Today, as part of its Computex 2026 programme, AMD has announced worldwide availability for the RX 9070 GRE. Technically, this GPU launched over a year ago as a Chinese exclusive, but it is now rolling out globally. Offering what is essentially a cut-down RX 9070 die, the 9070 GRE features fewer cores, 4GB less memory and a narrower memory interface. But what does that mean for performance, and is it still worth buying? We find out today.
Timestamps
00:00 Intro
00:50 Key info, pricing
01:21 Core specs
02:15 Test setup
03:02 Game benchmarks
03:57 Performance summary, comparison vs RTX 5070
05:01 Cost per frame analysis
05:39 Ray tracing performance
07:07 FSR 4 upscaling
07:51 The Sapphire Pulse card
08:32 Thermals and noise
09:23 Power + efficiency
10:12 Closing thoughts
The RX 9070 GRE launched to little fanfare back in April 2025, given it was only on sale in China, but today AMD has announced global availability, with a $549 price tag. In the UK we're told to expect pricing in the £469-479 region, so it's a similar ballpark to the RTX 5070 that can be regularly found at £500.
Like the rest of the RDNA 4 family, there's no reference model, and instead we have tested the Sapphire Pulse edition. This card does come factory overclocked, but as regular readers will know, we manually test all GPUs at reference specifications for these day one reviews, so we can ensure that all testing is ‘apples to apples.'
|
RX 9070 XT |
RX 9070 |
RX 9070 GRE |
RX 9060 XT |
| Architecture |
RDNA 4 |
RDNA 4 |
RDNA 4 |
RDNA 4 |
| Manufacturing Process |
TSMC N4 |
TSMC N4 |
TSMC N4 |
TSMC N4 |
| Transistor Count |
53.9 billion |
53.9 billion |
53.9 billion |
29.7 billion |
| Die Size |
357 mm² |
357 mm² |
357 mm² |
199 mm² |
| Compute Units |
64 |
56 |
48 |
32 |
| Ray Accelerators |
64 |
56 |
48 |
32 |
| Stream Processors |
4096 |
3584 |
3072 |
2048 |
| Game GPU Clock |
2400 MHz |
2070 MHz |
2220 MHz |
2530 MHz |
| Boost GPU Clock |
Up to 2970 MHz |
Up to 2520 MHz |
Up to 2790 MHz |
Up to 3130 MHz |
| ROPs |
128 |
128 |
96 |
64 |
| AMD Infinity Cache |
64MB |
64MB |
48MB |
32MB |
| Memory |
16GB GDDR6 |
16GB GDDR6 |
12GB GDDR6 |
8/16GB GDDR6 |
| Memory Data Rate |
20 Gbps |
20 Gbps |
18 Gbps |
20 Gbps |
| Memory Bandwidth |
645 GB/s |
645 GB/s |
432 GB/s |
322 GB/s |
| Memory Interface |
256-bit |
256-bit |
192-bit |
128-bit |
| Board Power |
304W |
220W |
220W |
160W |
First, let's take a quick look at the specs. The RX 9070 GRE is built on the same Navi 48 die as the RX 9070 and 9070 XT. It measures 357mm², and it's worth pointing out this is a monolithic chip, so AMD is not using a chiplet-based design as per RDNA 3. In total, Navi 48 silicon packs in 53.9 billion transistors.
While a full Navi 48 GPU packs in 64 Compute Units, the 9070 GRE is cut down to 48 CUs, and each CU houses 64 Stream Processors, for a total of 3072 shaders. There's also 48 Ray Accelerators – one per CU – and 96 ROPs.
As for clock speed, the RX 9070 GRE runs faster than the regular 9070, sporting a rated game clock of 2220MHz and a boost clock of up to 2790MHz.
The memory configuration is cut down however, with 12GB of GDDR6 memory clocked at 18Gbps, operating over a 192-bit memory interface, for total memory bandwidth of 432 GB/s. 48MB of Infinity Cache is also present.
Power draw for the RX 9070 GRE is rated at 220W Total Board Power (TBP), but we are using our updated GPU power testing methodology in this review, so read on for our most detailed power and efficiency testing yet.
The post
AMD RX 9070 GRE Review ft. Sapphire first appeared on
KitGuru.
Login
