Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Highest possible frequency...

Status
Not open for further replies.

lord loh.

Member
Someone told me of a 70 GHz Oscilloscope... I was really shocked... I never believed that silicon amplifiers could do better and a few MHz...

What is the maximum possible frequency that solid state components can operate ?

If basic TTL gates have propagation delay time of 4 ns then it can operate at maximum of 250 MHz... So how can PCs operate at GHz range ?

Thanks in advance...
 
lord loh. said:
Someone told me of a 70 GHz Oscilloscope... I was really shocked... I never believed that silicon amplifiers could do better and a few MHz...

Did it say it was silicon?, that's not the only semiconductor, various others are often used for ultra high frequencies - gallium arsenide is a common one.

However, silicon certainly works a great deal higher than a few MHz, low GHz is no problem, and even cheap little silicon transistors reach 100's of MHz.

What is the maximum possible frequency that solid state components can operate ?

It's limited by the technology of the day, and increases all the time - but generally the higher the frequency, the more expensive the technology!.

If basic TTL gates have propagation delay time of 4 ns then it can operate at maximum of 250 MHz... So how can PCs operate at GHz range ?

They don't use TTL 8)

I suggest you try researching on Intel, IBM etc. they have all been pushing the boundaries of technology, usually in slightly different ways. You have probably noticed that the voltage supplies used for CPU's has fallen as their speed increases?.
 
my cpu runs at 0.95V @ 1.56Ghz, dont forget that ur transistor is fairly larger (well incredible large) when compared to a CPU's transistor. But they are cheaper to make to desired accuracy's.

Terraherz? possible? maybe. But not soon, CPU manaufacturers are slowing down the speed increases to pile more features in such as hyperthreading and dual CPU's and quad CPU's onto one silicone slice. Which means ur os can use one cpu for office work, one for games / music, one for SETI and another for spam, anti-virus and firewall protection ;)
 
yes indeed these days whats hot is not the GHz race. its the 64-bit architechture, low power consumption and the other things that pittuck pointed out. i have been reading reviews and benchmarks of processors for years and i have definitely seen a reduction in the amount of increase in the operating speed of processors per month. look at the newest processor from intel; the p4-600 which has a 2MB L2 cache. the speed is still there 3.6~3.8GHz. so these days engineers are trying to improve the architechture of the microprocessors and the manufacturing technology.

oh okay i went wayyyyyyyyy off topic.

i think that a 70GHz+ oscilloscope is practicable but the costs will be huge. ICs made with gallium arsenide or simillar materials as Nigel has pointed out can do things at such high frequencies but they are really expensive and hard to get. only some high tech R&D lab would require such an oscilloscope!!!!!!
 
Dr.EM said:
I've heard of Gallium arsenide LEDs. Also heard Gallium arsenide was really awful stuff, presumably contains arsenic among other nasties.

Most semiconductors contain nasty poisons, it's best not to digest them :lol:
 
Nigel Goodwin said:
They don't use TTL 8)

I suggest you try researching on Intel, IBM etc. they have all been pushing the boundaries of technology, usually in slightly different ways. You have probably noticed that the voltage supplies used for CPU's has fallen as their speed increases?.


Oh....Thanks...

I really thought that increasing the voltage increased the speed...
And isn't CMOS always slower than TTL ? although it consumes less power..

I shall check IBM and Intel's works...anyway.
 
lord loh. said:
I really thought that increasing the voltage increased the speed...
And isn't CMOS always slower than TTL ? although it consumes less power..

Nope CMOS consumes less power because the insulation layers on the chip die are way thinner, and as a result faster. The only draw back is that the maximum operating voltage is lower and it is more susceptible to ESD (electro static discharge) which renders the chip useless. And more expensive!!
 
Intel hit a brick wall and gave up on trying to hit the 4ghz mark because of frequency limitations. Now they are focusing on dual core and 64 bit processors instead...

The Characteristic Impedance of wire goes to inifinite pretty much at 17ghz... but it starts getting harsh at 3ghz, and basically unusable at 4ghz. Processors are having the same types of problems when pushing for speed, the cost verses the measily gain of a few MHz simply isn't worth it.
 
Fascinating! , and yet nobody has mentioned the core problem of heat.
Heat is what places constraints upon the upper limit of the nano scale transistors switching, clocking a computer above 4ghz can be done, but not for very long ! a few tenths of a second and the substrate becomes molten slag. Of course as Nigel pointed out silicon is not the only semi conductor, now silicon has a fairly low melting point and most computer chips start to show serious distress if allowed to exceed a mere 70 degees centigrade. The solution .. a semi conductor more tolerent to heat is required, well one does exist. Crystals of pure carbon are at present the focus of much research, though quite how an average home user is going to cool a motherboard running at over a 1000 degrees centigrade will present a challenge all by it'self.

Tinkering with the architechture gets around some of the problems of silicon, IBM's approach of four processors in a single package is being watched with some interest. Wether Microsoft is going to embrace this distributed computing is open to much speculation with many large manufatures and server operators already running Linux.

Of course technology might make a U turn and ditch silcon transistors as an engineering dead end , returning to vacuum thermionic devices wich are heat tolerent and resistant to damage from electromagnetic pulses.
Whilst I doubt the glowing tubes of our grandfathers will make a comeback, the lessons learned from etching silicon allows for the devices to made immensely small.
 
i remember once seeing a 5Ghz comp on slashdot, pentium 4 with some liquid nitrogen to cool it. Was quite a set up, but it is possible ;)
 
A video of a 5ghz computer project....

**broken link removed**

Warning : Link contains images of nudity,not suitable for persons under 18
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top