PPU variants

Beyond the well-studied 2C02G, we know of the following PPU revisions, both made by Ricoh and other manufacturers:

Official

Chips officially licensed by Nintendo for use in official consoles and arcade systems.

Composite

NTSC

All official NTSC PPUs expect a 21.477272 MHz master clock.

Part	First Seen	Last Seen	Notes
RP2C02	1983-06 3F4 13	1983-08 3H2 10	Extremely rare. Likely only a few thousand made. http://web.archive.org/web/20160315221802/kitayama3800.publog.jp/archives/cat_915765.html (pictures not archived)]
RP2C02A	1983-08 3H1 43	1983-10 3K3 59	Sometimes erroneous sprite pixels appear in X=255. Some modern PCBs generate almost exclusively glitchy pattern fetches. PPUMASK and PPUCTRL seem to be entirely asynchronous, and writes to PPUMASK can disable rendering for one pixel with the commensurate bugs resulting.
RP2C02B	1983-12 3M3 73	1984-05 4E1 54	Erroneous sprite pixels appear at X=255, just like 2C02A. Production seems to have halted to produce the 2C02C, but 2C02C production stopped and 2C02B production resumed, for unknown reasons.
RP2C02C	1983-12 3M1 10	1984-02 4B4 98	Eventually stopped being produced in favor of resuming 2C02B production, for unknown reasons.
RC2C02C	1984-01 4A3 15		Comes in a ceramic package.[1] Currently only found inside serviced Famicoms.
RP2C02D	1984-07 4G4 27	1984-12 4M2 29
RP2C02D-0	1984-10 4K3 63	1984-12 4M2 58
RP2C02E	1984-12 4M3 14	1985-10 5K4 36
RP2C02E-0	1985-03 5C5 46	1987-03 7C4 29	In this and all previous revisions, OAMDATA and palette RAM are not readable. Various OAM evaluation bugs
RP2C02G-0	1987-05 7E2 80	1993-10 3KM 1H	Writes to OAMADDR cause corruption of OAM. Leaving OAMADDR at a value of 8 or greater causes OAM corruption when rendering starts. Particularly susceptible to reflections causing OAM corruption, fixed by putting series resistors between the CPU and the cartridge ROM.
RP2C02H-0	1993-12 3MM 40	1999-05 9EM 5B	Thought have to have fixed some of the glitches in the previous 2C02G-0 revision^(which?).
RP2C02H-0 (laser)	2000-10 0KL 40	2003-01 3AL 4B	Reported (along with PAL PPUs) to have some kind of difference that caused problems with address bus filtering in earlier versions of the Hi-Def NES firmware.

PAL

All official PAL PPUs expect a 26.601712 MHz master clock.

Part	First Seen	Last Seen	Notes
RP2C07	1985-12 5M4 26		PAL-B PPU. Vblanking is 71 scanlines long. OAM evaluation can never be fully disabled. Red/green color emphasis swapped. OAMADDR works correctly, not corrupting OAM contents]
RP2C07-0	1987-10 7K3 27	1991-10 1KM 13
RP2C07A-0	1992-06 2FM 22		Some subtle differences in PPU that make this work better with Kevtris's HDNES, but otherwise believed identical to 2C07.

RGB

All official RGB PPUs are NTSC and expect a 21.477272 MHz master clock.

Part	First Seen	Last Seen	Notes
RC2C03	1983-11 3L4 15		Found in the Sharp C1 TV. Suspected to be same core as 2C02 letterless.
RP2C03B	1984-02 4B2 36	1989-03 9C4 23	RGB PPU. Color emphasis bits set the corresponding channel to full brightness. Otherwise believed identical to 2C02B. OAMDATA and PPU palette are not readable.
RC2C03B	1984-01 4A2 14	1984-01 4A4 30
RP2C03C	1984-03 4C2 63		RGB PPU. Believed to be same core as 2C02C.
RC2C03C	1984-01 4A2 10		Believed identical to RP2C03C.
RP2C04 0001	1984-03 4C2 01	1984-05 4E4 35	RGB PPU with a scrambled palette and some added colors, for copy protection purposes. Equivalent to 2C02D, but keeps 2C03 timing (no skipped dot).
RP2C04 0002	1984-07 4G4 19	1984-10 4K2 35	Scrambles the palette uniquely compared to the previous revision.
RP2C04 0003	1984-10 4K5 13	1984-11 4L4 33	Scrambles the palette uniquely compared to the previous revisions.
RP2C04 0004	1984-11 4L3 18	1984-11 4L5 36	Scrambles the palette uniquely compared to the previous revisions.
RC2C05-01	1985-06 5F5 10	1985-06 5F5 11	PPUCTRL and PPUMASK swap locations. Five LSBs of PPUSTATUS return a constant. Otherwise believed to behave like 2C03.
RC2C05-02			Believed to be the same as 2C05-01, with the PPUSTATUS constant changed.
RC2C05-03	1985-07 5G1 12		Believed to be the same as 2C05-01, with the PPUSTATUS constant changed.
RC2C05-04	1987-03 7C3 12		Believed to be the same as 2C05-01, with the PPUSTATUS constant changed.
RC2C05-99	1988-11 8L4 11	1988-12 8M4 14	An RGB PPU with RP2C02E behavior: OAM and palettes are unreadable, but OAM corruption added in revision E is present. PPUSTATUS behavior is normal (the low 5 bits are PPU open bus). Grayscale is normal (palette & $30). Uses 2C02 timing (a dot is skipped every other frame), unlike other RGB PPUs. The palette and emphasis behavior match other RGB PPUs. PPUCTRL and PPUMASK aren't swapped. Used in the Famicom Titler. (picture of Titler PCBs)

Unofficial

These chips are found exclusively inside of Famiclone systems, made by multiple companies.

