Commodore introduced the Commodore 128 in 1985 as an upgrade path from the Commodore 64, the most popular model of computer of all time. The 128 addressed the 64’s biggest shortcomings while remaining mostly compatible with its hardware and software. That makes the Commodore 64 vs 128 a natural comparison, even more natural than comparing the 64 with the VIC-20.
As its name implies, the 128 had 128 kilobytes of memory, double the amount of the 64. Because the 8-bit 6502 series of CPUs could only address 64K of memory at a time, the 128 had two banks of 64K.
The most notable features of the 128 were a faster disk drive, an 80-column display, a larger keyboard with a separate numeric keypad, and built-in CP/M compatibility.
The design team
Most of the C-64’s key engineers left the company soon after its release, including Charles Winterble, Al Charpentier, and Bob Yannes, designer of the SID chip. This limited Commodore’s options when it came to adding more voices to the SID chip, or adding more sprites or colors or higher resolution to the VIC-II chip.
There also wasn’t a lot of time. Commodore rushed the 128 through when it realized the Amiga wouldn’t be ready for market in early 1985. So the team responsible for the Plus/4, including Bil Herd and Dave Haynie, did most of the 128’s design work using slightly modified chips from the 64 as well as chips intended for other computers that Commodore never released.
Commodore 64 vs 128: Original price
The Commodore 64 originally sold for $595 when it was released in 1982, but by 1985, it was an inexpensive machine, selling for around $149. Kmart would sometimes use it as a loss leader, selling it for $99. The 1541 disk drive and 1702 composite monitor often weren’t discounted quite as heavily, usually selling for around $199. It was generally possible to put together a full system for around $500.
The Commodore 128’s original price was $349 for the system alone. The 1571 disk drive and 1902 RGB monitor sold for $299 each. So a complete Commodore 128 system as Commodore envisioned it retailed for around $950. In practice, it was often possible to pick up a 128 for $299 and a 1571 for $249.
The faster disk drive
Commodore had intended for the 64 to have faster disk access, but a mistake in the manufacturing of the original motherboard made this impossible without a revision that would have delayed the machine’s introduction. Commodore expected most people would use a tape drive anyway, so they were able to justify it to themselves, mostly. Then disk drives proved more popular than anyone expected. So Commodore addressed the problem in the combination of the 128 and the new 1571 disk drive. In 64 mode, however, drive access slowed down to the same speed as a real 64.
One of the most popular products ever made for the 64 was the fast load cartridge. The combination of a 128 and 1571 disk drive made fast loaders less necessary–in c128 mode. The 1541’s frailties led to lots of clones. The 1571 had a much better track record.
The 1541 worked with the 128 too, but it was just as slow with a 128 as it was with a 64. But nothing stopped someone from buying a 128 and using a 1541 with it as a stopgap until they could afford a 1571, and even keeping the 1541 as a secondary drive after upgrading. Some did.
Disk drives still connect the same way on both the 64 and 128.
The 80-column display
The 80-column display came via a separate video chip with its own dedicated RGBI video output. It was completely independent of the VIC-II chip. Commodore salvaged this chip from another computer that never came to market, the ill-fated Commodore 900, which was intended to be a Unix-compatible workstation. So this chip, the 8563 Video Display Controller (VDC), wasn’t compatible with the VIC-II.
The 80-column display was popular for word processing and spreadsheets, but aside from those and a handful of Infocom titles, there wasn’t a lot of commercial software that took advantage of this capability. Had Commodore used a better known chip, such as a 6545/6845, this may have played out differently.
Commodore itself didn’t really understand the capabilities of the VDC when it used it, because its original engineers left the company before they documented it. It was cumbersome to program and Commodore had a hard time with the yields on the chip initially. Commodore initially stated the chip was only capable of displaying text. Later it turned out the VDC did in fact have bitmap graphics capability. When outfitted with enough video RAM, it was capable of 640×200 resolution comparable to the IBM PC’s CGA display.
Not every 128 owner used this capability as it required an RGB monitor, which initially cost about $350. The 80-column display would work with a composite monitor like the Commodore 1702, used with the 64, but it only provided a monochrome display.
A faster CPU
One key difference between the Commodore 64 vs 128, at least on paper, was CPU speed. The C-128 had an 8502 CPU capable of running at 2 MHz, twice the speed of the C-64’s 6510 CPU. Unfortunately the VIC-II chip was tied to the 1 MHz speed of the 64, so when you used the 128 in 40-column mode, the CPU was limited to 1 MHz. You could use a software trick to switch the CPU to 2 MHz while the 128 wasn’t rendering the visible part of the screen, which resulted in about a 20% improvement.
In practice, to get the full benefit of the faster CPU, you had to run the 128 in 80-column mode.
The 128’s CP/M mode was a bit of an accident. Commodore marketed a CP/M cartridge for the 64, intending to give it access to popular business software. The problem was that the CP/M cartridge didn’t work with post-1983 revisions of the 64. It was one of the few things that worked well on the silver-label 64. Commodore tried to get this cartridge working with the 128, but failed in that attempt as well. Its designers ended up putting a Z-80 CPU on the board itself so nobody would want a CP/M cartridge. The decision was controversial in 1985 and remains so today.
But Commodore solved compatibility issues with certain other C-64 cartridges and peripherals by having the Z-80 handle initialization before handing control over to the 8502. Commodore was able to license the CP/M operating system cheaply and the Z-80 wasn’t an expensive chip, so the increased cost of the CP/M capability wasn’t high.
Today it’s common to say Commodore should have used MS-DOS instead and made the 128 IBM compatible, but that would have increased the cost of the machine dramatically–more than double.
