Intel’s Thirteenth-Gen Raptor Lake processors will convey extra cores, extra connectivity, a revamped core structure, help for PCIe 5.0 SSDs, and probably even a rumored 6.0 GHz peak increase clock to bear. And that is to not point out any potential IPC enhancements. These chips will arrive in October to sq. off with AMD’s Zen 4 Ryzen 7000 processors, setting the stage for a fierce battle for desktop PC supremacy — notably for the crown of the finest CPU for gaming because the Intel vs AMD rivalry enters a brand new stage.
Intel’s Alder Lake introduced the corporate again from what had been a sluggish erosion of its management place in our CPU benchmarks rankings as AMD relentlessly iterated on its Ryzen processor lineup. AMD’s string of improvements finally culminated in an embarrassing lack of the efficiency crown for Intel because the Ryzen 5000 processors outclassed Intel’s chips in each efficiency, value, and energy metric that mattered again in 2020, capping Intel’s decline from grace after incessant delays shifting to its oft-delayed and seemingly doomed 10nm course of node.
Alder Lake righted the ship. These chips introduced the perfect of Intel’s newly re-worked 10nm course of, now re-named ‘Intel 7,’ enabling greater clock charges and decrease energy consumption, paving the best way for Raptor Lake. Intel will etch the Raptor Lake processors on a refined model of that very same course of node and pair it with its newly-revamped x86 hybrid structure, a design that mixes a mixture of bigger high-performance cores paired with smaller high-efficiency cores.
Like its predecessor, Raptor Lake may even help disruptive new options like PCIe 5.0 and DDR5 however preserves DDR4 help for less-expensive construct choices. Raptor Lake may even drop into current motherboards to supply an improve path for Alder Lake customers, however there will likely be new 700-series motherboards at launch that provide higher connectivity choices. Intel can be introducing extra CPU overclock options for Raptor Lake, too.
Though Raptor Lake is clearly on the cusp of coming to market this 12 months — we have even seen chips bought at public sale and benchmarks within the wild — Intel has been uncharacteristically silent about its pending line of chips for desktop PCs. The truth is, the corporate has mentioned extra about its next-next-gen Meteor Lake chips than it has about Raptor Lake. That hasn’t stopped us from gathering all the data we all know from official and unofficial sources into this text. We even have footage of the brand new processor die. We’ll replace the article as we study extra, however here is what we all know up to now.
Intel Thirteenth-Gen Raptor Lake Collection at a Look
- Codename Raptor Lake
- Desktop and cell launch in This fall 2022 (October 20)
- As much as 24 cores and 32 threads on ‘Intel 7’ course of node
- As much as 8 Raptor Cove Efficiency cores (P-Cores) and 16 Gracemont Effectivity cores (E-Cores)
- Raptor Lake-S (65W to 125W desktop) and Raptor Lake-P (15 to 45W cell) confirmed
- Rumored 5.8 GHz increase
- As much as 36MB of L3 Cache (20% improve), as much as 32MB L2 (2.3x improve)
- Twin-Channel DDR4-3200 and DDR5-5600 reminiscence help, x16 PCIe 5.0 and x4 PCIe 4.0 interface, Thunderbolt 4 / USB 4
- Help for PCIe 5.0 M.2 SSDs and AI M.2 Module
- “As much as double-digit efficiency increase”
- Socket LGA 1700, backward suitable with current coolers, cell chips are BGA suitable
- 700-Collection Chipset: Z790, H770, B760 Motherboards
- Chipset: As much as 20 PCH PCIe 4.0 and eight PCIe 3.0
- Enhanced CPU overclocking options, together with per-core and Environment friendly Thermal Velocity Increase
Intel Thirteenth-Gen Raptor Lake Launch Date Window
Intel hasn’t given an official launch date for Raptor Lake but, however all indicators level to a This fall 2022 launch for the desktop processors. Intel has additionally introduced that it’s going to convey the Raptor Lake cell CPUs to market this 12 months, too. Our personal sources inform us that we’ll see an announcement for the desktop chips in late September, however the launch will land in mid-to-late October. Naturally, that is early data and topic to alter — distributors usually push again timelines — however the chips are mentioned to be at present scheduled to launch in that timeframe.
Our data was subsequently backed up by information from Enthusiastic Citizen, a leaker with a stellar monitor document, that claims Intel will announce the total Raptor Lake particulars on the Intel Innovation occasion on September 28 in tandem with evaluations being printed that very same day. Preorders will purportedly start on October 13, and retail availability for the Ok-series SKUs will arrive on October 20 together with the Z790 motherboards. Non-Ok CPUs and the B760 and H760 chipsets reportedly will not be introduced till January 5, 2023, at CES, with retail availability within the second half of that month.
