UK’s trusted IT infrastructure partner since 2003
Servnet
FinanceToolsConfiguratorGet in Touch
Fujitsu PRIMERGY Model Names Explained — networkFujitsu PRIMERGY Model Names Explained — reach
Storage & Hardware

Fujitsu PRIMERGY Model Names Explained

Servnet Editorial · IT infrastructure analysis6 min read
Share

Fujitsu PRIMERGY model names such as RX2540 M8 pack three pieces of information into a short code: the family (TX, RX or GX), a four-digit model number and a generation. The servers are now sold and supported by Fsas Technologies, Fujitsu’s data centre hardware company, but the naming has carried on unchanged. Fsas does not publish a key to the digits, so this guide sets out what the letters mean officially and what the numbers show across the 13 models in the current range.

Reading a PRIMERGY name: RX2540 M8
5RXFamily: TX tower, RX rack, GX GPU42Most sockets in the current range: 1, 2, 4 or 835Line: 3 entry Intel, 4 AMD, 5 two-socket Intel, 7 scale-up24Size within a pair: RX2530 is 1U, RX2540 is 2U1M8Generation; current M8 models use Xeon 6
View the data behind this chart
Reading a PRIMERGY name: RX2540 M8
LayerDetail
RXFamily: TX tower, RX rack, GX GPU
2Most sockets in the current range: 1, 2, 4 or 8
5Line: 3 entry Intel, 4 AMD, 5 two-socket Intel, 7 scale-up
4Size within a pair: RX2530 is 1U, RX2540 is 2U
M8Generation; current M8 models use Xeon 6

The letters: TX, RX and GX

Fsas’s positioning card splits PRIMERGY into three families. TX is the tower range, described as robust and cost-efficient: the TX1310 M5, TX1320 M6, TX1330 M6 and TX2550 M7. RX is the rack range, described as versatile, from the single-socket RX1330 M6 up to the eight-socket RX8770 M7. GX is the GPU range, built for AI, HPC and data-intensive work: the GX2550 M8s and GX2580 M8s.

The first letter is the useful one: T for tower, R for rack, G for GPU. The TX1330 M6 and TX2550 M7 also come as rack-mount base units, but they stay in the TX family.

The first digit: sockets

Across the current range the first digit matches the most processors a model takes. Every model starting with 1 is single-socket: the TX1310, TX1320 and TX1330 towers, the RX1330 and the AMD RX1440. Every model starting with 2 takes two processors: the TX2550, RX2450, RX2530, RX2540 and both GX servers. The RX4770 M8 takes two or four, and the RX8770 M7 up to eight.

That makes the first digit a quick filter when you are reading an asset register: a 1 is a small, single-processor server, a 2 the mainstream dual-socket class, and a 4 or 8 a scale-up machine, which Fsas positions for in-memory databases such as SAP S/4HANA.

The second and third digits: line and size

The second digit groups models into lines. In today’s range a 3 marks the entry single-socket Intel servers (TX13x0 and RX1330), a 4 the AMD EPYC servers (RX1440 and RX2450), a 5 the mainstream two-socket Intel servers (TX2550, RX2530, RX2540 and both GX models), and a 7 the four- and eight-socket scale-up servers (RX4770 and RX8770).

Within a pair, a higher third digit is the bigger box: the RX2530 M8 is 1U and the RX2540 M8 2U; the GX2550 M8s is 4U with PCIe GPUs and the GX2580 M8s 8U with an NVIDIA HGX B300 board. All current models end in 0.

