Variant comparison: Difference between revisions
From CMP Wiki
More actions
Import: 8 GB vs 10 GB comparison table, @djpandaman, @wakeupmbrc1, @markshowa, @c133, @lia_oc, @rh3dai, @programmerboi, @cuddylac997, @lxtend, @matstatman, @darkmaniac7, @zorg33 |
Import: 8 GB vs 10 GB comparison table, @djpandaman, @wakeupmbrc1, @markshowa, @c133, @lia_oc, @rh3dai, @programmerboi, @cuddylac997, @lxtend, @matstatman, @darkmaniac7, @zorg33 |
||
| Line 14: | Line 14: | ||
| Memory, unlocked || '''64 GB''' (65536 MiB) {{Status|Proven}} || '''40 GB''' (40960 MiB) {{Status|Proven}}; 48 GB {{Status|Reported}} if the sixth stack is healthy, about 1 card in 5 (6 of 30 cards seen) || {{Source|handle=zorg33|date=2026-10-07|link=https://discord.com/channels/1518634379604136178/1518692084729511967/1557343927336697877}}<br />{{Source|handle=rh3dai|date=2026-10-08|link=https://discord.com/channels/1518634379604136178/1555480078089986138/1557643237282553956|repo=https://github.com/amoghmunikote/cmpunlocker}}<br />{{Source|handle=newman_08462|date=2026-10-09|link=https://discord.com/channels/1518634379604136178/1518695328008638516/1558178969264848927}}<br />{{Source|handle=djpandaman|date=2026-10-17|link=https://discord.com/channels/1518634379604136178/1518692084729511967/1560920383782260797}}<br />{{Source|handle=programmerboi|date=2026-10-25|link=https://discord.com/channels/1518634379604136178/1534370835220004885/1563753581637537904}} | | Memory, unlocked || '''64 GB''' (65536 MiB) {{Status|Proven}} || '''40 GB''' (40960 MiB) {{Status|Proven}}; 48 GB {{Status|Reported}} if the sixth stack is healthy, about 1 card in 5 (6 of 30 cards seen) || {{Source|handle=zorg33|date=2026-10-07|link=https://discord.com/channels/1518634379604136178/1518692084729511967/1557343927336697877}}<br />{{Source|handle=rh3dai|date=2026-10-08|link=https://discord.com/channels/1518634379604136178/1555480078089986138/1557643237282553956|repo=https://github.com/amoghmunikote/cmpunlocker}}<br />{{Source|handle=newman_08462|date=2026-10-09|link=https://discord.com/channels/1518634379604136178/1518695328008638516/1558178969264848927}}<br />{{Source|handle=djpandaman|date=2026-10-17|link=https://discord.com/channels/1518634379604136178/1518692084729511967/1560920383782260797}}<br />{{Source|handle=programmerboi|date=2026-10-25|link=https://discord.com/channels/1518634379604136178/1534370835220004885/1563753581637537904}} | ||
|- | |- | ||
| Memory beyond that || 80 or 96 GB rumoured, nothing shown || 80 or 96 GB not achieved {{Status|Contested}}: the evidence points to aliasing, not real memory || {{Source|handle=djpandaman|date=2026-10-12|link=https://discord.com/channels/1518634379604136178/1534370835220004885/ | | Memory beyond that || 80 or 96 GB rumoured, nothing shown || 80 or 96 GB not achieved {{Status|Contested}}: the evidence points to aliasing, not real memory || {{Source|handle=djpandaman|date=2026-10-12|link=https://discord.com/channels/1518634379604136178/1534370835220004885/1559044323545456671}}<br />{{Source|handle=colo2442|date=2026-10-14|link=https://discord.com/channels/1518634379604136178/1518692084729511967/1559906143373885462}} | ||
|- | |- | ||
| HBM clock, stock || 1,458 MHz (NDIV 54, VBIOS 92.00.67.00.01) or 1,728 MHz (NDIV 64, VBIOS 92.00.6D.00.0A) || 1,215 MHz || {{Source|handle=markshowa|date=2026-10-18|link=https://discord.com/channels/1518634379604136178/1534370835220004885/1561324784598384784|repo=https://github.com/amoghmunikote/170th-Street}}<br />{{Source|handle=c133|date=2026-10-20|link=https://discord.com/channels/1518634379604136178/1528826330454429706/1562161732137386127|repo=https://github.com/PixelML/club-170hx}} | | HBM clock, stock || 1,458 MHz (NDIV 54, VBIOS 92.00.67.00.01) or 1,728 MHz (NDIV 64, VBIOS 92.00.6D.00.0A) || 1,215 MHz || {{Source|handle=markshowa|date=2026-10-18|link=https://discord.com/channels/1518634379604136178/1534370835220004885/1561324784598384784|repo=https://github.com/amoghmunikote/170th-Street}}<br />{{Source|handle=c133|date=2026-10-20|link=https://discord.com/channels/1518634379604136178/1528826330454429706/1562161732137386127|repo=https://github.com/PixelML/club-170hx}} | ||
| Line 46: | Line 46: | ||