Part	Clock	Video	Notes
UA6528	21.48 MHz	NTSC	UMC-made clone of 2C02E (or something earlier) [2]
UA6528P	21.49 MHz	PAL-N	UMC-made variant for PAL-N (Argentina) [3] System crystal is 21492337.5 Hz (exactly 6×229.2516×15625)
UA6538(P??)	26.60 MHz	PAL-B	UMC-made variant for playing NTSC games in PAL countries. Emits PAL-B video. Vblank interrupt intentionally emitted 50 scanlines later than 2C07. Video DAC is brighter and more saturated ^{(how so?)}. See also Clock rate.
UA6541	26.60 MHz	PAL	UMC-made clone of 2C07 [4]
UA6548	21.45 MHz	PAL-M	UMC-made variant for PAL-M (Brazil) System crystal is 21453671… Hz (exactly 6×227.25×4500000÷286)
UM6558	21.31 MHz	SECAM	UMC-made variant of UA6538 for SECAM countries. Emits 8-bit "Color Data" digital bus, for conversion into SECAM by UM6559 IC. Color palette noticeably off. System crystal is 21312500 Hz (exactly 4×341×15625). Maybe supports both 50 and 60 Hz operation?
UM6561xx-1	21.48 MHz	NTSC	UMC-made NES-on-a-chip for NTSC. PPU half believed to be mostly identical to UA6528. Revisions "xx" F, AF, BF, CF known. Emphasis is much stronger than on UA6528 and official PPUs.
UM6561xx-2	?	PAL-B	UMC-made NES-on-a-chip for PAL-B. PPU half believed to be mostly identical to UA6538. Revisions "xx" F, AF, BF, CF known. AF revision has graphical and timing glitches in Prince of Persia perhaps caused by sprite 0 hit being missed for reasons not yet understood. F and BF revisions are not affected. Emphasis is much stronger than on UA6538 and official PPUs.
TA-02N	21.48 MHz	NTSC	??-made die-mask clone of 2C02G, despite the "6528" label.[5] Chip underside also has two codes of currently unknown purpose.
TA-02NP	?	PAL	??-made PPU, Dendy compatible (not a 6538 clone). Pins 14-17 are background color in, like normal 2C03[6]
TA-02NPB	?	Select	??-made PPU, "NTSC for PAL-B" (NPB) Dendy-compatible. Has a unique video switching capability shared between it and other TA-02NPx chipsets, as well as the WDL chipset.
TA-08	?	SECAM	??-made PPU, Second source alternative to UM6558. Color Data bus bit-reversed?
MG-N-502	?	NTSC
MG-P-502	26.60 MHz	PAL-B	Micro Genius / TXC. Die shot matches UA6538.
1818N	21.48 MHz	NTSC	??-made NES-on-a-chip, NTSC timing.
1818P	26.60 MHz	PAL-B	SiS-made[7]] NES-on-a-chip.[8]. Requires external 2KiB RAMs for CPU and PPU. UA6538 timing.
PM02-1	?	PAL-M	Gradiente-made variant for PAL-M (Brazil). [9]
VT01	?	Select	V.R.Tech-made clone of UA6561. Only seen as chip-on-board. Supports composite out; RGB out; or 2bpp STN LCD, either greyscale or red/cyan checkerboard, 240 px wide, 80/120/160/240px tall. The chip was extended significantly in VT02 and newer NOACs.
GS87008	21.48 MHz	NTSC	(Goldstar??)-made NTSC clone [10] Found in MicroGenius clone with PAL/NTSC switch.^[1]
KC-6078	26.60 MHz	PAL-B	Found in MT777-DX famiclone, behaves exactly like UA6538
6022	?	NTSC	??-made NTSC clone. Details unknown.
2010	?	?
2A02E	?	?	Dendy Timing. On a large solid-color background (especially yellow, green), every other row has a horizontal stripe.
02	?	?
GT-01	?	?	In PLCC
HA6538	?	PAL	PAL video output.
SENITON 6538U-8	?	PAL	PAL video output.
SENITON 6538A	?	PAL	PAL video output.
6528 (WDL6528)	21.48 MHz	NTSC	WDL made chip. NTSC timing. Palette like RP2C02. Often referred to as "WDL6528".
8Z02	21.48 MHz	NTSC	Found in Family Game by NTDEC.
TECH28	?	PAL	PAL video output.