Hardware Reference
In-Depth Information
large as common EPROMs and they are only half as fast. EEPROMs cannot com-
pete with DRAMs or SRAMs because they are 10 times slower, 100 times smaller
in capacity, and much more expensive. They are used only in situations where
their nonvolatility is crucial.
A more recent kind of EEPROM is flash memory . Unlike EPROM, which is
erased by exposure to ultraviolet light, and EEPROM, which is byte erasable, flash
memory is block erasable and rewritable. Like EEPROM, flash memory can be
erased without removing it from the circuit. Various manufacturers produce small
printed-circuit cards with up to 64 GB of flash memory on them for use as ''film''
for storing pictures in digital cameras and many other purposes. As we discussed
in Chap. 2, flash memory is now starting to replace mechanical disks. As a disk,
flash memory provides faster access times at lower power, but with a much higher
cost per bit. A summary of the various kinds of memory is given in Fig. 3-32.
Byte
alterable Volatile
Type
Category
Erasure
Typical use
SRAM
Read/write
Electrical
Yes
Yes
Level 2 cache
DRAM
Read/write
Electrical
Yes
Yes
Main memory (old)
SDRAM Read/write
Electrical
Yes
Yes
Main memory (new)
ROM
Read-only
Not possible No
No
Large-volume appliances
PROM
Read-only
Not possible No
No
Small-volume equipment
EPROM Read-mostly UV light
No
No
Device prototyping
EEPROM Read-mostly Electrical
Yes
No
Device prototyping
Flash
Read/write
Electrical
No
No
Film for digital camera
Figure 3-32. A comparison of various memory types.
Field-Programmable Gate Arrays
As we saw in Chap. 1, field-programmable gate arrays ( FPGA s) are chips
which contain programmable logic such that we can form arbitrary logic circuit by
simply loading the FPGA with appropriate configuration data. The main advantage
of FPGAs is that new hardware circuits can be built in hours, rather than the
months it takes to fabricate ICs. Integrated circuits are not going the way of the
dodo, however, as they still hold a significant cost advantage over FPGAs for
high-volume applications, and they run faster and use much less power. Because of
their design-time advantages, however, FPGAs are often used for design prototyp-
ing and low-volume applications.
Let's now look inside an FPGA and understand how it can be used to imple-
ment a wide range of logic circuits. The FPGA chip contains two primary compo-
nents that are replicated many times: LUT s( LookUp Tables ) and programmable
interconnects . Let us now examine how they are used.
 
 
Search WWH ::




Custom Search