Intel can be accelerating its cell launch, with CEO Pat Gelsinger just lately asserting that the cell platforms will launch this 12 months. We have even seen benchmarks emerge from the cell Raptor Lake processors. An engineering pattern of the flagship Core i9-13900K has been auctioned on-line, and a number of other evaluations of production-quality silicon have leaked on-line. We’ll cowl these benchmark outcomes under, however all of those indicators level to an impending launch.
Intel Thirteenth-Gen Raptor Lake Specs and Options
Official details about Raptor Lake is skinny on the bottom. Intel’s most complete official itemizing of options touts as much as 24 cores and 32 threads on the ‘Intel 7’ course of, which it claims delivers an ‘as much as double-digit efficiency increase.’ Raptor Lake additionally comes with enhanced overclocking options, help for an AI M.2 module, and the chips are suitable with Alder Lake methods.
The Raptor Lake chips could have Efficiency Cores (P-cores) with a purportedly new microarchitecture, rumored to be named Raptor Cove (extra under). These cores are designed for single- and lightly-threaded duties, like gaming and productiveness workloads. The Effectivity Cores (E-cores) additionally bear indicators of a revamped microarchitecture, however these cores are nonetheless rumored to have the Gracemont design. These cores step in for heavily-threaded workloads, background duties, and multi-tasking.
| Worth | Cores / Threads (P+E) | Base / Increase Clock (GHz) | Cache (L2/L3) | TDP / PBP / MTP | Reminiscence | |
| Ryzen 9 7950X | $699 | 16 / 32 | 4.5 / 5.7 | 80MB (16+64) | 170W / 230W | DDR5-5200 |
| Core i9-13900K / KF | ? | 24 / 32 (8+16) | 3.0 / 5.8 | 64MB (32+36) | 125W / 253W | DDR4-3200 / DDR5-5600 |
| Core i9-12900K / KF | $589 (Ok) – $564 (KF) | 16 / 24 (8+8) | 3.2 / 5.2 | 44MB (14+30) | 125W / 241W | DDR4-3200 / DDR5-4800 |
| Ryzen 9 7900X | $549 | 12 / 24 | 4.7 / 5.6 | 76MB (12+64) | 170W / 230W | DDR5-5200 |
| Core i7-13700K / KF | ? | 16 / 24 (8+8) | 3.4 / 5.4 | 54MB (24+30) | 125W / 253W | DDR4-3200 / DDR5-5600 |
| Core i7-12700K / KF | $409 (Ok) – $384 (KF) | 12 / 20 (8+4) | 3.6 / 5.0 | 37MB (12+25) | 125W / 190W | DDR4-3200 / DDR5-4800 |
| Ryzen 7 7700X | $399 | 8 / 16 | 4.5 / 5.4 | 40MB (8+32) | 105W / ? | DDR5-5200 |
| Ryzen 5 7600X | $299 | 6 / 12 | 4.7 / 5.3 | 38MB (6+32) | 105W / ? | DDR5-5200 |
| Core i5-13600K / KF | ? | 14 / 20 (6+8) | 3.5 / 5.1 | 44MB (20+24) | 125W / 181W | DDR4-3200 / DDR5-5600 |
| Core i5-12600K / KF | $289 (Ok) – $264 (KF) | 10 / 16 (6+4) | 3.7 / 4.9 | 29.5MB (9.5+20) | 125W / 150W | DDR4-3200 / DDR5-4800 |
| Core i5-13400 / F | ? | 10 / 16 (6+4) | 3.4 / ? | 24MB | 65W / ? | DDR4-3200 / DDR5-5600 |
| Core i5-12400 / F | $199 – $167 (F) | 6 / 12 (4+0) | 4.4 / 2.5 | 25.5MB (7.5+18) | 65W / 117W | DDR4-3200 / DDR5-4800 |
| Core i5-13100 | ? | 4P + 0E | 4 Cores / 8 Threads | ~3.4 GHz / ? | ? | DDR4-3200 / DDR5-5600 | 12MB |
Above, we are able to see the Core i9, i7, and i5 flagships from Intel’s previous-gen household, together with what we all know in regards to the new Raptor Lake fashions from the assorted leaks and knowledge we have collected. Remember that not all the Intel Thirteenth-Gen information within the above desk has been formally confirmed. We have additionally added within the newly-announced Ryzen 7000 processors.
Intel will purportedly solely launch Core i9, i7, and a few i5 fashions with the Raptor Lake die, whereas some i5 and all i3, Pentium, and Celeron will leverage refreshed previous-gen Alder Lake chips (Intel took an identical strategy with its Eleventh-Gen Rocket Lake processors).
The Core i9-13900K will include a complete of 24 cores representing a mixture of eight P-cores and 16 E-cores. That is a further eight E-cores over the previous-gen flagship (however the identical variety of P-cores). These extra E-cores come from a brand new bigger 8+16 die (8 P-core + 16 E-core) that Intel will use for the Core i9, i7, and Ok-series i5 chips solely. This bigger die comes with extra cache capability for the cores (extra on that within the structure part), however Core i3 and under could have the identical quantity of cache as discovered with the present Alder Lake fashions.