  • •The first- and second-digit patterns hold across all 13 current models; the size pattern shows only in the RX2530/RX2540 and GX pairs (the compact TX1320 M6 is smaller than the TX1310 M5). Fsas documents none of them as a rule, so check the datasheet before relying on one.
  • •Some datasheets name base units SFF (2.5-inch drives) or LFF (3.5-inch drives), for example RX2540 M8 LFF.
The three PRIMERGY families today
TX towerRX rackGX GPUModels472Sockets1 or 21 to 82Entry modelTX1310 M5RX1330 M6 (1U)GX2550 M8s (4U)Top modelTX2550 M7RX8770 M7 (6U)GX2580 M8s (8U)AMD optionNoRX1440, RX2450No
View the data behind this chart
The three PRIMERGY families today
TX towerRX rackGX GPU
Models472
Sockets1 or 21 to 82
Entry modelTX1310 M5RX1330 M6 (1U)GX2550 M8s (4U)
Top modelTX2550 M7RX8770 M7 (6U)GX2580 M8s (8U)
AMD optionNoRX1440, RX2450No

The M number: generation

The M suffix marks the generation. Fsas’s October 2025 launch release speaks of “the PRIMERGY M8 generation”, but Fsas publishes no key to the numbers and they are not shared across every line: the Intel RX2540 is at M8 while the AMD RX2450 is at M2, and the entry towers are at M5 and M6. The two GX servers launched new in 2026 as M8s.

In the current Intel models the M number lines up with the processor platform. The M8 models (RX2530, RX2540, RX4770 and the GX pair) use Xeon 6. The M7 models use the Xeon Scalable processors that came before Xeon 6: the TX2550 M7 takes 4th and 5th Gen Xeon Scalable and the RX8770 M7 4th Gen Xeon Scalable Platinum 84xxH. The entry servers count separately: the M6 models take Xeon E-2400 or Xeon 6300P, and the TX1310 M5 Xeon E-2300 with DDR4 memory.

The two GX servers add a lower-case s after the generation, as in GX2550 M8s. Fsas’s documents do not explain the letter, so treat it as part of the name rather than a feature code.

Reading a name in practice

Put together, RX2540 M8 reads as a rack server, two sockets, the mainstream Intel line, the 2U member of its pair, M8 generation. TX1320 M6 reads as a single-socket entry tower of the M6 generation.

The generation matters most when buying parts or support, because each generation has its own datasheet and parts list, so quote the full name, generation included. For a warranty check you need the serial number from the nameplate rather than the model name; see how to run a Fujitsu warranty check. For the current models in detail, start at Fsas Technologies PRIMERGY and ETERNUS.

Share
Key takeaways
  • ✓TX is tower, RX is rack and GX is GPU, per Fsas’s positioning card.
  • ✓In the current range the first digit matches the maximum socket count: 1, 2, 4 or 8.
  • ✓The second digit groups the line: 3 entry Intel, 4 AMD EPYC, 5 mainstream two-socket Intel, 7 scale-up.
  • ✓The M number marks the generation, but lines are numbered separately; M8 models use Xeon 6.
  • ✓Fsas publishes no naming key, so confirm against the datasheet.
Frequently asked

FAQs — Fujitsu PRIMERGY Model Names Explained

The names

What does RX mean in Fujitsu PRIMERGY?

RX is the rack server family. Fsas’s positioning card splits PRIMERGY into TX towers, RX rack servers and GX GPU servers. The current RX range runs from the single-socket RX1330 M6 to the eight-socket RX8770 M7; see Fsas Technologies for the full line-up.

What does M8 mean on a PRIMERGY server?

It marks the generation. Fsas calls its Xeon 6 servers the PRIMERGY M8 generation, and every current M8 model uses Intel Xeon 6. The numbers are not shared across lines, so the AMD RX2450 is at M2 while the Intel RX2540 is at M8.

Using them

Do I need the model name to check a Fujitsu warranty?

No. The Fsas support portal checks warranty by serial number, which is on the server’s nameplate. The model name and generation matter when ordering parts or a replacement. See our Fujitsu warranty check guide.

Is a PRIMERGY RX2540 M8 the same server as an RX2540 M7?

No. They share a model number, but the generation is a different platform: the M8 is built for Intel Xeon 6 processors with P-cores. Treat each generation as its own server when buying parts or support, and check part numbers against the datasheet for that generation.

Related

Got a question this article didn't answer?

One conversation with an engineer who's done this before. No sales script.

Talk to Servnet →

Talk to a UK specialist

Get expert advice or a no-obligation quote — servers, storage, networking, maintenance, finance and cloud. We reply the same working day.

or call 0800 987 4111