| Power connector || one EPS12V 8-pin || same || {{Source|handle=civiloid|date=2026-10-15|link=https://discord.com/channels/1518634379604136178/1534370835220004885/1560436414829559809}} | | Power connector || one EPS12V 8-pin || same || {{Source|handle=civiloid|date=2026-10-15|link=https://discord.com/channels/1518634379604136178/1534370835220004885/1560436414829559809}} | ||
|- | |- | ||
| Cooling || passive heatsink with IHS; forced airflow required || same; memory runs a few degrees warmer (41 C vs 37 C at about 33 W idle) || {{Source|handle=ghostinthedatacenter_87281|date=2026-10-13|link=https://discord.com/channels/1518634379604136178/1534370835220004885/ | | Cooling || passive heatsink with IHS; forced airflow required || same; memory runs a few degrees warmer (41 C vs 37 C at about 33 W idle) || {{Source|handle=ghostinthedatacenter_87281|date=2026-10-13|link=https://discord.com/channels/1518634379604136178/1534370835220004885/1559651382816936108}}<br />{{Source|handle=spectakkular|date=2026-10-27|link=https://discord.com/channels/1518634379604136178/1545982693903306813/1564563013053452412}} | ||
|- | |- | ||
| Price, reported (USD per card unless noted) || 100-300 before the unlock; 1,900 (eBay, 2026-10-01); 3,000-3,500 (2026-10-13); 3,300-3,600 delivered from China (2026-10-25) || about 200 before the unlock news; 20,000 RMB in China, about the same as an 8 GB card (2026-10-13); cheaper than the 8 GB card on eBay (2026-10-13); 1,200 for a "for parts" card that worked (2026-10-25) || {{Source|handle=cuddylac997|date=2026-10-07|link=https://discord.com/channels/1518634379604136178/1518695328008638516/1557383580106227724}}<br />{{Source|handle=djpandaman|date=2026-10-12|link=https://discord.com/channels/1518634379604136178/1528513735575081070/1559074773533724702}}<br />{{Source|handle=lenxtroz|date=2026-10-13|link=https://discord.com/channels/1518634379604136178/1528513735575081070/1559511799831863317}}<br />{{Source|handle=wakeupmbrc1|date=2026-10-13|link=https://discord.com/channels/1518634379604136178/1518692084729511967/1559705431155155044}}<br />{{Source|handle=froggao|date=2026-10-13|link=https://discord.com/channels/1518634379604136178/1518692084729511967/1559705220467007631}}<br />{{Source|handle=c133|date=2026-10-20|link=https://discord.com/channels/1518634379604136178/1528826330454429706/1562161732137386127|repo=https://github.com/PixelML/club-170hx}}<br />{{Source|handle=kellertreppe123|date=2026-10-25|link=https://discord.com/channels/1518634379604136178/1528513735575081070/1563910041487147110}}<br />{{Source|handle=programmerboi|date=2026-10-25|link=https://discord.com/channels/1518634379604136178/1534370835220004885/1563753581637537904}} | | Price, reported (USD per card unless noted) || 100-300 before the unlock; 1,900 (eBay, 2026-10-01); 3,000-3,500 (2026-10-13); 3,300-3,600 delivered from China (2026-10-25) || about 200 before the unlock news; 20,000 RMB in China, about the same as an 8 GB card (2026-10-13); cheaper than the 8 GB card on eBay (2026-10-13); 1,200 for a "for parts" card that worked (2026-10-25) || {{Source|handle=cuddylac997|date=2026-10-07|link=https://discord.com/channels/1518634379604136178/1518695328008638516/1557383580106227724}}<br />{{Source|handle=djpandaman|date=2026-10-12|link=https://discord.com/channels/1518634379604136178/1528513735575081070/1559074773533724702}}<br />{{Source|handle=lenxtroz|date=2026-10-13|link=https://discord.com/channels/1518634379604136178/1528513735575081070/1559511799831863317}}<br />{{Source|handle=wakeupmbrc1|date=2026-10-13|link=https://discord.com/channels/1518634379604136178/1518692084729511967/1559705431155155044}}<br />{{Source|handle=froggao|date=2026-10-13|link=https://discord.com/channels/1518634379604136178/1518692084729511967/1559705220467007631}}<br />{{Source|handle=c133|date=2026-10-20|link=https://discord.com/channels/1518634379604136178/1528826330454429706/1562161732137386127|repo=https://github.com/PixelML/club-170hx}}<br />{{Source|handle=kellertreppe123|date=2026-10-25|link=https://discord.com/channels/1518634379604136178/1528513735575081070/1563910041487147110}}<br />{{Source|handle=programmerboi|date=2026-10-25|link=https://discord.com/channels/1518634379604136178/1534370835220004885/1563753581637537904}} | ||
| Line 55: | Line 55: | ||
== Which one to choose == | == Which one to choose == | ||