Intel’s Core i7-13700K just lately emerged within the Geekbench database, displaying the chip has eight P-cores and eight E-cores, for a complete of 16 cores and 24 threads, paired with 30MB of L3 cache. These primary specs match the previous-gen Core i7-12700K, however we aren’t conscious of any variations within the official clock speeds but. Nonetheless, this processor has been overclocked to a wonderful 6 GHz and 6.1 GHz, as seen in two separate benchmarks, the latter with liquid cooling.
It is clear that Intel has modified its Core i5 E-core technique. The present Ok-series Core i5, the Core i5-12600K, comes with 4 E-cores, however a current leak factors to the Core i5-13600K having 4 extra E-cores, for a complete of eight. Historically, the remainder of the non-Ok Core i5 fashions, just like the Core i5-12400, haven’t got E-cores. Intel is alleged to be including E-cores to the non-Ok Core i5 Raptor Lake chips, which may make its mid-range Core i5 all of the stronger [Edit: as you can see below, this has proven to be true].
Above we are able to see leaked Raptor Lake specs for the whole Thirteenth-Era Intel household. Take these with a grain of salt, however this comes from a identified leaker with entry to early {hardware}. You will discover the non-Ok Core i5 fashions have been revamped with e-cores, however they’re bolded on the chart as a result of these will purportedly additionally include previous-gen Alder Lake dies, identical to the i3 fashions. Time will inform.
We haven’t any concrete clock velocity data on the Raptor Lake SKUs but, because the benchmarks we have seen of Engineering Samples (ES) chips aren’t consultant of the ultimate clock speeds. These chips undergo completely different revisions, like ES1, ES2, and so forth, and the rumor mill factors to as much as 5.5 GHz with the ES3 model of the Core i9-13900K — however keep in mind that these aren’t the ultimate clock speeds. We have since seen a 5.7 GHz peak clock price in a leaked benchmark, and that was later eclipsed by a 5.8 GHz pattern.
As a reminder, Intel has redefined its energy terminology to have a ‘Processor Increase Energy’ (PBP) worth representing the assured base efficiency degree (PL1). This replaces TDP. CPU-Z entries present the Core i9-13900K with a 125W PBP, together with screenshots for a 65W variant with the identical 8+16 core counts that’s doubtless the 65W Core i9-13900. Intel additionally lists a ‘Most Turbo Energy’ (MTP) specification that quantifies the facility consumption throughout Turbo Increase (PL2). That is rumored to stay at 241W for the Core i9-13900K. By assigning the identical peak PBP for the Core i9, it is truthful to imagine that the Core i7, i5, and i3 ranges could have comparable energy limits because the Alder Lake processors. Which means we’ll see 125W, 65W, and 35W variations for desktop PCs and sub-45W for the cell chips.
There are additionally rumblings of a 350W ‘Turbo Mode,’ however take these reviews with some salt. It is notable that Intel’s present chips may function at such excessive wattage if the facility limits are eliminated. As such, this is likely to be some type of new advertising tactic of an current function, but it surely feels uncertain that Intel would name consideration to egregious energy consumption.
Intel has confirmed that the Alder Lake chips will drop into the LGA 1700 socket, which means they’re backward suitable with the present 600-series chipsets, and older coolers are additionally ahead suitable with the brand new Raptor Lake motherboards. As well as, the 16 PCIe 5.0 PCIe lanes coming off the CPU can now be cut up into twin x8 preparations, thus enabling help for PCIe 5.0 M.2 SSDs. We’ll dive a lot deeper into this subject within the motherboard part.
The above slide was just lately leaked, confirming the main points we have already shared on this article. The Raptor Lake chips help DDR4-3200, identical to the previous-gen chips, and are at present certified for DDR5-5200, which is quicker than the DDR5-4800 with previous-gen chips. Nonetheless, we’re informed that Intel is engaged on getting DDR5-5600 certified in time for launch, so reminiscence information switch charges aren’t remaining.
Intel’s previous-gen Alder Lake chips marked the trade debut of DDR5 reminiscence, however the brand new reminiscence usually does not present huge speedups over DDR4, particularly at inventory settings. That would change, although, as two particular person units of benchmarks have proven spectacular positive aspects, with the Core i7-13700K being both 20% or 60% quicker than DDR4 in a number of leaked benchmarks. We’ll have to attend to see how that appears in the actual world, however DDR5 does look promising on the Raptor Lake platform.
As earlier than, Raptor Lake could have a dual-channel reminiscence interface. Intel may even carry over ECC reminiscence help for its consumer-class W-series motherboards. We are going to nonetheless additionally see a sophisticated DDR5 reminiscence help matrix that sees speeds decline if the motherboard has a couple of DIMM slot per channel (SPC), should you populate greater than 1 DIMM per channel (DPC) on boards which have two SPC, or primarily based on various DIMM ranks. Nonetheless, now you can count on elevated velocity with every sort of configuration. Non-Ok and Core i3 chips may even have slower supported peak speeds than the Ok-series chips.