CP/M provided a large library of software available for the 128, but few Commodore users knew how to get their hands on it. You pretty much had to be in a large metro area that had a CP/M users group. Even then, you needed a 1571 drive since the software usually came in disk formats that weren’t compatible with the 1541. Commodore made the 1571 capable of reading MFM disk formats that other CP/M computers used in order to gain that capability.
A new Basic
Less controversially, the 128 included an enhanced version of the Basic programming language with graphics and sound commands. One of the criticisms of the 64 was that included industry-leading sound and graphics capabilities but one had to resort to cryptic POKE commands to make any use of them. With the 128, hobbyist programmers could easily create their own programs with graphics and sound. The 128’s commands were generally more intuitive than equivalent 64 commands.
The biggest difference when considering the Commodore 64 vs 128 was their respective software libraries. The Commodore 64 had one of the largest software libraries of its generation, eventually growing to around 10,000 titles. It was more software than anyone could ever hope to own or use during the useful life expectancy of the machine.
There was a much smaller selection of software that ran in the native C128 mode. Only a couple hundred commercial titles for 128 mode existed.
Commodore software was generally written for the least common denominator, a 64 with a single 1541 drive. That assumption assured the highest possible sales. Software that took advantage of the 128’s additional capabilities was rare, but so was software that took advantage of two disk drives with a 64. Apple software was much more prone to push the envelope, requiring dual drives and the 80-column display. Apple software developers tended to assume more affluence, and therefore more inclination to buy additional peripherals. They were probably correct.
Commodore 64 mode
In C64 mode, the 128 reverted into a close copy of its predecessor, with a 40-column screen, slow disk access, the new keys disabled, and the old, limited Basic. The machines weren’t 100% compatible, but compatibility issues were rare. Commodore claimed 99% compatibility because they couldn’t test absolutely everything, and in reality they probably hit better than 99%. The only compatibility issue I ever had was with a copy of the game Archon II, which worked fine on a real 64 but locked up on my 128. The problem likely was something the cracking group introduced in this particular copy, as I’ve never heard of anyone else having the issue.
While I’ve heard people say the 128 wasn’t 100% compatible with the 64, I’ve never heard anyone else actually give a specific example of something that didn’t work.
It was possible to re-enable the keypad and the 80-column display in C64 mode. Things like this could be a source of incompatibility if software wasn’t aware of those memory locations and how they behaved. But prior to the 128’s introduction, there was no reason for 64 software to touch those memory locations.
From a reliability standpoint, the biggest contrast between the Commodore 64 vs 128 was the power supply. The C-64 shipped with a cheaply made, epoxy-filled brick that tended to fail a lot, fail in a way that blew up memory chips, and couldn’t be repaired when it failed. The 128, in contrast, shipped with a reliable power supply that could be repaired and had an easy to replace fuse. Bil Herd went against orders and designed the power supply he would have wanted as a user, rather than the one management wanted.
While it’s not safe to use a vintage Commodore 64 power supply anymore, the 128 power supply generally fails in a safe manner. It’s very uncommon for a 128 power supply to damage the machine.
Commodore also used an odd square connector on the Commodore 128 power supply. This kept someone from accidentally plugging the power supply into the port intended for the disk drive or printer and damaging the machine.
Commodore 64 vs 128: Success
The Commodore 64 was the most successful home computer of all time. The sales figures of both machines are a matter of frequent debate but there’s no question the 64 outsold the 128 by a wide margin, likely around 5 to 1. This has led people to call it a flop, but compared to the 64, every other 8-bit computer was a flop.
People forget that for the first couple of years, Commodore couldn’t keep up with demand for the 128. When we bought ours in 1986, the dealer put us on a waiting list. I hear the same stories from people who tried to buy the 128D in 1987. Commodore probably would have sold more 128s if it had been able to make them. Had Commodore invested more money in expanding or modernizing its chipmaking plant instead of paying for Irving Gould’s lavish lifestyle, both computers might have been more successful.
I think it’s more fair to compare the 128 to the Plus/4, which sold around 250,000 units. Nobody disputes the Plus/4 was a flop. The 128 came out in 1985, facing a much more hostile market with 68000-based computers looming on the horizon, and still sold several million units. Its success rivaled that of the VIC-20 and Amiga 500.
The 128 never lived up to the 64’s popularity, but it wasn’t really intended to. That said, 4 million units was still a lot of machines in the 1980s. No single model of the Apple II ever sold that many. Perhaps surprisingly, it increased C-64 sales by giving the 64 an upgrade path. People could buy a $149 C-64, and later, when they could afford a 128, they could use all their old software and peripherals with it.
Commodore 64 vs 128: Legacy
Commodore expected the 64 to last between 2-4 years on the market. They learned the hard way in 1984 that any direct replacement for the 64 needed to be compatible with it. That’s why the 128 had compatibility with the 64. For that matter, 64 emulators for the Amiga were popular even though the early 64 emulators couldn’t emulate the 64 at full speed.
C-64 sales declined rapidly in the late 1980s. But the existence of the 128 slowed that decline by a couple of years. In the end the 64 sold 20-25 million units and had a number of different incarnations.
Commodore only expected the 128 to last about two years on the market. It ended up lasting closer to four, and it sold better than Commodore expected. In that regard, it was very much like the 64, whose sustained success was also an accident. Unfortunately, Commodore was never able to fully duplicate it again. The 128 bought Commodore time until the Amiga was ready. Unfortunately for Commodore, the Amiga didn’t really match the 128’s success, let alone the 64’s.
Commodore really needed another product to sell in 64-like quantities to survive beyond 1994. Had the 128 sold in better numbers, perhaps Commodore would have lasted longer before it bled out. But to really survive into the 1990s, Commodore would have needed the Amiga to sell better than it did.