{{Status|Reported}} | {{Status|Reported}} | ||
* If you run LLMs, memory per card decides what fits. The 8 GB card gives 64 GB; the 10 GB card gives 40 GB (48 GB if lucky) plus 128 SMs and Gen3. {{Source|handle=68x0|date=2026-10-26|link=https://discord.com/channels/1518634379604136178/1518695328008638516/1564104089560678491}} PCIe only matters sometimes; VRAM matters basically all of the time. {{Source|handle=sojun80|date=2026-10-20|link=https://discord.com/channels/1518634379604136178/1518695328008638516/1562100982584967180}} | * If you run LLMs, memory per card decides what fits. The 8 GB card gives 64 GB; the 10 GB card gives 40 GB (48 GB if lucky) plus 128 SMs and Gen3. {{Source|handle=68x0|date=2026-10-26|link=https://discord.com/channels/1518634379604136178/1518695328008638516/1564104089560678491}} PCIe only matters sometimes; VRAM matters basically all of the time. {{Source|handle=sojun80|date=2026-10-20|link=https://discord.com/channels/1518634379604136178/1518695328008638516/1562100982584967180}} In the same thread @shudza answered that "128 SM matters too", and @programmerboi that VRAM "only matters sometimes": his pretraining work is compute bound. {{Source|handle=shudza|date=2026-10-20|link=https://discord.com/channels/1518634379604136178/1518695328008638516/1562102220525404221}}<br />{{Source|handle=programmerboi|date=2026-10-20|link=https://discord.com/channels/1518634379604136178/1518695328008638516/1562105200980795524}}<br />{{Source|handle=programmerboi|date=2026-10-20|link=https://discord.com/channels/1518634379604136178/1518695328008638516/1562105604984807527}} | ||
* If you do not need tensor parallel or training, the 8 GB card serves better; think in memory per watt. {{Source|handle=djpandaman|date=2026-10-07|link=https://discord.com/channels/1518634379604136178/1518692084729511967/1557244136157872133}} The 10 GB card gets more compute and interconnect bandwidth today, the 8 GB card more total memory. {{Source|handle=_arcanum_|date=2026-10-07|link=https://discord.com/channels/1518634379604136178/1518692084729511967/1557246965526298625}} | * If you do not need tensor parallel or training, the 8 GB card serves better; think in memory per watt. {{Source|handle=djpandaman|date=2026-10-07|link=https://discord.com/channels/1518634379604136178/1518692084729511967/1557244136157872133}} The 10 GB card gets more compute and interconnect bandwidth today, the 8 GB card more total memory. {{Source|handle=_arcanum_|date=2026-10-07|link=https://discord.com/channels/1518634379604136178/1518692084729511967/1557246965526298625}} | ||
* If you already own 8 GB cards, trading them for 10 GB cards is not worth it for most users. {{Source|handle=lia_oc|date=2026-10-22|link=https://discord.com/channels/1518634379604136178/1518695328008638516/1562931544371372094}} | * If you already own 8 GB cards, trading them for 10 GB cards is not worth it for most users: "realistically for most of us, the bottleneck is still VRAM" (@lia_oc); @reala7096: "I wouldn't trade my 8GB cards for 10GB ones". {{Source|handle=lia_oc|date=2026-10-22|link=https://discord.com/channels/1518634379604136178/1518695328008638516/1562931544371372094}}<br />{{Source|handle=lia_oc|date=2026-10-22|link=https://discord.com/channels/1518634379604136178/1518695328008638516/1562931590353653774}}<br />{{Source|handle=reala7096|date=2026-10-22|link=https://discord.com/channels/1518634379604136178/1518695328008638516/1562931684616441979}} | ||
* Mixing both in one rig works with a per-card unlock profile, but a multi-card setup runs at the pace of the slowest card and link. {{Source|handle=djpandaman|date=2026-10-07|link=https://discord.com/channels/1518634379604136178/1518692084729511967/1557238098662793267}} | * Mixing both in one rig works with a per-card unlock profile, but a multi-card setup runs at the pace of the slowest card and link. {{Source|handle=djpandaman|date=2026-10-07|link=https://discord.com/channels/1518634379604136178/1518692084729511967/1557238098662793267}} | ||
Latest revision as of 10:02, 30 October 2026
Reported 8 GB and 10 GB
The 8 GB card and the 10 GB card side by side. Prices are reported figures from the Discord on the dates given, not quotes. How to tell which card you have: Which card do I have.