As evidenced by Intel’s personal statements, Raptor Lake will proceed to have a heavy concentrate on CPU overclock options. Tantalizing unofficial particulars just lately emerged through an replace to Intel’s personal eXtreme Tuning Utility (XTU). Intel has added help for ‘future platforms’ to leverage added help for per-core and package-level Thermal Velocity Increase (TVB) tech, which permits the processor to spice up greater than the bottom specification if the chip is below a sure temperature threshold. Intel additionally added help for a brand new sort of TVB, referred to as ‘Effectivity TVB.’ We have additionally heard rumblings of a 6 GHz clock price for a Core i9-13900KS, the successor to the Core i9-12900KS, enabled by this new increase. Take this with a grain of salt, as it’s primarily based on a single declare.
The Raptor Lake iGPU is alleged to be primarily based on the identical Xe-LP Gen 12.2 structure discovered with Alder Lake. A current benchmark emerged displaying the Raptor Lake Core i9-13900’s iGPU with 32 EUs operating at 1.65 GHz in an OpenGL benchmark. Which means the iGPU has the identical variety of cores however runs 100 MHz (6.5%) quicker than the previous-gen model. Except for minor clock velocity enhancements, we do not count on any significant architectural adjustments to the iGPU. Intel did just lately add help for Raptor Lake-P (cell) and Raptor Lake-S (desktop) processors to its media driver, thus tipping its hat that cell variants of Raptor Lake are additionally within the works.
Intel has additionally teased a brand new AI accelerator that can slot into an M.2 slot. It is onerous to inform what sensible objective this may serve for many makes use of, although some edge use-cases would possibly profit. Intel hasn’t shared any extra details about this product, and there hasn’t been every other data, so we’ll have to attend to study extra.
We’ve got loads of different architectural data to share within the sections under. However first, let’s take a break to take a look at the leaked Raptor Lake benchmarks.
Intel Thirteenth-Gen Raptor Lake vs AMD Zen 4 Ryzen 7000
| AMD Zen 4 Ryzen 7000 | Intel Raptor Lake | |
| Launch Date: Rumored | September twenty seventh | Mid- to Late-October |
| Node / Design | TSMC 5nm Compute die, 6nm I/O DIe | Intel 7 – Monolithic Die |
| Cores / Threads | As much as 16 Cores / 32 Threads | As much as 8P + 16E | 24 Cores / 32 Threads |
| Peak Clocks | ~5.7 GHz | 5.8 Noticed / 6.0 Rumored |
| TDP / PBP / MTP | 170W / 230W | 125W / 241W |
| Reminiscence | DDR5 Solely (No DDR4 help) | DDR4-3200 / DDR5-5600 |
| PCIe | PCIe 5.0 – 24 Lanes | PCIe 5.0 x16, PCIe 4.0 x4 (SSD) |
| Graphics | ? | ? |
This is what we expect the Raptor Lake vs Ryzen 7000 competitors will appear to be, although keep in mind that not all specs are confirmed. Each chipmakers are pushing clocks to the very best we have seen with their fashionable chips, and that can naturally include elevated energy consumption, notably in multi-threaded work.
AMD’s Zen 4 Ryzen 7000 chips solely help DDR5 reminiscence, whereas Raptor Lake helps DDR4 and DDR5. That offers Intel a leg up within the general system price class, as DDR5 nonetheless instructions a value premium. Nonetheless, we not see DDR5 shortages, and costs proceed to plummet as extra provide comes on-line and demand recedes. Which means the pricing delta ought to get noticeably slimmer, although we absolutely count on DDR5 to all the time command a premium.
Intel Thirteenth-Gen Raptor Lake Die Footage
We have already seen loads of Thirteenth-gen Intel Raptor Lake benchmarks emerge (extra about that under), so it is clear that loads of Raptor Lake chips are floating round within the wild. Now we have now footage of the particular die, too.
