even under varying system loads. |
| Passive Release |
This mechanism enables more efficient ISA Bus
Master cycles to PCI by absorbing ISA overhead
before requesting PCI. On multiple data cycles,
the PCI bus is free for other agents’ use in
between each ISA bus data phase. |
For systems with ISA and PCI masters, passive release
enables more efficient PCI transactions, reducing both
PCI Master latencies and CPU utilization. As an example,
audio cards place less load on the system, enabling
more CPU and PCI concurrency.
|
| PCI Delayed
Transaction |
This mechanism latches PCI-to-ISA cycle
information into the Intel PCIset, freeing the
PCI bus. Masters then use the PCI bus while
the PCIset performs necessary transactions on
the ISA bus. Data is transferred later to
complete the transaction. |
CPU-to-ISA transactions are made more efficient for
PCI masters since the PCI bus is not slowed down
waiting on CPU-to-ISA transactions. Delayed
transactions to any ISA card and within the PCIset
itself improve overall system efficiency, resulting in
smooth display of video frames in multimedia
applications. |
|
Enhanced Write
Performance |
The Write Performance of the 430 family of
PCIsets has been improved in four ways: deeper
CPU-to-DRAM and PCI-to-DRAM buffers, write
merging on CPU-to-DRAM writes, and faster
retiring to DRAM. The write merging feature
collects multiple byte, word, and Dword cycles
and merges them into a single write to DRAM. |
With larger posting buffers DRAM availability has less
impact on write cycles, reducing overall system latencies.
By merging CPU writes to DRAM of less than 64 bits
wide, DRAM performance is increased on both 16-bit
and 32-bit applications. Overall, multimedia compression
and decompression code runs faster. |
Intel’s Concurrent PCI Architecture optimizes performance of Pentium® processor-based systems.
Concurrent PCI Architecture is available on motherboards designed with the Intel 430TX PCIset.
It’s the easy way to optimize multimedia performance and quality, while unleashing the power
that the Pentium processors were designed to deliver.
|
![[INTEL NAVIGATION HEADER]](../../../PIX/HEADER.GIF){kind=small}