| Property | 8 GB card | 10 GB card | Source |
|---|---|---|---|
| PCI ID | 10de:20c2 (subsystem 1585) | 10de:2082 (subsystem 1557) | Source: repo github.com/amoghmunikote/170th-Street by amoghmunikote on GitHub, posted by @markshowa, 2026-10-18 |
| HBM | SK hynix HBM2e, 16 of 24 FBPA, 4096-bit | Samsung HBM2e, 20 of 24 FBPA, 5120-bit | Source: repo github.com/matstatman/gpcprobe by matstatman on GitHub, posted by @matstatman, 2026-10-20 Source: @darkmaniac7, 2026-10-30 (site owner, from his own cards); earlier posts that conflict on HBM2 versus HBM2e: Which card do I have#Memory vendor |
| Memory, stock | 8 GB | 10 GB | Source: repo github.com/amoghmunikote/170th-Street by amoghmunikote on GitHub, posted by @markshowa, 2026-10-18 |
| Memory, unlocked | 64 GB (65536 MiB) Proven | 40 GB (40960 MiB) Proven; 48 GB Reported if the sixth stack is healthy, about 1 card in 5 (6 of 30 cards seen) | Source: @zorg33, 2026-10-07 Source: repo github.com/amoghmunikote/cmpunlocker by amoghmunikote on GitHub, posted by @rh3dai, 2026-10-08 Source: @newman_08462, 2026-10-09 Source: @djpandaman, 2026-10-17 Source: @programmerboi, 2026-10-25 |
| Memory beyond that | 80 or 96 GB rumoured, nothing shown | 80 or 96 GB not achieved Contested: the evidence points to aliasing, not real memory | Source: @djpandaman, 2026-10-12 Source: @colo2442, 2026-10-14 |
| HBM clock, stock | 1,458 MHz (NDIV 54, VBIOS 92.00.67.00.01) or 1,728 MHz (NDIV 64, VBIOS 92.00.6D.00.0A) | 1,215 MHz | Source: repo github.com/amoghmunikote/170th-Street by amoghmunikote on GitHub, posted by @markshowa, 2026-10-18 Source: repo github.com/PixelML/club-170hx by PixelML on GitHub, posted by @c133, 2026-10-20 |
| Bandwidth, theoretical | 1.49 TB/s at 1,458 MHz; 1.77 TB/s at 1,728 MHz; about 1.94 TB/s at NDIV 70 | 1.56 TB/s; about 1.87 TB/s with all six stacks | Source: repo github.com/amoghmunikote/170th-Street by amoghmunikote on GitHub, posted by @markshowa, 2026-10-18; higher values computed as bus width x 2 x clock |
| Bandwidth, reported | about 1,493 GB/s on an unlocked card (tool readout) | 1,741 GB/s on an unlocked card at stock clocks | Source: @cuddylac997, 2026-10-07 Source: @kio063218, 2026-10-16 |
| HBM overclock | NDIV 68-70 typical, 74 fails | not reported | Source: @alanoo, 2026-10-14 |
| SMs | 70 stock, 74 unlocked | 70 stock, 74 public; 112-128 after the RIR burn | Source: @lia_oc, 2026-10-22 Source: @wakeupmbrc1, 2026-10-27 |
| RIR fuse records | 15 of 16 used, last one partly burned: no burn path | 16 of 16 free | Source: @wakeupmbrc1, 2026-10-13 Source: @djpandaman, 2026-10-17 |
| PCIe link, stock | Gen1 x4 | Gen1 x4 | Source: repo github.com/amoghmunikote/170th-Street by amoghmunikote on GitHub, posted by @markshowa, 2026-10-18 |
| PCIe link, unlocked | Gen2; Gen3 blocked by fuse | Gen2; Gen3 after the RIR burn; Gen4 reported, not public | Source: @djpandaman, 2026-10-09 Source: @wakeupmbrc1, 2026-10-25 |
| PCIe width | x16 needs the PCIe capacitor mod | same | Source: @lxtend, 2026-10-17 |
| NVLink | no path | not public; needs a board-level mod | Source: @lia_oc, 2026-10-22 |
| ECC | yes (cmpunlocker v0.5) | yes | Source: @reala7096, 2026-10-22 |