A video posted to Bilibili reveals the size of the Raptor Lake die, which measures 23.8 x 10.8mm. Which means the chips have a barely wider and for much longer die than the Twelfth-gen Alder Lake processors.
| Die Space | Die Dimensions | Cores | Course of | |
| Raptor Lake Core i9-13900K | 257 mm^2 | 23.8 x 10.8 mm | 8 P-Cores | 16 E-Cores | Intel 7 |
| Alder Lake Core i9-12900K | 208 mm^2 | 20.4 x 10.2 mm | 8 P-Cores | 8 E-Cores | Intel 7 |
| Rocket Lake Core i9-11900K | 281 mm^2 | 24 x 11.7 mm | 8 P-Cores | 14nm |
| Comet Lake Core i9-10900K | 206 mm^2 | 9.2 x 22.4 mm | 10 P-Cores | 14nm |
Right here we are able to see the 11900K, 12900K, and Raptor Lake Core i9-13900K collectively. Intel elevated the core rely for Raptor Lake by shifting from a peak of eight E-cores to 16 by including two extra quad-core clusters, elevated the L2 cache by 60%, and in addition elevated the L3 cache capability (extra particulars under). All of those elements necessitated an 8+16 die that’s 49mm^2 bigger than the previous-gen 8+8 die. Every E-core cluster weighs in at roughly ~8.65 mm^2, which means solely 17.3 mm^2 of the die improve is devoted to the E-cores, whereas the remaining is allotted to the bigger L2 and L3 caches. That is an actual testomony to the extremely area-efficient nature of the e-cores.
The bigger chip measurement quantities to extra pricey manufacturing as a result of a bigger die leads to fewer dies per wafer. Intel has mentioned that it’s going to increase chip costs as a result of inflation, but it surely is not clear how that can impression the Raptor Lake household — Intel will both improve pricing much more to offset the bigger die, or it can increase costs additional to offset the elevated manufacturing prices.
Intel Thirteenth-Gen Raptor Lake Benchmarks and IPC
Intel demoed Raptor Lake at its Investor Day 2022 utilizing a chip with eight P-Cores and 16 E-Cores, so it was a flagship Core i9 mannequin. The demo of a Blender and After Results workload did not give us any efficiency information to work with. Fortunately, we have seen loads of leaked benchmarks that give us a greater thought of what Raptor Lake will appear to be. Nonetheless, as with all leaked take a look at outcomes, take all the benchmarks under with a grain of salt — they’re usually performed with pre-production silicon and unoptimized platforms, which means remaining outcomes may very well be materially higher.
The most effective comes from a Core i9-13900K video overview posted in China which you could watch right here. For non-Chinese language audio system, we have now the rundown of these checks under:
| Benchmark | Core i9-13900K | Core i9-12900KF | % Change |
| CPU-Z bench 1T | 892.2 | 815.5 | +9.4% |
| CPU-Z bench nT | 16,606 | 11,348 | +46.3% |
| Geekbench 1T | 2,133 | 1,939 | +10% |
| Geekbench nT | 23,701 | 19,304 | +22.8% |
| Cinebench R23 1T | 2,206 | 1,940 | +13.7% |
| Cinebench R23 nT | 37,385 | 26,939 | +38.8% |
| 3D Mark Timespy CPU | 23,839 | 20,121 | +18.5% |
| PugetBench Premiere Professional | 1,213 | 1,003 | +20.1% |
The examined 13900K is a Qualification Pattern (QS), which means it’s almost equivalent to the retail fashions. The 13900K had a 3 GHz base and 5.8 GHz increase for the P-cores and was examined on a ROG Maximus Z690 Excessive, however with out optimized firmware. The chip was dialed into the proper 125W/250W energy vary for the checks.
On common, throughout a broad number of checks, the 13900K was 10% quicker in single-thread and 35% quicker in multi-threaded workloads than the previous-gen 12900K. That is fairly a powerful multi-threaded acquire on condition that it stems nearly totally from the additional eight E-Cores and cache. The checks expose even greater positive aspects in a number of the particular person threaded benchmarks, like CPU-z and Cinebench, which registered a whopping 46% and 40% enchancment, respectively. The outlet later added energy measurements with the proper energy settings, displaying a extra reasonable energy draw of 253W. That resulted within the chip operating at 77C throughout an prolonged take a look at.
| Core i9-13900K | Core i9-12900K | AMD 5950X | |
| Cores / Threads | 24C/32T | 16C/24T | 16C/32T |
| 1T rating | 2,133 | 1,987 | 1,686 |
| Multi-thread rating | 23,701 | 17,272 | 16,508 |
Above, we are able to see the outcomes from a unique Geekbench 5 take a look at submission that confirmed the Core i9-13900K, which was paired with 32GB of DDR4-6400 reminiscence, boosting to five.7 GHz. This take a look at confirmed the 13900K as being 7% quicker in single-thread and 37% quicker in multi-threaded checks than the previous-gen 12900K. Extra importantly, it was 26% quicker in single-and 43% quicker in multi-thread than the Ryzen 9 5950X. Geekbench later eliminated this end result from its database as a result of a timing error in the course of the benchmark run, so it won’t signify the precise efficiency or increase clock speeds. Given the take a look at outcomes we noticed within the full overview we shared above, this seems to be near correct.
| Core i5-13600K | Core i5-12600K | Core i9-12900K | Ryzen 9 5950X | |
| CPU-Z 1T | 830 | 753 | 803 | 684 |
| CPU-Z nT | 10,032 | 6,692 | 10,921 | 12,078 |
| CB23 | 1,387 | 1,886 | 1,965 | 1,652 |
| CB23 nT | 24,420 | 17,161 | 27,287 | 26,271 |
The Core i5-13600K has purportedly been examined by a identified leaker with entry to early {hardware}. These outcomes level to an distinctive quantity of efficiency positive aspects from the additional 4 e-cores over the previous-gen Core i5. The Core i5-13600K is 8% and 44% quicker within the CPU-z single- and multi-threaded checks, respectively, than the previous-gen i5-12600K.
