Radeon RX 7000
Radeon RX 7000 Base design
There are several differences between RDNA 2 and RDNA 3. However, the RX 7000 series has 20% more Compute Units (CUs) and doubles the number of FP32 (32-bit floating point) computing functionality per CU. AMD still employs 64 Streaming Processors (SPs) per CU; however, that amount is more like double in many respects. Each GPU holds a single GCD chiplet (Graphics Compute Die) and up to six MCD chiplets (Memory Cache Die). The XTX has a fully enabled GCD and six MCDs. In contrast, the 7900 XT has a partially enabled GCD with 84 CUs and five MCDs — the sixth MCD slot will remain occupied, either with a dummy piece of silicon or a non-functional MCD (non-functional because it is fused off, though it may or may not be defective). Transistor counts have increased significantly compared to the previous generation RX 6000-series, with 58 billion total transistors compared to only 26.8 billion on Navi 21. Theoretical computation has increased by up to 159% compared to the RX 6950 XT, with 61.4 teraflops on the XTX card and 51.6 teraflops on the XT variant.
Radeon RX | 7900 XTX | 7900 XT | 7900 GRE | 7800 XT | 7700 XT | 7600 |
---|---|---|---|---|---|---|
Architecture | RDNA 3 | |||||
GPU | Navi 31 | Navi 32 | Navi 33 | |||
Design | Chiplet1 × GCD + 6 × MCD | Chiplet1 × GCD + 5 × MCD | Chiplet1 × GCD + 4 × MCD | Chiplet1 × GCD + 3 × MCD | Monolithic | |
production | TSMC N5 (GCD)TSMC N6 (MCD) | TSMC N6 | ||||
transistors | 57,7B | 28.1B | 13,3B | |||
chip size | 300 mm (GCD) | 200mm (GDC) | 204 mm² | |||
CU/MS | 96 | 84 | 80 | 60 | 54 | 32 |
FP32-ALUs | 6.144/12.288 | 5.376/10.752 | 5.120/10.240 | 3.840/7.680 | 3.456/6.912 | 2.048/4.096 |
RT Core | 96, 2nd Gen | 84, 2nd Gen | 80, 2nd Gen | 60, 2nd Gen | 54, 2nd Gen | 32, 2nd Gen |
ki-kerne | 192, 1st Gen | 168, 1st Gen | 160, 1st Gen | 120, 1st Gen | 108, 1st Gen | 64, 1st Gen |
Game Tact | 2.300 MHz | 2.000 MHz | 1.880 MHz | 2.124 MHz | 2.171 MHz | 2.250 MHz |
Boost Rate | 2.500 MHz | 2.400 MHz | 2.245 MHz | 2.430 MHz | 2.544 MHz | 2.625 MHz |
FP32 processing power | 61,4 TFLOPS | 51,6 TFLOPS | 46 TFLOPS | 37,3 TFLOPS | 35,2 TFLOPS | 10,8 TFLOPS |
FP16 processing power | 122.9 TFLOPS | 103,2 TFLOPS | 92 TFLOPS | 74,6 TFLOPS | 70,3 TFLOPS | 21,5 TFLOPS |
texture units | 384 | 336 | 320 | 240 | 216 | 128 |
ROPs | 192 | 196 | 64 | |||
Storage | 24GB GDDR6 | 20GB GDDR6 | 16 GB GDDR6 | 12GB GDDR6 | 8 GB GDDR6 | |
memory throughput | 20 Gbps | 18 Gbps | 19,5 Gbps | 18 Gbps | ||
memory interface | 384 Bit | 320 Bit | 256 Bit | 192 Bit | 128 Bit | |
memory bandwidth | 960 GB/s | 880 GB/s | 576 GB/s | 624 GB/s | 432 GB/s | 288 GB/s |
Infinity-Cache | 96 MB, 2nd Gen | 80 MB, 2nd Gen | 64 MB, 2nd Gen | 48 MB, 2nd Gen | 32 MB, 2nd Gen | |
slot connection | PCIe 4.0 ×16 | PCIe 4.0 ×8 | ||||
TBP | 355 Watt | 315 Watt | 260 Watt | 263 Watt | 245 Watt | 165 Watt |
Price | $999 | $899 | $799 | $499 | $449 | $299 |