| Windows | no Windows unlock reported; cmpunlocker is Linux only (an A800 device-ID change is an untested idea) | yes, with the A100 40G / A800 40G driver | Source: repo github.com/amoghmunikote/cmpunlocker by amoghmunikote on GitHub, posted by @rh3dai, 2026-10-08 Source: @wakeupmbrc1, 2026-10-11 |
| Power limit (VBIOS) | 250 W (92.00.67.00.01) or 300 W (92.00.6D.00.0A) | 250 W; 300 W reported with a no-flash VBIOS loader | Source: @zagitta, 2026-10-12 Source: @lxtend, 2026-10-20 |
| Power, in use | most owners cap at 140-200 W | power-bound above 80 SMs | Source: @djpandaman, 2026-10-13 Source: repo github.com/PixelML/club-170hx by PixelML on GitHub, posted by @c133, 2026-10-20 |
| Power connector | one EPS12V 8-pin | same | Source: @civiloid, 2026-10-15 |
| Cooling | passive heatsink with IHS; forced airflow required | same; memory runs a few degrees warmer (41 C vs 37 C at about 33 W idle) | Source: @ghostinthedatacenter_87281, 2026-10-13 Source: @spectakkular, 2026-10-27 |
| Price, reported (USD per card unless noted) | 100-300 before the unlock; 1,900 (eBay, 2026-10-01); 3,000-3,500 (2026-10-13); 3,300-3,600 delivered from China (2026-10-25) | about 200 before the unlock news; 20,000 RMB in China, about the same as an 8 GB card (2026-10-13); cheaper than the 8 GB card on eBay (2026-10-13); 1,200 for a "for parts" card that worked (2026-10-25) | Source: @cuddylac997, 2026-10-07 Source: @djpandaman, 2026-10-12 Source: @lenxtroz, 2026-10-13 Source: @wakeupmbrc1, 2026-10-13 Source: @froggao, 2026-10-13 Source: repo github.com/PixelML/club-170hx by PixelML on GitHub, posted by @c133, 2026-10-20 Source: @kellertreppe123, 2026-10-25 Source: @programmerboi, 2026-10-25 |
| Good for | LLM inference: memory matters most | training, image and video generation, diffusion; compute-bound work | Source: @wakeupmbrc1, 2026-10-13 |
Which one to choose
Reported
- If you run LLMs, memory per card decides what fits. The 8 GB card gives 64 GB; the 10 GB card gives 40 GB (48 GB if lucky) plus 128 SMs and Gen3. Source: @68x0, 2026-10-26 PCIe only matters sometimes; VRAM matters basically all of the time. Source: @sojun80, 2026-10-20 In the same thread @shudza answered that "128 SM matters too", and @programmerboi that VRAM "only matters sometimes": his pretraining work is compute bound. Source: @shudza, 2026-10-20
Source: @programmerboi, 2026-10-20
Source: @programmerboi, 2026-10-20 - If you do not need tensor parallel or training, the 8 GB card serves better; think in memory per watt. Source: @djpandaman, 2026-10-07 The 10 GB card gets more compute and interconnect bandwidth today, the 8 GB card more total memory. Source: @_arcanum_, 2026-10-07
- If you already own 8 GB cards, trading them for 10 GB cards is not worth it for most users: "realistically for most of us, the bottleneck is still VRAM" (@lia_oc); @reala7096: "I wouldn't trade my 8GB cards for 10GB ones". Source: @lia_oc, 2026-10-22
Source: @lia_oc, 2026-10-22
Source: @reala7096, 2026-10-22 - Mixing both in one rig works with a per-card unlock profile, but a multi-card setup runs at the pace of the slowest card and link. Source: @djpandaman, 2026-10-07