Moreover, the chip beat the Ryzen 9 5950X in each single- and multi-threaded CPU-z benchmarks, and solely lagged ~7% within the Cinbench threaded benchmark. A subsequent itemizing within the GeekBench database was simply as spectacular, displaying comparable positive aspects. Making an allowance for that the Core i5 will retail for round $300 and the 5950X retails for over $500, these outcomes are spectacular. Total, the Core i5-13600K is within the vary of the current-gen Core i7 in these checks.
Lastly, EXP Evaluation additionally posted a full overview, but it surely used an Engineering Pattern (ES) Core i9-13900 to check to a Core i9-12900K locked to the identical frequencies, thus giving us an thought of potential IPC positive aspects. Nonetheless, do not forget that that is nonetheless not remaining silicon, and the chip was examined on a motherboard that is not optimized for Raptor Lake but. The massive takeaway is that, on the identical clock speeds, the Raptor Lake mannequin was roughly 20% quicker in non-gaming multi-threaded work than the Core i9-12900K.
Extra impressively, the Raptor Lake chip took a 12% lead in single-threaded functions, suggesting we’ll see a major IPC efficiency improve that can prolong Intel’s lead additional over the Ryzen 5000 chips and probably protect the lead when AMD’s Zen 4 involves market. The outlet additionally ran gaming benchmarks, however the distinction between the 2 chips fell throughout the margin of error.
A number of different rounds of gaming benchmarks have emerged, with one collection of checks displaying considerable positive aspects in minimal body charges, however no important positive aspects in different metrics. That is hardly stunning as a result of video games are very delicate to poor latency of any type, like unrefined reminiscence help or motherboard BIOSes. As such, we would not take the gaming outcomes too critically till we see the ultimate silicon emerge.
Intel Thirteenth-Gen Raptor Lake Structure
Intel hasn’t shared the names of the chip microarchitectures that it’s going to etch onto the Intel 7 node for the Raptor Lake CPUs, however rumors point out the P-Cores will use ‘Raptor Cove’ naming whereas the E-Cores will follow Gracemont. Nonetheless, each cores do have considerably extra L2 cache, suggesting a re-working of the underlying designs. We warning that we will not discover any indication of the ‘Raptor Cove’ codename being actual, so take that exact naming with a pinch of salt. We do know that software program will handle the Raptor Cove cores the identical because the Golden Cove structure discovered on the previous-gen Alder Lake processors, however we’ll have to attend to see if Intel makes use of a unique naming conference.
One factor is for positive: Intel has considerably re-worked the cache hierarchies of each sorts of cores, so we must always count on new code names. The Raptor Lake chip will now share as much as 36 MB of L3 cache, representing a rise of 6MB over the earlier era. Nonetheless, this merely represents the addition of two extra 3MB L3 cache clusters and never a rise in per-core cache capability.
As you’ll be able to see within the picture of the Alder Lake die above, the hybrid structure pairs a 3MB slice of L3 cache (labeled as LLC) subsequent to every ‘block’ of cores. The P-cores, in darkish blue, every have an L3 slice close by, whereas the E-cores, in mild blue, are available quad-core clusters that even have a 3MB slice of L3 cache close by. These slices of cache are shared amongst all cores.
The brand new Raptor Lake 8+ 16 die (lined within the earlier part) comes with 16 E-Cores. These extra E-Cores would equate to the above Alder Lake die being stretched to accommodate two extra mild blue quad-core clusters of E-Cores. These two clusters would include two extra 3MB L3 cache slices, bringing the overall capability to the 36MB we see with the Core i9-13900K.
Apparently, Alder Lake’s cache scheme launched a brand new means for Intel to disable cache for decrease core rely SKUs. Prior to now, to create lower-end SKUs with fewer cores, Intel disabled total slices of L3 cache (the LLC blocks within the picture) in lockstep with any cores it disabled. With Alder Lake, Intel transitioned to not disabling any total slice of L3, as a substitute merely disabling some banks in every slice, thus decreasing per-slice capability (from 3MB to 2MB, for example). This makes loads of sense because it retains stops on the ring bus lively, and it additionally permits for creating SKUs with greater cache capability than we’d sometimes count on primarily based on the variety of disabled cores.
As such, we will not merely guess the quantity of L3 cache per Raptor Lake chip primarily based on the core rely alone. As an alternative, we’ll have to attend to see extra detailed data. For now, we merely have affirmation that the Core i9-13900K and Core i9-13900 will include 36MB of complete L3 cache. Nonetheless, Intel’s new L3 cache coverage leaves room for cache capability enhancements on all SKUs that leverage the 8+16 die.
Intel has elevated per-core L2 cache capability for each the P-Cores and the E-Cores. The E-Cores see a rise to 2MB of personal L2 cache per core, a 60% improve over the 1.25 MB per core present in Alder Lake. Intel additionally boosted the quantity of L2 cache shared amongst every quad-core cluster of E-Cores to 4 MB, a doubling over the 2MB with Alder Lake. Which means we’ll see as much as 32MB of L2 cache. As we have seen with Intel’s earlier chips that received a giant improve in L2 capability, this sort of enchancment tends to end in higher efficiency in multi-threaded workloads, however it’s potential that it may end in greater IPC in some workloads as a result of protecting the cores fed with extra information. It must also assist free the ring from some site visitors that will in any other case be current for shuffling round L3 information, thus permitting higher scalability.
The L1 cache for each sorts of cores stays the identical (L1i$ 32kB, L1D$ 48Kb for P-Cores — L1i$ 64kB, L2D$ 32kB for E-Cores).
A leaked roadmap listed a brand new function, a Digital Linear Voltage Regulator (D-LVR), alongside the Raptor Lake chips. In line with an Intel patent, this function helps cut back the CPU VID and energy consumed by the cores, probably decreasing energy consumption by as much as 25% in some circumstances. Naturally, on condition that it’s in a pitched battle for efficiency supremacy within the desktop PC house, we’d count on Intel to make use of the headroom afforded from these energy financial savings to ship but more practical energy to the cores. We aren’t positive of this tech’s function within the Raptor Lake chips but, or how efficient it is going to be (if current). We aren’t positive if this will likely be used for the desktop PC or cell chips, or each, but it surely may very well be a part of Intel’s recipe to enhance Raptor Lake’s performance-per-watt.
Raptor Lake’s E-Cores nonetheless don’t help AVX-512, so we count on that Intel will maintain the function disabled, which is odd on condition that AMD’s Ryzen 7000 will absolutely help the extensions. As earlier than, AVX2 and VNNI stay enabled for the E-Cores. We’ve got rather more to study in regards to the Raptor Lake design, however particulars are nonetheless scarce. We’ll replace this part as we study extra.
Intel Thirteenth-Gen Raptor Lake 700-Collection Motherboards, Z790, H770, B760, H610
The Raptor Lake chips will use the identical LGA1700 socket as Alder Lake, which means they’ve the identical socket and pinout, and each Raptor and Alder will likely be suitable with each the 600- and 700-series motherboards, offering fairly a little bit of flexibility for each generations. Nonetheless, should you use a Raptor Lake chip on a 600-series motherboard, you may lose the enhancements in PCH PCIe lane configurations that we’ll define under. All LGA1700 coolers are suitable, so you will not want a brand new CPU cooler for Raptor Lake. Naturally, current 600-series boards would require a BIOS replace to help Raptor Lake, and a number of other motherboard producers, like Asus and ASRock, have already launched the updates.
Raptor Lake brings a major variety of huge steps ahead in connectivity. The previous-gen Alder Lake chips help 16 PCIe 5.0 lanes for a discrete GPU and 4 PCIe 4.0 lanes from the CPU for an M.2 SSD. Those self same lanes are nonetheless current on Raptor Lake, however a brand new connection scheme permits for expanded performance.
For Raptor Lake, motherboard distributors can now cut up the 16 PCIe 5.0 PCIe lanes from the CPU into twin x8 preparations, thus enabling help for PCIe 5.0 M.2 SSDs. This does imply that the connection to the discrete GPU will likely be cut up right into a x8 connection (a change may very well be used right here), however the current PCIe 4.0 hyperlink from the CPU for an M.2 SSD may even stay lively, offering a complete of three M.2 SSD ports that hold immediately off the CPU.
| PCH Connectivity (as much as) | 700-series (Raptor Lake) | 600-series (Alder Lake) |
| PCIe 4.0 | 20 | 12 |
| PCIe 3.0 | 8 | 16 |
| USB 3.2 Gen 2×2 (20G) | 5 | 4 |
Intel itself leaked the 700-series chipset adjustments. The x8 DMI 4.0 connection between the CPU and the chipset (PCH) stays current, however the chipset additionally has enhancements. Prior to now, the PCH supported as much as 16 PCIe 3.0 lanes and as much as 12 PCIe 4.0 lanes, however Intel has elevated the variety of PCIe 4.0 lanes to twenty and diminished the variety of PCIe 3.0 lanes to eight, thus increasing connectivity. Intel additionally elevated the variety of USB 3.2 Gen 2×2 (20G) connections from a peak of 4 to 5.
All different PCH connectivity choices discovered with the Alder Lake motherboards stay unchanged with the Raptor Lake chipset itself. Nonetheless, we’ll see considerably completely different allocations with the Z790, H770, and B760 motherboards primarily based on the brand new options. You may see Intel’s accidentally-released breakdown, which we have independently confirmed, above. The highlighted areas record the adjustments (Z690 = Z790, H670 = H770, B660 = B760).
Intel Raptor Lake Thread Director
Alder Lake pioneered the hybrid period for x86 desktop PCs, and Intel’s Thread Director, which ensures that threads are positioned on the proper cores, is the magic that glues the design collectively. It is doubtless that Raptor Lake will include enhancements to the Thread Director know-how, a results of what Intel says is an upgradeable and tunable design that can enhance over time.
Similar to we noticed with Alder, Raptor Lake will use each quicker and slower cores which are optimized for various voltage/frequency profiles. As such, unlocking the utmost efficiency and effectivity requires the working system and functions to have an consciousness of the chip topology to make sure workloads (threads) land within the right core primarily based on the kind of software.
That is the place Intel’s Thread Director know-how is available in. This hardware-based know-how gives enhanced telemetry information to Home windows 11 to guarantee that threads are scheduled to both the P or E cores in an optimized and clever method, however in a means that is clear to software program.
This know-how works by feeding the Home windows 11 working system with low-level telemetry information collected from throughout the processor itself, thus informing the scheduler in regards to the state of the core, be it energy, thermal, or in any other case. Alder Lake chips may even work tremendous with a bog-standard Home windows 10 working system – current thread-scheduling methods proceed to work with the processors, simply not as nicely. Whereas the chips work, you may miss out on the improved capabilities of Thread Director (that is Home windows 11 solely), which could have a various impression on efficiency and energy consumption primarily based on instruction sort and software utilization fashions. In different phrases, your mileage will fluctuate.
Intel Thirteenth-Gen Raptor Lake Pricing
| Core rely | Itemizing Worth, After Conversion – Rumored | Generational Worth Enhance | |
| Core i9-13900K / KF | 24C / 32T | $725 / $695 | +15% / +18% |
| Core i7-13700K / KF | 16C / 24T | $511 / $483 | +17% / +17% |
| Core i5-13600K / KF | 14C / 20T | $355 / $327 | +16% / +19% |
Intel hasn’t launched any official pricing data but, however we have now seen listings at a Canadian retailer. After changing from CAD to USD, we are able to see that the chips are roughly 15% extra throughout the board (as in comparison with the itemizing for the current-gen fashions on the identical retailer). These may very well be placeholder values, so take them with a grain of salt. Nonetheless, a value improve would not be totally stunning – Intel has introduced that it’s going to improve its pricing as a result of varied elements, like inflation and the growing prices of supplies as a result of provide chain disruptions. Total, we may see the Ok-series Core i9 fashions peaking at~ $725, Core i7 Ok’s at ~$511, and Core i5 Ok’s at ~$355, however Intel hasn’t confirmed pricing but.
Intel has already taken a no-holds-barred bare-knuckle strategy to pricing with Alder Lake because it makes an attempt to steal again market share from AMD, so the corporate does have some room to maneuver. Nonetheless, Intel just lately introduced that it might improve pricing as a result of elevated price of supplies, provide chain disruptions, and inflation, so we are able to count on Raptor Lake to hold a premium over the prior-generation fashions. AMD has introduced pricing for its Ryzen 7000 processors, and other than shaving $100 off of its flagship 16-core Ryzen 9 7950X, it stays largely much like the previous-gen fashions. AMD has a excessive $299 bar for entry to its latest household of chips, displaying that the corporate will proceed to prioritize promoting its priciest high-margin silicon. Intel hasn’t absolutely quantified its deliberate value will increase but, however AMD’s Ryzen 7000 pricing units a tricky bar.
Ultimately, a whole lot of the distinction in pricing between Intel’s Raptor Lake and AMD’s Ryzen 7000 will boil right down to platform prices. We are able to count on a number of the identical tendencies that we noticed with Alder Lake motherboards: DDR5-supporting boards will carry a premium over their DDR4 counterparts as a result of dearer manufacturing methods and supplies required to help the quicker interface. Nonetheless, the choice for DDR4 motherboards will likely be a major benefit over competing Ryzen 7000 platforms, that are DDR5-only. We do count on DDR5 pricing to recede considerably by the point each of those platforms are in the marketplace, however the jury remains to be out on DDR5 pricing.
AMD’s AM5 motherboard product stack will embody provisions for PCIe 4.0-only motherboards, one other solution to save on price, whereas Intel’s lineup will not have that possibility. That may present two tiers of motherboard pricing for AMD, too, however prices will likely be adjusted on a unique axis. It would definitely be fascinating to see which strategy affords the most important pricing deltas between the full-fledged implementations.